zrepl/zfs/zfs.go

1922 lines
54 KiB
Go
Raw Permalink Normal View History

2017-04-14 19:26:32 +02:00
package zfs
import (
"bufio"
"bytes"
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
"context"
"encoding/json"
"fmt"
2017-04-26 20:25:53 +02:00
"io"
"math"
"os"
2017-04-26 20:25:53 +02:00
"os/exec"
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
"regexp"
"sort"
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
"strconv"
2017-04-26 20:25:53 +02:00
"strings"
"sync"
"github.com/pkg/errors"
"github.com/prometheus/client_golang/prometheus"
"github.com/zrepl/zrepl/util/circlog"
"github.com/zrepl/zrepl/util/envconst"
"github.com/zrepl/zrepl/util/nodefault"
zfsprop "github.com/zrepl/zrepl/zfs/property"
"github.com/zrepl/zrepl/zfs/zfscmd"
)
type DatasetPath struct {
comps []string
}
func (p *DatasetPath) ToString() string {
return strings.Join(p.comps, "/")
}
func (p *DatasetPath) Empty() bool {
return len(p.comps) == 0
}
func (p *DatasetPath) Extend(extend *DatasetPath) {
p.comps = append(p.comps, extend.comps...)
}
func (p *DatasetPath) HasPrefix(prefix *DatasetPath) bool {
if len(prefix.comps) > len(p.comps) {
return false
}
for i := range prefix.comps {
if prefix.comps[i] != p.comps[i] {
return false
}
}
return true
}
func (p *DatasetPath) TrimPrefix(prefix *DatasetPath) {
if !p.HasPrefix(prefix) {
return
}
prelen := len(prefix.comps)
newlen := len(p.comps) - prelen
oldcomps := p.comps
p.comps = make([]string, newlen)
for i := 0; i < newlen; i++ {
p.comps[i] = oldcomps[prelen+i]
}
}
func (p *DatasetPath) TrimNPrefixComps(n int) {
if len(p.comps) < n {
n = len(p.comps)
}
if n == 0 {
return
}
p.comps = p.comps[n:]
}
func (p DatasetPath) Equal(q *DatasetPath) bool {
if len(p.comps) != len(q.comps) {
return false
}
for i := range p.comps {
if p.comps[i] != q.comps[i] {
return false
}
}
return true
}
func (p *DatasetPath) Length() int {
return len(p.comps)
}
func (p *DatasetPath) Copy() (c *DatasetPath) {
c = &DatasetPath{}
c.comps = make([]string, len(p.comps))
copy(c.comps, p.comps)
return
}
func (p *DatasetPath) MarshalJSON() ([]byte, error) {
return json.Marshal(p.comps)
}
func (p *DatasetPath) UnmarshalJSON(b []byte) error {
p.comps = make([]string, 0)
return json.Unmarshal(b, &p.comps)
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
func (p *DatasetPath) Pool() (string, error) {
if len(p.comps) < 1 {
return "", fmt.Errorf("dataset path does not have a pool component")
}
return p.comps[0], nil
}
func NewDatasetPath(s string) (p *DatasetPath, err error) {
p = &DatasetPath{}
if s == "" {
p.comps = make([]string, 0)
return p, nil // the empty dataset path
}
const FORBIDDEN = "@#|\t<>*"
/* Documentation of allowed characters in zfs names:
https://docs.oracle.com/cd/E19253-01/819-5461/gbcpt/index.html
Space is missing in the oracle list, but according to
https://github.com/zfsonlinux/zfs/issues/439
there is evidence that it was intentionally allowed
*/
if strings.ContainsAny(s, FORBIDDEN) {
err = fmt.Errorf("contains forbidden characters (any of '%s')", FORBIDDEN)
return
}
p.comps = strings.Split(s, "/")
if p.comps[len(p.comps)-1] == "" {
err = fmt.Errorf("must not end with a '/'")
return
}
return
2017-04-26 18:36:01 +02:00
}
func toDatasetPath(s string) *DatasetPath {
p, err := NewDatasetPath(s)
if err != nil {
panic(err)
}
return p
}
type ZFSError struct {
Stderr []byte
WaitErr error
}
func (e *ZFSError) Error() string {
return fmt.Sprintf("zfs exited with error: %s\nstderr:\n%s", e.WaitErr.Error(), e.Stderr)
}
var ZFS_BINARY string = "zfs"
func ZFSList(ctx context.Context, properties []string, zfsArgs ...string) (res [][]string, err error) {
2017-04-14 19:26:32 +02:00
args := make([]string, 0, 4+len(zfsArgs))
args = append(args,
"list", "-H", "-p",
"-o", strings.Join(properties, ","))
args = append(args, zfsArgs...)
ctx, cancel := context.WithCancel(ctx)
defer cancel()
cmd := zfscmd.CommandContext(ctx, ZFS_BINARY, args...)
stdout, stderrBuf, err := cmd.StdoutPipeWithErrorBuf()
if err != nil {
return
}
if err = cmd.Start(); err != nil {
return
}
// in case we return early, we want to kill the zfs list process and wait for it to exit
defer func() {
_ = cmd.Wait()
}()
defer cancel()
s := bufio.NewScanner(stdout)
buf := make([]byte, 1024)
s.Buffer(buf, 0)
res = make([][]string, 0)
for s.Scan() {
fields := strings.SplitN(s.Text(), "\t", len(properties))
if len(fields) != len(properties) {
err = errors.New("unexpected output")
return
}
res = append(res, fields)
}
2017-04-26 20:25:53 +02:00
if waitErr := cmd.Wait(); waitErr != nil {
err := &ZFSError{
Stderr: stderrBuf.Bytes(),
WaitErr: waitErr,
}
return nil, err
}
return
}
2017-05-07 12:18:54 +02:00
type ZFSListResult struct {
Fields []string
Err error
}
// ZFSListChan executes `zfs list` and sends the results to the `out` channel.
// The `out` channel is always closed by ZFSListChan:
// If an error occurs, it is closed after sending a result with the Err field set.
// If no error occurs, it is just closed.
// If the operation is cancelled via context, the channel is just closed.
//
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
// If notExistHint is not nil and zfs exits with an error,
// the stderr is attempted to be interpreted as a *DatasetDoesNotExist error.
//
// However, if callers do not drain `out` or cancel via `ctx`, the process will leak either running because
// IO is pending or as a zombie.
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
func ZFSListChan(ctx context.Context, out chan ZFSListResult, properties []string, notExistHint *DatasetPath, zfsArgs ...string) {
defer close(out)
args := make([]string, 0, 4+len(zfsArgs))
args = append(args,
"list", "-H", "-p",
"-o", strings.Join(properties, ","))
args = append(args, zfsArgs...)
sendResult := func(fields []string, err error) (done bool) {
select {
case <-ctx.Done():
return true
case out <- ZFSListResult{fields, err}:
return false
}
}
ctx, cancel := context.WithCancel(ctx)
defer cancel()
cmd := zfscmd.CommandContext(ctx, ZFS_BINARY, args...)
stdout, stderrBuf, err := cmd.StdoutPipeWithErrorBuf()
if err != nil {
sendResult(nil, err)
return
}
if err = cmd.Start(); err != nil {
sendResult(nil, err)
return
}
defer func() {
// discard the error, this defer is only relevant if we return while parsing the output
// in which case we'll return an 'unexpected output' error and not the exit status
_ = cmd.Wait()
}()
defer cancel() // in case we return before our regular call to cmd.Wait(), kill the zfs list process
s := bufio.NewScanner(stdout)
buf := make([]byte, 1024) // max line length
s.Buffer(buf, 0)
for s.Scan() {
fields := strings.SplitN(s.Text(), "\t", len(properties))
if len(fields) != len(properties) {
sendResult(nil, errors.New("unexpected output"))
return
}
if sendResult(fields, nil) {
return
}
}
if err := cmd.Wait(); err != nil {
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
if _, ok := err.(*exec.ExitError); ok {
var enotexist *DatasetDoesNotExist
if notExistHint != nil {
enotexist = tryDatasetDoesNotExist(notExistHint.ToString(), stderrBuf.Bytes())
}
if enotexist != nil {
sendResult(nil, enotexist)
} else {
sendResult(nil, &ZFSError{
Stderr: stderrBuf.Bytes(),
WaitErr: err,
})
}
} else {
sendResult(nil, &ZFSError{WaitErr: err})
}
return
}
if s.Err() != nil {
sendResult(nil, s.Err())
return
}
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
// FIXME replace with EntityNamecheck
func validateZFSFilesystem(fs string) error {
if len(fs) < 1 {
return errors.New("filesystem path must have length > 0")
}
return nil
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
// v must not be nil and be already validated
func absVersion(fs string, v *ZFSSendArgVersion) (full string, err error) {
if err := validateZFSFilesystem(fs); err != nil {
return "", err
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
return fmt.Sprintf("%s%s", fs, v.RelName), nil
}
func pipeWithCapacityHint(capacity int) (r, w *os.File, err error) {
if capacity < 0 {
panic(fmt.Sprintf("capacity must be non-negative, got %v", capacity))
}
stdoutReader, stdoutWriter, err := os.Pipe()
if err != nil {
return nil, nil, err
}
trySetPipeCapacity(stdoutWriter, capacity)
return stdoutReader, stdoutWriter, nil
}
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
type sendStreamState int
const (
sendStreamOpen sendStreamState = iota
sendStreamClosed
)
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
type SendStream struct {
cmd *zfscmd.Cmd
kill context.CancelFunc
stdoutReader io.ReadCloser // not *os.File for mocking during platformtest
stderrBuf *circlog.CircularLog
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
mtx sync.Mutex
state sendStreamState
exitErr *ZFSError
}
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
func (s *SendStream) Read(p []byte) (n int, _ error) {
s.mtx.Lock()
defer s.mtx.Unlock()
switch s.state {
case sendStreamClosed:
return 0, os.ErrClosed
case sendStreamOpen:
n, readErr := s.stdoutReader.Read(p)
if readErr != nil {
debug("sendStream: read: readErr=%T %s", readErr, readErr)
if readErr == io.EOF {
// io.EOF must be bubbled up as is so that consumers can handle it properly.
return n, readErr
}
// Assume that the error is not retryable.
