zrepl/platformtest/platformtest_ops.go

327 lines
7.6 KiB
Go
Raw Normal View History

package platformtest
import (
"bufio"
"bytes"
"context"
"fmt"
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
"io/ioutil"
"os"
"os/exec"
"strings"
"unicode"
)
type Execer interface {
RunExpectSuccessNoOutput(ctx context.Context, cmd string, args ...string) error
RunExpectFailureNoOutput(ctx context.Context, cmd string, args ...string) error
}
type Stmt interface {
Run(context context.Context, e Execer) error
}
type Op string
const (
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
Comment Op = "#"
AssertExists Op = "!E"
AssertNotExists Op = "!N"
Add Op = "+"
Del Op = "-"
RunCmd Op = "R"
DestroyRoot Op = "DESTROYROOT"
CreateRoot Op = "CREATEROOT"
)
type DestroyRootOp struct {
Path string
}
func (o *DestroyRootOp) Run(ctx context.Context, e Execer) error {
// early-exit if it doesn't exist
if err := e.RunExpectSuccessNoOutput(ctx, "zfs", "get", "-H", "name", o.Path); err != nil {
GetLog(ctx).WithField("root_ds", o.Path).Info("assume root ds doesn't exist")
return nil
}
return e.RunExpectSuccessNoOutput(ctx, "zfs", "destroy", "-r", o.Path)
}
type FSOp struct {
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