// Try to kill now so that we can return a nice *ZFSError with captured stderr.
// If the kill doesn't work, it doesn't matter because the caller must by contract call Close() anyways.
killErr := s.killAndWait()
debug("sendStream: read: killErr=%T %s", killErr, killErr)
if killErr == nil {
s.state = sendStreamClosed
return n, s.exitErr // return the nice error
} else {
// we remain open so that we retry
return n, readErr // return the normal error
}
}
return n, readErr
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
default:
panic("unreachable")
}
}
func (s *SendStream) Close() error {
debug("sendStream: close called")
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
s.mtx.Lock()
defer s.mtx.Unlock()
switch s.state {
case sendStreamOpen:
err := s.killAndWait()
if err != nil {
return err
} else {
s.state = sendStreamClosed
return nil
}
case sendStreamClosed:
return os.ErrClosed
default:
panic("unreachable")
}
}
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
// returns nil iff the child process is gone (has been successfully waited upon)
// in that case, s.exitErr is set
func (s *SendStream) killAndWait() error {
debug("sendStream: killAndWait enter")
defer debug("sendStream: killAndWait leave")
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
// send SIGKILL
s.kill()
// Close our read-end of the pipe.
//
// We must do this before .Wait() because in some (not all) versions/build configs of ZFS,
// `zfs send` uses a separate kernel thread (taskq) to write the send stream (function `dump_bytes`).
// The `zfs send` thread then waits uinterruptably for the taskq thread to finish the write.
// And signalling the `zfs send` thread doesn't propagate to the taskq thread.
// So we end up in a state where we .Wait() forever.
// (See https://github.com/openzfs/zfs/issues/12500 and
// https://github.com/zrepl/zrepl/issues/495#issuecomment-902530043)
//
// By closing our read end of the pipe before .Wait(), we unblock the taskq thread if there is any.
// If there is no separate taskq thread, the SIGKILL to `zfs end` would suffice and be most precise,
// but due to the circumstances above, there is no other portable & robust way.
//
// However, the fallout from closing the pipe is that (in non-taskq builds) `zfs sends` will get a SIGPIPE.
// And on Linux, that SIGPIPE appears to win over the previously issued SIGKILL.
// And thus, on Linux, the `zfs send` will be killed by the default SIGPIPE handler.
// We can observe this in the WaitStatus below.
// This behavior is slightly annoying because the *exec.ExitError's message ("signal: broken pipe")
// isn't as clear as ("signal: killed").
// However, it seems like we just have to live with that. (covered by platformtest)
var closePipeErr error
if s.stdoutReader != nil {
closePipeErr = s.stdoutReader.Close()
if closePipeErr == nil {
// avoid double-closes in case waiting below doesn't work
// and someone attempts Close again
s.stdoutReader = nil
} else {
return closePipeErr
}
}
waitErr := s.cmd.Wait()
// distinguish between ExitError (which is actually a non-problem for us)
// vs failed wait syscall (for which we give upper layers the chance to retyr)
var exitErr *exec.ExitError
if waitErr != nil {
if ee, ok := waitErr.(*exec.ExitError); ok {
exitErr = ee
} else {
return waitErr
}
}
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
// invariant: at this point, the child is gone and we cleaned up everything related to the SendStream
if exitErr != nil {
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
// zfs send exited with an error or was killed by a signal.
s.exitErr = &ZFSError{
Stderr: []byte(s.stderrBuf.String()),
WaitErr: exitErr,
}
} else {
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
// zfs send exited successfully (we know that since waitErr was either nil or wasn't an *exec.ExitError)
s.exitErr = nil
}
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
return nil
}
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
func (s *SendStream) TestOnly_ReplaceStdoutReader(f io.ReadCloser) (prev io.ReadCloser) {
prev = s.stdoutReader
s.stdoutReader = f
return prev
}
zfs: rewrite SendStream, fix bug in Close() on FreeBSD, add platformtests This commit was motivated by https://github.com/zrepl/zrepl/issues/495 where, on FreeBSD with OpenZFS 2.0, a SendStream.Close() call might wait indefinitely for `zfs send` to exit. The reason is that, due to the refactoring done for redacted send & recv (https://github.com/openzfs/zfs/commit/30af21b02569ac192f52ce6e6511015f8a8d5729), the `dump_bytes` function, which writes to the pipe, executes in a separate thread (synctask taskq) iff not `HAVE_LARGE_STACKS`. The `zfs send` process/thread waits for that taskq thread using an uninterruptible primitive. So when we SIGKILL `zfs send`, that signal doesn't reach the right thread to interrupt the pipe write. Theoretically this affects both Linux and FreeBSD, but most Linux users `HAVE_LARGE_STACKS` and since https://github.com/penzfs/zfs/pull/12350/files OpenZFS on FreeBSD `HAVE_LARGE_STACKS` as well. However, at least until FreeBSD 13.1, possibly for the entire 13 lifecycle, we're going to have to live with that oddity. Measures taken in this commit: - Report the behavior as an upstream bug https://github.com/openzfs/zfs/issues/12500 - Change SendStream code so that it closes zrepl's read-end of the pipe (see comment in code) - Clean up and make explicit SendStream's state handling - Write extensive platformtests for SendStream - They pass on my Linux install and on FreeBSD 12 - FreeBSD 13 still needs testing. fixes https://github.com/zrepl/zrepl/issues/495
2021-08-20 17:43:28 +02:00
func (s *SendStream) TestOnly_ExitErr() *ZFSError { return s.exitErr }
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
// NOTE: When updating this struct, make sure to update funcs Validate ValidateCorrespondsToResumeToken
type ZFSSendArgVersion struct {
RelName string
GUID uint64
}
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
func (v ZFSSendArgVersion) GetGuid() uint64 { return v.GUID }
func (v ZFSSendArgVersion) ToSendArgVersion() ZFSSendArgVersion { return v }
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
func (v ZFSSendArgVersion) ValidateInMemory(fs string) error {
if fs == "" {
panic(fs)
}
if len(v.RelName) == 0 {
return errors.New("`RelName` must not be empty")
}
var et EntityType
switch v.RelName[0] {
case '@':
et = EntityTypeSnapshot
case '#':
et = EntityTypeBookmark
default:
return fmt.Errorf("`RelName` field must start with @ or #, got %q", v.RelName)
}
full := v.fullPathUnchecked(fs)
if err := EntityNamecheck(full, et); err != nil {
return err
}
return nil
}
func (v ZFSSendArgVersion) mustValidateInMemory(fs string) {
if err := v.ValidateInMemory(fs); err != nil {
panic(err)
}
}
// fs must be not empty
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
func (a ZFSSendArgVersion) ValidateExistsAndGetVersion(ctx context.Context, fs string) (v FilesystemVersion, _ error) {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if err := a.ValidateInMemory(fs); err != nil {
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
return v, nil
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
realVersion, err := ZFSGetFilesystemVersion(ctx, a.FullPath(fs))
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if err != nil {
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
return v, err
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
if realVersion.Guid != a.GUID {
return v, fmt.Errorf("`GUID` field does not match real dataset's GUID: %q != %q", realVersion.Guid, a.GUID)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
return realVersion, nil
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
func (a ZFSSendArgVersion) ValidateExists(ctx context.Context, fs string) error {
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
_, err := a.ValidateExistsAndGetVersion(ctx, fs)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
return err
}
func (v ZFSSendArgVersion) FullPath(fs string) string {
v.mustValidateInMemory(fs)
return v.fullPathUnchecked(fs)
}
func (v ZFSSendArgVersion) fullPathUnchecked(fs string) string {
return fmt.Sprintf("%s%s", fs, v.RelName)
}
func (v ZFSSendArgVersion) IsSnapshot() bool {
v.mustValidateInMemory("unimportant")
return v.RelName[0] == '@'
}
func (v ZFSSendArgVersion) MustBeBookmark() {
v.mustValidateInMemory("unimportant")
if v.RelName[0] != '#' {
panic(fmt.Sprintf("must be bookmark, got %q", v.RelName))
}
}
// When updating this struct, check Validate and ValidateCorrespondsToResumeToken (POTENTIALLY SECURITY SENSITIVE)
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
type ZFSSendArgsUnvalidated struct {
FS string
From, To *ZFSSendArgVersion // From may be nil
ZFSSendFlags
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
type ZFSSendArgsValidated struct {
ZFSSendArgsUnvalidated
FromVersion *FilesystemVersion
ToVersion FilesystemVersion
}
type ZFSSendFlags struct {
Encrypted *nodefault.Bool
Properties bool
BackupProperties bool
Raw bool
LargeBlocks bool
Compressed bool
EmbeddedData bool
Saved bool
// Preferred if not empty
ResumeToken string // if not nil, must match what is specified in From, To (covered by ValidateCorrespondsToResumeToken)
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
type zfsSendArgsValidationContext struct {
encEnabled *nodefault.Bool
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
type ZFSSendArgsValidationErrorCode int
const (
ZFSSendArgsGenericValidationError ZFSSendArgsValidationErrorCode = 1 + iota
ZFSSendArgsEncryptedSendRequestedButFSUnencrypted
ZFSSendArgsFSEncryptionCheckFail
ZFSSendArgsResumeTokenMismatch
)
type ZFSSendArgsValidationError struct {
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
Args ZFSSendArgsUnvalidated
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
What ZFSSendArgsValidationErrorCode
Msg error
}
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
func newValidationError(sendArgs ZFSSendArgsUnvalidated, what ZFSSendArgsValidationErrorCode, cause error) *ZFSSendArgsValidationError {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
return &ZFSSendArgsValidationError{sendArgs, what, cause}
}
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
func newGenericValidationError(sendArgs ZFSSendArgsUnvalidated, cause error) *ZFSSendArgsValidationError {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
return &ZFSSendArgsValidationError{sendArgs, ZFSSendArgsGenericValidationError, cause}
}
func (e ZFSSendArgsValidationError) Error() string {
return e.Msg.Error()
}
// - Recursively call Validate on each field.
// - Make sure that if ResumeToken != "", it reflects the same operation as the other parameters would.
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
//
// This function is not pure because GUIDs are checked against the local host's datasets.
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
func (a ZFSSendArgsUnvalidated) Validate(ctx context.Context) (v ZFSSendArgsValidated, _ error) {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if dp, err := NewDatasetPath(a.FS); err != nil || dp.Length() == 0 {
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
return v, newGenericValidationError(a, fmt.Errorf("`FS` must be a valid non-zero dataset path"))
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
if a.To == nil {
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
return v, newGenericValidationError(a, fmt.Errorf("`To` must not be nil"))
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
toVersion, err := a.To.ValidateExistsAndGetVersion(ctx, a.FS)
if err != nil {
return v, newGenericValidationError(a, errors.Wrap(err, "`To` invalid"))
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
var fromVersion *FilesystemVersion
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if a.From != nil {
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
fromV, err := a.From.ValidateExistsAndGetVersion(ctx, a.FS)
if err != nil {
return v, newGenericValidationError(a, errors.Wrap(err, "`From` invalid"))
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
fromVersion = &fromV
// fallthrough
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
if err := a.ZFSSendFlags.Validate(); err != nil {
return v, newGenericValidationError(a, errors.Wrap(err, "send flags invalid"))
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
valCtx := &zfsSendArgsValidationContext{}
fsEncrypted, err := ZFSGetEncryptionEnabled(ctx, a.FS)
if err != nil {
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
return v, newValidationError(a, ZFSSendArgsFSEncryptionCheckFail,
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
errors.Wrapf(err, "cannot check whether filesystem %q is encrypted", a.FS))
}
valCtx.encEnabled = &nodefault.Bool{B: fsEncrypted}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if a.Encrypted.B && !fsEncrypted {
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
return v, newValidationError(a, ZFSSendArgsEncryptedSendRequestedButFSUnencrypted,
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
errors.Errorf("encrypted send requested, but filesystem %q is not encrypted", a.FS))
}
if err := a.validateEncryptionFlagsCorrespondToResumeToken(ctx, valCtx); err != nil {
return v, newValidationError(a, ZFSSendArgsResumeTokenMismatch, err)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
return ZFSSendArgsValidated{
ZFSSendArgsUnvalidated: a,
FromVersion: fromVersion,
ToVersion: toVersion,
}, nil
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
func (f ZFSSendFlags) Validate() error {
if err := f.Encrypted.ValidateNoDefault(); err != nil {
return errors.Wrap(err, "flag `Encrypted` invalid")
}
return nil
}
// If ResumeToken is empty, builds a command line with the flags specified.
// If ResumeToken is not empty, build a command line with just `-t {{.ResumeToken}}`.
//
// SECURITY SENSITIVE it is the caller's responsibility to ensure that a.Encrypted semantics
// hold for the file system that will be sent with the send flags returned by this function
func (a ZFSSendFlags) buildSendFlagsUnchecked() []string {
args := make([]string, 0)
// ResumeToken takes precedence, we assume that it has been validated
// to reflect what is described by the other fields.
if a.ResumeToken != "" {
args = append(args, "-t", a.ResumeToken)
return args
}
if a.Encrypted.B || a.Raw {
args = append(args, "-w")
}
if a.Properties {
args = append(args, "-p")
}
if a.BackupProperties {
args = append(args, "-b")
}
if a.LargeBlocks {
args = append(args, "-L")
}
if a.Compressed {
args = append(args, "-c")
}
if a.EmbeddedData {
args = append(args, "-e")
}
if a.Saved {
args = append(args, "-S")
}
return args
}
func (a ZFSSendArgsValidated) buildSendCommandLine() ([]string, error) {
flags := a.buildSendFlagsUnchecked()
if a.ZFSSendFlags.ResumeToken != "" {
return flags, nil
}
toV, err := absVersion(a.FS, a.To)
if err != nil {
return nil, err
}
fromV := ""
if a.From != nil {
fromV, err = absVersion(a.FS, a.From)
if err != nil {
return nil, err
}
}
if fromV == "" { // Initial
flags = append(flags, toV)
} else {
flags = append(flags, "-i", fromV, toV)
}
return flags, nil
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
type ZFSSendArgsResumeTokenMismatchError struct {
What ZFSSendArgsResumeTokenMismatchErrorCode
Err error
}
func (e *ZFSSendArgsResumeTokenMismatchError) Error() string { return e.Err.Error() }
type ZFSSendArgsResumeTokenMismatchErrorCode int
// The format is ZFSSendArgsResumeTokenMismatch+WhatIsWrongInToken
const (
ZFSSendArgsResumeTokenMismatchGeneric ZFSSendArgsResumeTokenMismatchErrorCode = 1 + iota
ZFSSendArgsResumeTokenMismatchEncryptionNotSet // encryption not set in token but required by send args
ZFSSendArgsResumeTokenMismatchEncryptionSet // encryption not set in token but not required by send args
ZFSSendArgsResumeTokenMismatchFilesystem
)
func (c ZFSSendArgsResumeTokenMismatchErrorCode) fmt(format string, args ...interface{}) *ZFSSendArgsResumeTokenMismatchError {
return &ZFSSendArgsResumeTokenMismatchError{
What: c,
Err: fmt.Errorf(format, args...),
}
}
// Validate that the encryption settings specified in `a` correspond to the encryption settings encoded in the resume token.
//
// This is SECURITY SENSITIVE:
// It is possible for an attacker to craft arbitrary resume tokens.
// Those malicious resume tokens could encode different parameters in the resume token than expected:
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
// for example, they may specify another file system (e.g. the filesystem with secret data) or request unencrypted send instead of encrypted raw send.
//
// Note that we don't check correspondence of all other send flags because
// a) the resume token does not capture all send flags (e.g. send -p is implemented in libzfs and thus not represented in the resume token)
// b) it would force us to either reject resume tokens with unknown flags.
func (a ZFSSendArgsUnvalidated) validateEncryptionFlagsCorrespondToResumeToken(ctx context.Context, valCtx *zfsSendArgsValidationContext) error {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if a.ResumeToken == "" {
return nil // nothing to do
}
debug("decoding resume token %q", a.ResumeToken)
t, err := ParseResumeToken(ctx, a.ResumeToken)
debug("decode resume token result: %#v %T %v", t, err, err)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if err != nil {
return err
}
tokenFS, _, err := t.ToNameSplit()
if err != nil {
return err
}
gen := ZFSSendArgsResumeTokenMismatchGeneric
if a.FS != tokenFS.ToString() {
return ZFSSendArgsResumeTokenMismatchFilesystem.fmt(
"filesystem in resume token field `toname` = %q does not match expected value %q", tokenFS.ToString(), a.FS)
}
// If From is set, it must match.
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if (a.From != nil) != t.HasFromGUID { // existence must be same
if t.HasFromGUID {
return gen.fmt("resume token not expected to be incremental, but `fromguid` = %v", t.FromGUID)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
} else {
return gen.fmt("resume token expected to be incremental, but `fromguid` not present")
}
} else if t.HasFromGUID { // if exists (which is same, we checked above), they must match
if t.FromGUID != a.From.GUID {
return gen.fmt("resume token `fromguid` != expected: %v != %v", t.FromGUID, a.From.GUID)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
} else {
_ = struct{}{} // both empty, ok
}
// To must never be empty
if !t.HasToGUID {
return gen.fmt("resume token does not have `toguid`")
}
if t.ToGUID != a.To.GUID { // a.To != nil because Validate checks for that
return gen.fmt("resume token `toguid` != expected: %v != %v", t.ToGUID, a.To.GUID)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
if a.Encrypted.B {
if !(t.RawOK && t.CompressOK) {
return ZFSSendArgsResumeTokenMismatchEncryptionNotSet.fmt(
"resume token must have `rawok` and `compressok` = true but got %v %v", t.RawOK, t.CompressOK)
}
// fallthrough
} else {
if t.RawOK || t.CompressOK {
return ZFSSendArgsResumeTokenMismatchEncryptionSet.fmt(
"resume token must not have `rawok` or `compressok` set but got %v %v", t.RawOK, t.CompressOK)
}
// fallthrough
}
return nil
}
var zfsSendStderrCaptureMaxSize = envconst.Int("ZREPL_ZFS_SEND_STDERR_MAX_CAPTURE_SIZE", 1<<15)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
var ErrEncryptedSendNotSupported = fmt.Errorf("raw sends which are required for encrypted zfs send are not supported")
// if token != "", then send -t token is used
// otherwise send [-i from] to is used
// (if from is "" a full ZFS send is done)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
//
// Returns ErrEncryptedSendNotSupported if encrypted send is requested but not supported by CLI
func ZFSSend(ctx context.Context, sendArgs ZFSSendArgsValidated) (*SendStream, error) {
args := make([]string, 0)
args = append(args, "send")
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
// pre-validation of sendArgs for plain ErrEncryptedSendNotSupported error
// we tie BackupProperties (send -b) and SendRaw (-w, same as with Encrypted) to this
// since these were released together.
if sendArgs.Encrypted.B || sendArgs.Raw || sendArgs.BackupProperties {
encryptionSupported, err := EncryptionCLISupported(ctx)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if err != nil {
return nil, errors.Wrap(err, "cannot determine CLI native encryption support")
}
if !encryptionSupported {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
return nil, ErrEncryptedSendNotSupported
}
}
sargs, err := sendArgs.buildSendCommandLine()
if err != nil {
return nil, err
}
args = append(args, sargs...)
2017-05-07 12:18:54 +02:00
ctx, cancel := context.WithCancel(ctx)
2017-05-07 12:18:54 +02:00
// setup stdout with an os.Pipe to control pipe buffer size
stdoutReader, stdoutWriter, err := pipeWithCapacityHint(getPipeCapacityHint("ZFS_SEND_PIPE_CAPACITY_HINT"))
if err != nil {
cancel()
return nil, err
}
stderrBuf := circlog.MustNewCircularLog(zfsSendStderrCaptureMaxSize)
cmd := zfscmd.CommandContext(ctx, ZFS_BINARY, args...)
cmd.SetStdio(zfscmd.Stdio{
Stdin: nil,
Stdout: stdoutWriter,
Stderr: stderrBuf,
})
if err := cmd.Start(); err != nil {
cancel()
stdoutWriter.Close()
stdoutReader.Close()
return nil, errors.Wrap(err, "cannot start zfs send command")
}
// close our writing-end of the pipe so that we don't wait for ourselves when reading from the reading end
stdoutWriter.Close()
stream := &SendStream{
2019-03-22 19:41:12 +01:00
cmd: cmd,
kill: cancel,
stdoutReader: stdoutReader,
stderrBuf: stderrBuf,
}
return stream, nil
2017-05-07 12:18:54 +02:00
}
2017-05-07 12:22:57 +02:00
type DrySendType string
const (
DrySendTypeFull DrySendType = "full"
DrySendTypeIncremental DrySendType = "incremental"
)
func DrySendTypeFromString(s string) (DrySendType, error) {
switch s {
2019-03-22 19:41:12 +01:00
case string(DrySendTypeFull):
return DrySendTypeFull, nil
case string(DrySendTypeIncremental):
return DrySendTypeIncremental, nil
default:
return "", fmt.Errorf("unknown dry send type %q", s)
}
}
type DrySendInfo struct {
2019-03-22 19:41:12 +01:00
Type DrySendType
Filesystem string // parsed from To field
From, To string // direct copy from ZFS output
SizeEstimate uint64 // 0 if size estimate is not possible
}
var (
// keep same number of capture groups for unmarshalInfoLine homogeneity
sendDryRunInfoLineRegexFull = regexp.MustCompile(`^(?P<type>full)\t()(?P<to>[^\t]+@[^\t]+)(\t(?P<size>[0-9]+))?$`)
2019-03-22 19:41:12 +01:00
// cannot enforce '[#@]' in incremental source, see test cases
sendDryRunInfoLineRegexIncremental = regexp.MustCompile(`^(?P<type>incremental)\t(?P<from>[^\t]+)\t(?P<to>[^\t]+@[^\t]+)(\t(?P<size>[0-9]+))?$`)
)
// see test cases for example output
func (s *DrySendInfo) unmarshalZFSOutput(output []byte) (err error) {
debug("DrySendInfo.unmarshalZFSOutput: output=%q", output)
lines := strings.Split(string(output), "\n")
for _, l := range lines {
regexMatched, err := s.unmarshalInfoLine(l)
if err != nil {
return fmt.Errorf("line %q: %s", l, err)
}
if !regexMatched {
continue
}
return nil
}
return fmt.Errorf("no match for info line (regex1 %s) (regex2 %s)", sendDryRunInfoLineRegexFull, sendDryRunInfoLineRegexIncremental)
}
// unmarshal info line, looks like this:
// full zroot/test/a@1 5389768
// incremental zroot/test/a@1 zroot/test/a@2 5383936
// => see test cases
func (s *DrySendInfo) unmarshalInfoLine(l string) (regexMatched bool, err error) {
mFull := sendDryRunInfoLineRegexFull.FindStringSubmatch(l)
mInc := sendDryRunInfoLineRegexIncremental.FindStringSubmatch(l)
var matchingExpr *regexp.Regexp
var m []string
if mFull == nil && mInc == nil {
return false, nil
} else if mFull != nil && mInc != nil {
panic(fmt.Sprintf("ambiguous ZFS dry send output: %q", l))
} else if mFull != nil {
matchingExpr, m = sendDryRunInfoLineRegexFull, mFull
} else if mInc != nil {
matchingExpr, m = sendDryRunInfoLineRegexIncremental, mInc
}
fields := make(map[string]string, matchingExpr.NumSubexp())
for i, name := range matchingExpr.SubexpNames() {
if i != 0 {
fields[name] = m[i]
}
}
s.Type, err = DrySendTypeFromString(fields["type"])
if err != nil {
return true, err
}
s.From = fields["from"]
s.To = fields["to"]
toFS, _, _, err := DecomposeVersionString(s.To)
if err != nil {
return true, fmt.Errorf("'to' is not a valid filesystem version: %s", err)
}
s.Filesystem = toFS
if fields["size"] == "" {
// workaround for OpenZFS 0.7 prior to https://github.com/openzfs/zfs/commit/835db58592d7d947e5818eb7281882e2a46073e0#diff-66bd524398bcd2ac70d90925ab6d8073L1245
// see https://github.com/zrepl/zrepl/issues/289
fields["size"] = "0"
}
s.SizeEstimate, err = strconv.ParseUint(fields["size"], 10, 64)
if err != nil {
return true, fmt.Errorf("cannot not parse size: %s", err)
}
return true, nil
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
// to may be "", in which case a full ZFS send is done
// May return BookmarkSizeEstimationNotSupported as err if from is a bookmark.
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
func ZFSSendDry(ctx context.Context, sendArgs ZFSSendArgsValidated) (_ *DrySendInfo, err error) {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if sendArgs.From != nil && strings.Contains(sendArgs.From.RelName, "#") {
/* TODO:
* XXX feature check & support this as well
* ZFS at the time of writing does not support dry-run send because size-estimation
* uses fromSnap's deadlist. However, for a bookmark, that deadlist no longer exists.
* Redacted send & recv will bring this functionality, see
* https://github.com/openzfs/openzfs/pull/484
*/
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
fromAbs, err := absVersion(sendArgs.FS, sendArgs.From)
2019-03-22 19:41:12 +01:00
if err != nil {
return nil, fmt.Errorf("error building abs version for 'from': %s", err)
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
toAbs, err := absVersion(sendArgs.FS, sendArgs.To)
2019-03-22 19:41:12 +01:00
if err != nil {
return nil, fmt.Errorf("error building abs version for 'to': %s", err)
}
return &DrySendInfo{
Type: DrySendTypeIncremental,
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
Filesystem: sendArgs.FS,
2019-03-22 19:41:12 +01:00
From: fromAbs,
To: toAbs,
SizeEstimate: 0}, nil
}
args := make([]string, 0)
args = append(args, "send", "-n", "-v", "-P")
sargs, err := sendArgs.buildSendCommandLine()
if err != nil {
return nil, err
}
args = append(args, sargs...)
cmd := zfscmd.CommandContext(ctx, ZFS_BINARY, args...)
output, err := cmd.CombinedOutput()
if err != nil {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
return nil, &ZFSError{output, err}
}
var si DrySendInfo
if err := si.unmarshalZFSOutput(output); err != nil {
return nil, fmt.Errorf("could not parse zfs send -n output: %s", err)
}
// There is a bug in OpenZFS where it estimates the size incorrectly.
// - zrepl: https://github.com/zrepl/zrepl/issues/463
// - resulting upstream bug: https://github.com/openzfs/zfs/issues/12265
//
// The wrong estimates are easy to detect because they are absurdly large.
// NB: we're doing the workaround for this late so that the test cases are not affected.
sizeEstimateThreshold := envconst.Uint64("ZREPL_ZFS_SEND_SIZE_ESTIMATE_INCORRECT_THRESHOLD", math.MaxInt64)
if sizeEstimateThreshold != 0 && si.SizeEstimate >= sizeEstimateThreshold {
debug("size estimate exceeds threshold %v, working around it: %#v %q", sizeEstimateThreshold, si, args)
si.SizeEstimate = 0
}
return &si, nil
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
type ErrRecvResumeNotSupported struct {
FS string
CheckErr error
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
func (e *ErrRecvResumeNotSupported) Error() string {
var buf strings.Builder
fmt.Fprintf(&buf, "zfs resumable recv into %q: ", e.FS)
if e.CheckErr != nil {
fmt.Fprint(&buf, e.CheckErr.Error())
} else {
fmt.Fprintf(&buf, "not supported by ZFS or pool")
}
return buf.String()
}
type RecvOptions struct {
// Rollback to the oldest snapshot, destroy it, then perform `recv -F`.
// Note that this doesn't change property values, i.e. an existing local property value will be kept.
RollbackAndForceRecv bool
// Set -s flag used for resumable send & recv
SavePartialRecvState bool
InheritProperties []zfsprop.Property
OverrideProperties map[zfsprop.Property]string
}
func (opts RecvOptions) buildRecvFlags() []string {
args := make([]string, 0)
if opts.RollbackAndForceRecv {
args = append(args, "-F")
}
if opts.SavePartialRecvState {
args = append(args, "-s")
}
if opts.InheritProperties != nil {
for _, prop := range opts.InheritProperties {
args = append(args, "-x", string(prop))
}
}
if opts.OverrideProperties != nil {
for prop, value := range opts.OverrideProperties {
args = append(args, "-o", fmt.Sprintf("%s=%s", prop, value))
}
}
return args
}
const RecvStderrBufSiz = 1 << 15
func ZFSRecv(ctx context.Context, fs string, v *ZFSSendArgVersion, stream io.ReadCloser, opts RecvOptions) (err error) {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if err := v.ValidateInMemory(fs); err != nil {
return errors.Wrap(err, "invalid version")
}
if !v.IsSnapshot() {
return errors.New("must receive into a snapshot")
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
fsdp, err := NewDatasetPath(fs)
if err != nil {
return err
}
if opts.RollbackAndForceRecv {
// destroy all snapshots before `recv -F` because `recv -F`
// does not perform a rollback unless `send -R` was used (which we assume hasn't been the case)
snaps, err := ZFSListFilesystemVersions(ctx, fsdp, ListFilesystemVersionsOptions{
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
Types: Snapshots,
})
if _, ok := err.(*DatasetDoesNotExist); ok {
snaps = []FilesystemVersion{}
} else if err != nil {
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
return fmt.Errorf("cannot list versions for rollback for forced receive: %s", err)
}
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
sort.Slice(snaps, func(i, j int) bool {
return snaps[i].CreateTXG < snaps[j].CreateTXG
})
// bookmarks are rolled back automatically
if len(snaps) > 0 {
// use rollback to efficiently destroy all but the earliest snapshot
// then destroy that earliest snapshot
// afterwards, `recv -F` will work
rollbackTarget := snaps[0]
rollbackTargetAbs := rollbackTarget.ToAbsPath(fsdp)
debug("recv: rollback to %q", rollbackTargetAbs)
if err := ZFSRollback(ctx, fsdp, rollbackTarget, "-r"); err != nil {
return fmt.Errorf("cannot rollback %s to %s for forced receive: %s", fsdp.ToString(), rollbackTarget, err)
}
debug("recv: destroy %q", rollbackTargetAbs)
if err := ZFSDestroy(ctx, rollbackTargetAbs); err != nil {
return fmt.Errorf("cannot destroy %s for forced receive: %s", rollbackTargetAbs, err)
}
}
}
2017-05-07 12:22:57 +02:00
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if opts.SavePartialRecvState {
if supported, err := ResumeRecvSupported(ctx, fsdp); err != nil || !supported {
return &ErrRecvResumeNotSupported{FS: fs, CheckErr: err}
}
}
args := []string{"recv"}
args = append(args, opts.buildRecvFlags()...)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
args = append(args, v.FullPath(fs))
2017-05-07 12:22:57 +02:00
ctx, cancelCmd := context.WithCancel(ctx)
defer cancelCmd()
cmd := zfscmd.CommandContext(ctx, ZFS_BINARY, args...)
2017-05-07 12:22:57 +02:00
// TODO report bug upstream
// Setup an unused stdout buffer.
// Otherwise, ZoL v0.6.5.9-1 3.16.0-4-amd64 writes the following error to stderr and exits with code 1
// cannot receive new filesystem stream: invalid backup stream
stdout := bytes.NewBuffer(make([]byte, 0, 1024))
stderr := bytes.NewBuffer(make([]byte, 0, RecvStderrBufSiz))
2017-05-07 12:22:57 +02:00
stdin, stdinWriter, err := pipeWithCapacityHint(getPipeCapacityHint("ZFS_RECV_PIPE_CAPACITY_HINT"))
if err != nil {
return err
}
2019-03-22 19:41:12 +01:00
cmd.SetStdio(zfscmd.Stdio{
Stdin: stdin,
Stdout: stdout,
Stderr: stderr,
})
2017-05-07 12:22:57 +02:00
if err = cmd.Start(); err != nil {
stdinWriter.Close()
stdin.Close()
return err
}
stdin.Close()
defer stdinWriter.Close()
2019-03-22 19:41:12 +01:00
pid := cmd.Process().Pid
debug := func(format string, args ...interface{}) {
debug("recv: pid=%v: %s", pid, fmt.Sprintf(format, args...))
2017-05-07 12:22:57 +02:00
}
debug("started")
copierErrChan := make(chan error)
go func() {
_, err := io.Copy(stdinWriter, stream)
copierErrChan <- err
stdinWriter.Close()
}()
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
waitErrChan := make(chan error)
go func() {
defer close(waitErrChan)
if err = cmd.Wait(); err != nil {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if rtErr := tryRecvErrorWithResumeToken(ctx, stderr.String()); rtErr != nil {
waitErrChan <- rtErr
} else if owErr := tryRecvDestroyOrOverwriteEncryptedErr(stderr.Bytes()); owErr != nil {
waitErrChan <- owErr
} else if readErr := tryRecvCannotReadFromStreamErr(stderr.Bytes()); readErr != nil {
waitErrChan <- readErr
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
} else {
waitErrChan <- &ZFSError{
Stderr: stderr.Bytes(),
WaitErr: err,
}
}
return
2017-05-07 12:22:57 +02:00
}
}()
copierErr := <-copierErrChan
debug("copierErr: %T %s", copierErr, copierErr)
if copierErr != nil {
debug("killing zfs recv command after copierErr")
2019-03-22 19:41:12 +01:00
cancelCmd()
2017-05-07 12:22:57 +02:00
}
waitErr := <-waitErrChan
debug("waitErr: %T %s", waitErr, waitErr)
if copierErr == nil && waitErr == nil {
return nil
} else if _, isReadErr := waitErr.(*RecvCannotReadFromStreamErr); isReadErr {
return copierErr // likely network error reading from stream
} else {
return waitErr // almost always more interesting info. NOTE: do not wrap!
}
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
type RecvFailedWithResumeTokenErr struct {
Msg string
ResumeTokenRaw string
ResumeTokenParsed *ResumeToken
}
var recvErrorResumeTokenRE = regexp.MustCompile(`A resuming stream can be generated on the sending system by running:\s+zfs send -t\s(\S+)`)
func tryRecvErrorWithResumeToken(ctx context.Context, stderr string) *RecvFailedWithResumeTokenErr {
if match := recvErrorResumeTokenRE.FindStringSubmatch(stderr); match != nil {
parsed, err := ParseResumeToken(ctx, match[1])
if err != nil {
return nil
}
return &RecvFailedWithResumeTokenErr{
Msg: stderr,
ResumeTokenRaw: match[1],
ResumeTokenParsed: parsed,
}
}
return nil
}
func (e *RecvFailedWithResumeTokenErr) Error() string {
return fmt.Sprintf("receive failed, resume token available: %s\n%#v", e.ResumeTokenRaw, e.ResumeTokenParsed)
}
type RecvDestroyOrOverwriteEncryptedErr struct {
Msg string
}
func (e *RecvDestroyOrOverwriteEncryptedErr) Error() string {
return e.Msg
}
var recvDestroyOrOverwriteEncryptedErrRe = regexp.MustCompile(`^(cannot receive new filesystem stream: zfs receive -F cannot be used to destroy an encrypted filesystem or overwrite an unencrypted one with an encrypted one)`)
func tryRecvDestroyOrOverwriteEncryptedErr(stderr []byte) *RecvDestroyOrOverwriteEncryptedErr {
debug("tryRecvDestroyOrOverwriteEncryptedErr: %v", stderr)
m := recvDestroyOrOverwriteEncryptedErrRe.FindSubmatch(stderr)
if m == nil {
return nil
}
return &RecvDestroyOrOverwriteEncryptedErr{Msg: string(m[1])}
}
type RecvCannotReadFromStreamErr struct {
Msg string
}
func (e *RecvCannotReadFromStreamErr) Error() string {
return e.Msg
}
var reRecvCannotReadFromStreamErr = regexp.MustCompile(`^(cannot receive: failed to read from stream)$`)
func tryRecvCannotReadFromStreamErr(stderr []byte) *RecvCannotReadFromStreamErr {
m := reRecvCannotReadFromStreamErr.FindSubmatch(stderr)
if m == nil {
return nil
}
return &RecvCannotReadFromStreamErr{Msg: string(m[1])}
}
type ClearResumeTokenError struct {
ZFSOutput []byte
2019-03-22 19:41:12 +01:00
CmdError error
}
func (e ClearResumeTokenError) Error() string {
return fmt.Sprintf("could not clear resume token: %q", string(e.ZFSOutput))
}
// always returns *ClearResumeTokenError
func ZFSRecvClearResumeToken(ctx context.Context, fs string) (err error) {
if err := validateZFSFilesystem(fs); err != nil {
return err
}
cmd := zfscmd.CommandContext(ctx, ZFS_BINARY, "recv", "-A", fs)
o, err := cmd.CombinedOutput()
if err != nil {
if bytes.Contains(o, []byte("does not have any resumable receive state to abort")) {
return nil
}
return &ClearResumeTokenError{o, err}
}
2017-05-07 12:22:57 +02:00
return nil
}
type PropertyValue struct {
Value string
Source PropertySource
}
type ZFSProperties struct {
m map[string]PropertyValue
}
func NewZFSProperties() *ZFSProperties {
return &ZFSProperties{make(map[string]PropertyValue, 4)}
}
func (p *ZFSProperties) Get(key string) string {
return p.m[key].Value
}
func (p *ZFSProperties) GetDetails(key string) PropertyValue {
return p.m[key]
}
func zfsSet(ctx context.Context, path string, props map[string]string) error {
args := make([]string, 0)
args = append(args, "set")
for prop, val := range props {
if strings.Contains(prop, "=") {
return errors.New("prop contains rune '=' which is the delimiter between property name and value")
}
args = append(args, fmt.Sprintf("%s=%s", prop, val))
}
2017-05-07 12:23:12 +02:00
args = append(args, path)
2017-05-07 12:23:12 +02:00
cmd := zfscmd.CommandContext(ctx, ZFS_BINARY, args...)
stdio, err := cmd.CombinedOutput()
if err != nil {
err = &ZFSError{
Stderr: stdio,
2017-05-07 12:23:12 +02:00
WaitErr: err,
}
}
return err
}
func ZFSSet(ctx context.Context, fs *DatasetPath, props map[string]string) error {
return zfsSet(ctx, fs.ToString(), props)
2017-05-07 12:23:12 +02:00
}
2017-07-01 18:21:18 +02:00
func ZFSGet(ctx context.Context, fs *DatasetPath, props []string) (*ZFSProperties, error) {
return zfsGet(ctx, fs.ToString(), props, SourceAny)
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
// The returned error includes requested filesystem and version as quoted strings in its error message
func ZFSGetGUID(ctx context.Context, fs string, version string) (g uint64, err error) {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
defer func(e *error) {
if *e != nil {
*e = fmt.Errorf("zfs get guid fs=%q version=%q: %s", fs, version, *e)
}
}(&err)
if err := validateZFSFilesystem(fs); err != nil {
return 0, err
}
if len(version) == 0 {
return 0, errors.New("version must have non-zero length")
}
if strings.IndexAny(version[0:1], "@#") != 0 {
return 0, errors.New("version does not start with @ or #")
}
path := fmt.Sprintf("%s%s", fs, version)
props, err := zfsGet(ctx, path, []string{"guid"}, SourceAny) // always local
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if err != nil {
return 0, err
}
return strconv.ParseUint(props.Get("guid"), 10, 64)
}
type GetMountpointOutput struct {
Mounted bool
Mountpoint string
}
func ZFSGetMountpoint(ctx context.Context, fs string) (*GetMountpointOutput, error) {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if err := EntityNamecheck(fs, EntityTypeFilesystem); err != nil {
return nil, err
}
props, err := zfsGet(ctx, fs, []string{"mountpoint", "mounted"}, SourceAny)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if err != nil {
return nil, err
}
o := &GetMountpointOutput{}
o.Mounted = props.Get("mounted") == "yes"
o.Mountpoint = props.Get("mountpoint")
if o.Mountpoint == "none" {
o.Mountpoint = ""
}
if o.Mounted && o.Mountpoint == "" {
panic("unexpected zfs get output")
}
return o, nil
}
func ZFSGetRawAnySource(ctx context.Context, path string, props []string) (*ZFSProperties, error) {
return zfsGet(ctx, path, props, SourceAny)
}
var zfsGetDatasetDoesNotExistRegexp = regexp.MustCompile(`^cannot open '([^)]+)': (dataset does not exist|no such pool or dataset)`) // verified in platformtest
type DatasetDoesNotExist struct {
Path string
}
func (d *DatasetDoesNotExist) Error() string { return fmt.Sprintf("dataset %q does not exist", d.Path) }
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
func tryDatasetDoesNotExist(expectPath string, stderr []byte) *DatasetDoesNotExist {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if sm := zfsGetDatasetDoesNotExistRegexp.FindSubmatch(stderr); sm != nil {
if string(sm[1]) == expectPath {
return &DatasetDoesNotExist{expectPath}
}
}
return nil
}
//go:generate enumer -type=PropertySource -trimprefix=Source
type PropertySource uint32
const (
SourceLocal PropertySource = 1 << iota
SourceDefault
SourceInherited
SourceNone
SourceTemporary
SourceReceived
SourceAny PropertySource = ^PropertySource(0)
)
var propertySourceParseLUT = map[string]PropertySource{
"local": SourceLocal,
"default": SourceDefault,
"inherited": SourceInherited,
"-": SourceNone,
"temporary": SourceTemporary,
"received": SourceReceived,
}
func parsePropertySource(s string) (PropertySource, error) {
fields := strings.Fields(s)
if len(fields) > 0 {
v, ok := propertySourceParseLUT[fields[0]]
if ok {
return v, nil
}
// fallthrough
}
return 0, fmt.Errorf("unknown property source %q", s)
}
func (s PropertySource) zfsGetSourceFieldPrefixes() []string {
prefixes := make([]string, 0, 7)
if s&SourceLocal != 0 {
2019-03-22 19:41:12 +01:00
prefixes = append(prefixes, "local")
}
if s&SourceDefault != 0 {
2019-03-22 19:41:12 +01:00
prefixes = append(prefixes, "default")
}
if s&SourceInherited != 0 {
2019-03-22 19:41:12 +01:00
prefixes = append(prefixes, "inherited")
}
if s&SourceNone != 0 {
2019-03-22 19:41:12 +01:00
prefixes = append(prefixes, "-")
}
if s&SourceTemporary != 0 {
2019-03-22 19:41:12 +01:00
prefixes = append(prefixes, "temporary")
}
if s&SourceReceived != 0 {
2019-03-22 19:41:12 +01:00
prefixes = append(prefixes, "received")
}
if s == SourceAny {
2019-03-22 19:41:12 +01:00
prefixes = append(prefixes, "")
}
return prefixes
}
func zfsGetRecursive(ctx context.Context, path string, depth int, dstypes []string, props []string, allowedSources PropertySource) (map[string]*ZFSProperties, error) {
args := []string{"get", "-Hp", "-o", "name,property,value,source"}
if depth != 0 {
args = append(args, "-r")
if depth != -1 {
args = append(args, "-d", fmt.Sprintf("%d", depth))
}
}
if len(dstypes) > 0 {
args = append(args, "-t", strings.Join(dstypes, ","))
}
args = append(args, strings.Join(props, ","), path)
cmd := zfscmd.CommandContext(ctx, ZFS_BINARY, args...)
stdout, err := cmd.Output()
if err != nil {
if exitErr, ok := err.(*exec.ExitError); ok {
if exitErr.Exited() {
// screen-scrape output
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if ddne := tryDatasetDoesNotExist(path, exitErr.Stderr); ddne != nil {
return nil, ddne
}
}
return nil, &ZFSError{
Stderr: exitErr.Stderr,
WaitErr: exitErr,
}
}
return nil, err
}
o := string(stdout)
lines := strings.Split(o, "\n")
propsByFS := make(map[string]*ZFSProperties)
allowedPrefixes := allowedSources.zfsGetSourceFieldPrefixes()
for _, line := range lines[:len(lines)-1] { // last line is an empty line due to how strings.Split works
fields := strings.FieldsFunc(line, func(r rune) bool {
return r == '\t'
})
if len(fields) != 4 {
return nil, fmt.Errorf("zfs get did not return name,property,value,source tuples")
}
for _, p := range allowedPrefixes {
// prefix-match so that SourceAny (= "") works
if strings.HasPrefix(fields[3], p) {
source, err := parsePropertySource(fields[3])
if err != nil {
return nil, errors.Wrap(err, "parse property source")
}
fsProps, ok := propsByFS[fields[0]]
if !ok {
fsProps = &ZFSProperties{
make(map[string]PropertyValue),
}
}
if _, ok := fsProps.m[fields[1]]; ok {
return nil, errors.Errorf("duplicate property %q for dataset %q", fields[1], fields[0])
}
fsProps.m[fields[1]] = PropertyValue{
Value: fields[2],
Source: source,
}
propsByFS[fields[0]] = fsProps
break
}
}
}
// validate we got expected output
for fs, fsProps := range propsByFS {
if len(fsProps.m) != len(props) {
return nil, errors.Errorf("zfs get did not return all requested values for dataset %q\noutput was:\n%s", fs, o)
}
}
return propsByFS, nil
}
func zfsGet(ctx context.Context, path string, props []string, allowedSources PropertySource) (*ZFSProperties, error) {
propMap, err := zfsGetRecursive(ctx, path, 0, nil, props, allowedSources)
if err != nil {
return nil, err
}
if len(propMap) == 0 {
// XXX callers expect to always get a result here
// They will observe props.Get("propname") == ""
// We should change .Get to return a tuple, or an error, or whatever.
return &ZFSProperties{make(map[string]PropertyValue)}, nil
}
if len(propMap) != 1 {
return nil, errors.Errorf("zfs get unexpectedly returned properties for multiple datasets")
}
res, ok := propMap[path]
if !ok {
return nil, errors.Errorf("zfs get returned properties for a different dataset that requested")
}
return res, nil
}
type DestroySnapshotsError struct {
RawLines []string
Filesystem string
Undestroyable []string // snapshot name only (filesystem@ stripped)
Reason []string
}
func (e *DestroySnapshotsError) Error() string {
if len(e.Undestroyable) != len(e.Reason) {
panic(fmt.Sprintf("%v != %v", len(e.Undestroyable), len(e.Reason)))
}
if len(e.Undestroyable) == 0 {
panic(fmt.Sprintf("error must have one undestroyable snapshot, %q", e.Filesystem))
}
if len(e.Undestroyable) == 1 {
return fmt.Sprintf("zfs destroy failed: %s@%s: %s", e.Filesystem, e.Undestroyable[0], e.Reason[0])
}
return strings.Join(e.RawLines, "\n")
}
var destroySnapshotsErrorRegexp = regexp.MustCompile(`^cannot destroy snapshot ([^@]+)@(.+): (.*)$`) // yes, datasets can contain `:`
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
var destroyOneOrMoreSnapshotsNoneExistedErrorRegexp = regexp.MustCompile(`^could not find any snapshots to destroy; check snapshot names.`)
var destroyBookmarkDoesNotExist = regexp.MustCompile(`^bookmark '([^']+)' does not exist`)
func tryParseDestroySnapshotsError(arg string, stderr []byte) *DestroySnapshotsError {
argComps := strings.SplitN(arg, "@", 2)
if len(argComps) != 2 {
return nil
}
filesystem := argComps[0]
lines := bufio.NewScanner(bytes.NewReader(stderr))
undestroyable := []string{}
reason := []string{}
rawLines := []string{}
for lines.Scan() {
line := lines.Text()
rawLines = append(rawLines, line)
m := destroySnapshotsErrorRegexp.FindStringSubmatch(line)
if m == nil {
return nil // unexpected line => be conservative
} else {
if m[1] != filesystem {
return nil // unexpected line => be conservative
}
undestroyable = append(undestroyable, m[2])
reason = append(reason, m[3])
}
}
if len(undestroyable) == 0 {
return nil
}
return &DestroySnapshotsError{
RawLines: rawLines,
Filesystem: filesystem,
Undestroyable: undestroyable,
Reason: reason,
}
}
func ZFSDestroy(ctx context.Context, arg string) (err error) {
2017-07-01 18:21:18 +02:00
var dstype, filesystem string
idx := strings.IndexAny(arg, "@#")
if idx == -1 {
dstype = "filesystem"
filesystem = arg
} else {
switch arg[idx] {
2019-03-22 19:41:12 +01:00
case '@':
dstype = "snapshot"
case '#':
dstype = "bookmark"
}
filesystem = arg[:idx]
}
defer prometheus.NewTimer(prom.ZFSDestroyDuration.WithLabelValues(dstype, filesystem))
cmd := zfscmd.CommandContext(ctx, ZFS_BINARY, "destroy", arg)
stdio, err := cmd.CombinedOutput()
if err != nil {
err = &ZFSError{
Stderr: stdio,
2017-07-01 18:21:18 +02:00
WaitErr: err,
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if destroyOneOrMoreSnapshotsNoneExistedErrorRegexp.Match(stdio) {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
err = &DatasetDoesNotExist{arg}
} else if match := destroyBookmarkDoesNotExist.FindStringSubmatch(string(stdio)); match != nil && match[1] == arg {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
err = &DatasetDoesNotExist{arg}
} else if dsNotExistErr := tryDatasetDoesNotExist(filesystem, stdio); dsNotExistErr != nil {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
err = dsNotExistErr
} else if dserr := tryParseDestroySnapshotsError(arg, stdio); dserr != nil {
err = dserr
}
2017-07-01 18:21:18 +02:00
}
return
}
func ZFSDestroyIdempotent(ctx context.Context, path string) error {
err := ZFSDestroy(ctx, path)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if _, ok := err.(*DatasetDoesNotExist); ok {
return nil
}
return err
}
func ZFSSnapshot(ctx context.Context, fs *DatasetPath, name string, recursive bool) (err error) {
2018-04-05 22:12:25 +02:00
promTimer := prometheus.NewTimer(prom.ZFSSnapshotDuration.WithLabelValues(fs.ToString()))
defer promTimer.ObserveDuration()
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
snapname := fmt.Sprintf("%s@%s", fs.ToString(), name)
if err := EntityNamecheck(snapname, EntityTypeSnapshot); err != nil {
return errors.Wrap(err, "zfs snapshot")
}
cmd := zfscmd.CommandContext(ctx, ZFS_BINARY, "snapshot", snapname)
stdio, err := cmd.CombinedOutput()
if err != nil {
err = &ZFSError{
Stderr: stdio,
WaitErr: err,
}
}
return
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
var zfsBookmarkExistsRegex = regexp.MustCompile("^cannot create bookmark '[^']+': bookmark exists")
type BookmarkExists struct {
zfsMsg string
fs, bookmark string
bookmarkOrigin ZFSSendArgVersion
bookGuid uint64
}
func (e *BookmarkExists) Error() string {
return fmt.Sprintf("bookmark %s (guid=%v) with #%s: bookmark #%s exists but has different guid (%v)",
e.bookmarkOrigin.FullPath(e.fs), e.bookmarkOrigin.GUID, e.bookmark, e.bookmark, e.bookGuid,
)
}
var ErrBookmarkCloningNotSupported = fmt.Errorf("bookmark cloning feature is not yet supported by ZFS")
// idempotently create bookmark of the given version v
//
// if `v` is a bookmark, returns ErrBookmarkCloningNotSupported
// unless a bookmark with the name `bookmark` exists and has the same idenitty (zfs.FilesystemVersionEqualIdentity)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
//
// v must be validated by the caller
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
//
func ZFSBookmark(ctx context.Context, fs string, v FilesystemVersion, bookmark string) (bm FilesystemVersion, err error) {
bm = FilesystemVersion{
Type: Bookmark,
Name: bookmark,
UserRefs: OptionUint64{Valid: false},
// bookmarks have the same createtxg, guid and creation as their origin
CreateTXG: v.CreateTXG,
Guid: v.Guid,
Creation: v.Creation,
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
promTimer := prometheus.NewTimer(prom.ZFSBookmarkDuration.WithLabelValues(fs))
2018-04-05 22:12:25 +02:00
defer promTimer.ObserveDuration()
bookmarkname := fmt.Sprintf("%s#%s", fs, bookmark)
if err := EntityNamecheck(bookmarkname, EntityTypeBookmark); err != nil {
return bm, err
}
if v.IsBookmark() {
existingBm, err := ZFSGetFilesystemVersion(ctx, bookmarkname)
if _, ok := err.(*DatasetDoesNotExist); ok {
return bm, ErrBookmarkCloningNotSupported
} else if err != nil {
return bm, errors.Wrap(err, "bookmark: idempotency check for bookmark cloning")
}
if FilesystemVersionEqualIdentity(bm, existingBm) {
return existingBm, nil
}
return bm, ErrBookmarkCloningNotSupported // TODO This is work in progress: https://github.com/zfsonlinux/zfs/pull/9571
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
snapname := v.FullPath(fs)
if err := EntityNamecheck(snapname, EntityTypeSnapshot); err != nil {
return bm, err
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
cmd := zfscmd.CommandContext(ctx, ZFS_BINARY, "bookmark", snapname, bookmarkname)
stdio, err := cmd.CombinedOutput()
if err != nil {
endpoint: refactor, fix stale holds on initial replication failure, zfs-abstractions subcmd, more efficient ZFS queries The motivation for this recatoring are based on two independent issues: - @JMoVS found that the changes merged as part of #259 slowed his OS X based installation down significantly. Analysis of the zfs command logging introduced in #296 showed that `zfs holds` took most of the execution time, and they pointed out that not all of those `zfs holds` invocations were actually necessary. I.e.: zrepl was inefficient about retrieving information from ZFS. - @InsanePrawn found that failures on initial replication would lead to step holds accumulating on the sending side, i.e. they would never be cleaned up in the HintMostRecentCommonAncestor RPC handler. That was because we only sent that RPC if there was a most recent common ancestor detected during replication planning. @InsanePrawn prototyped an implementation of a `zrepl zfs-abstractions release` command to mitigate the situation. As part of that development work and back-and-forth with @problame, it became evident that the abstractions that #259 built on top of zfs in package endpoint (step holds, replication cursor, last-received-hold), were not well-represented for re-use in the `zrepl zfs-abstractions release` subocommand prototype. This commit refactors package endpoint to address both of these issues: - endpoint abstractions now share an interface `Abstraction` that, among other things, provides a uniform `Destroy()` method. However, that method should not be destroyed directly but instead the package-level `BatchDestroy` function should be used in order to allow for a migration to zfs channel programs in the future. - endpoint now has a query facitilty (`ListAbstractions`) which is used to find on-disk - step holds and bookmarks - replication cursors (v1, v2) - last-received-holds By describing the query in a struct, we can centralized the retrieval of information via the ZFS CLI and only have to be clever once. We are "clever" in the following ways: - When asking for hold-based abstractions, we only run `zfs holds` on snapshot that have `userrefs` > 0 - To support this functionality, add field `UserRefs` to zfs.FilesystemVersion and retrieve it anywhere we retrieve zfs.FilesystemVersion from ZFS. - When asking only for bookmark-based abstractions, we only run `zfs list -t bookmark`, not with snapshots. - Currently unused (except for CLI) per-filesystem concurrent lookup - Option to only include abstractions with CreateTXG in a specified range - refactor `endpoint`'s various ZFS info retrieval methods to use `ListAbstractions` - rename the `zrepl holds list` command to `zrepl zfs-abstractions list` - make `zrepl zfs-abstractions list` consume endpoint.ListAbstractions - Add a `ListStale` method which, given a query template, lists stale holds and bookmarks. - it uses replication cursor has different modes - the new `zrepl zfs-abstractions release-{all,stale}` commands can be used to remove abstractions of package endpoint - Adjust HintMostRecentCommonAncestor RPC for stale-holds cleanup: - send it also if no most recent common ancestor exists between sender and receiver - have the sender clean up its abstractions when it receives the RPC with no most recent common ancestor, using `ListStale` - Due to changed semantics, bump the protocol version. - Adjust HintMostRecentCommonAncestor RPC for performance problems encountered by @JMoVS - by default, per (job,fs)-combination, only consider cleaning step holds in the createtxg range `[last replication cursor,conservatively-estimated-receive-side-version)` - this behavior ensures resumability at cost proportional to the time that replication was donw - however, as explained in a comment, we might leak holds if the zrepl daemon stops running - that trade-off is acceptable because in the presumably rare this might happen the user has two tools at their hand: - Tool 1: run `zrepl zfs-abstractions release-stale` - Tool 2: use env var `ZREPL_ENDPOINT_SENDER_HINT_MOST_RECENT_STEP_HOLD_CLEANUP_MODE` to adjust the lower bound of the createtxg range (search for it in the code). The env var can also be used to disable hold-cleanup on the send-side entirely. supersedes closes #293 supersedes closes #282 fixes #280 fixes #278 Additionaly, we fixed a couple of bugs: - zfs: fix half-nil error reporting of dataset-does-not-exist for ZFSListChan and ZFSBookmark - endpoint: Sender's `HintMostRecentCommonAncestor` handler would not check whether access to the specified filesystem was allowed.
2020-03-26 23:43:17 +01:00
if ddne := tryDatasetDoesNotExist(snapname, stdio); ddne != nil {
return bm, ddne
} else if zfsBookmarkExistsRegex.Match(stdio) {
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
// check if this was idempotent
bookGuid, err := ZFSGetGUID(ctx, fs, "#"+bookmark)
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
if err != nil {
return bm, errors.Wrap(err, "bookmark: idempotency check for bookmark creation") // guid error expressive enough
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
if v.Guid == bookGuid {
debug("bookmark: %q %q was idempotent: {snap,book}guid %d == %d", snapname, bookmarkname, v.Guid, bookGuid)
return bm, nil
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
return bm, &BookmarkExists{
fs: fs, bookmarkOrigin: v.ToSendArgVersion(), bookmark: bookmark,
zfsMsg: string(stdio),
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
bookGuid: bookGuid,
}
} else {
return bm, &ZFSError{
Stderr: stdio,
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
WaitErr: err,
}
}
new features: {resumable,encrypted,hold-protected} send-recv, last-received-hold - **Resumable Send & Recv Support** No knobs required, automatically used where supported. - **Hold-Protected Send & Recv** Automatic ZFS holds to ensure that we can always resume a replication step. - **Encrypted Send & Recv Support** for OpenZFS native encryption. Configurable at the job level, i.e., for all filesystems a job is responsible for. - **Receive-side hold on last received dataset** The counterpart to the replication cursor bookmark on the send-side. Ensures that incremental replication will always be possible between a sender and receiver. Design Doc ---------- `replication/design.md` doc describes how we use ZFS holds and bookmarks to ensure that a single replication step is always resumable. The replication algorithm described in the design doc introduces the notion of job IDs (please read the details on this design doc). We reuse the job names for job IDs and use `JobID` type to ensure that a job name can be embedded into hold tags, bookmark names, etc. This might BREAK CONFIG on upgrade. Protocol Version Bump --------------------- This commit makes backwards-incompatible changes to the replication/pdu protobufs. Thus, bump the version number used in the protocol handshake. Replication Cursor Format Change -------------------------------- The new replication cursor bookmark format is: `#zrepl_CURSOR_G_${this.GUID}_J_${jobid}` Including the GUID enables transaction-safe moving-forward of the cursor. Including the job id enables that multiple sending jobs can send the same filesystem without interfering. The `zrepl migrate replication-cursor:v1-v2` subcommand can be used to safely destroy old-format cursors once zrepl has created new-format cursors. Changes in This Commit ---------------------- - package zfs - infrastructure for holds - infrastructure for resume token decoding - implement a variant of OpenZFS's `entity_namecheck` and use it for validation in new code - ZFSSendArgs to specify a ZFS send operation - validation code protects against malicious resume tokens by checking that the token encodes the same send parameters that the send-side would use if no resume token were available (i.e. same filesystem, `fromguid`, `toguid`) - RecvOptions support for `recv -s` flag - convert a bunch of ZFS operations to be idempotent - achieved through more differentiated error message scraping / additional pre-/post-checks - package replication/pdu - add field for encryption to send request messages - add fields for resume handling to send & recv request messages - receive requests now contain `FilesystemVersion To` in addition to the filesystem into which the stream should be `recv`d into - can use `zfs recv $root_fs/$client_id/path/to/dataset@${To.Name}`, which enables additional validation after recv (i.e. whether `To.Guid` matched what we received in the stream) - used to set `last-received-hold` - package replication/logic - introduce `PlannerPolicy` struct, currently only used to configure whether encrypted sends should be requested from the sender - integrate encryption and resume token support into `Step` struct - package endpoint - move the concepts that endpoint builds on top of ZFS to a single file `endpoint/endpoint_zfs.go` - step-holds + step-bookmarks - last-received-hold - new replication cursor + old replication cursor compat code - adjust `endpoint/endpoint.go` handlers for - encryption - resumability - new replication cursor - last-received-hold - client subcommand `zrepl holds list`: list all holds and hold-like bookmarks that zrepl thinks belong to it - client subcommand `zrepl migrate replication-cursor:v1-v2`
2019-09-11 17:19:17 +02:00
}
return bm, nil
}
func ZFSRollback(ctx context.Context, fs *DatasetPath, snapshot FilesystemVersion, rollbackArgs ...string) (err error) {
snapabs := snapshot.ToAbsPath(fs)
if snapshot.Type != Snapshot {
return fmt.Errorf("can only rollback to snapshots, got %s", snapabs)
}
args := []string{"rollback"}
args = append(args, rollbackArgs...)
args = append(args, snapabs)
cmd := zfscmd.CommandContext(ctx, ZFS_BINARY, args...)
stdio, err := cmd.CombinedOutput()
if err != nil {
err = &ZFSError{
Stderr: stdio,
WaitErr: err,
}
}
return err
}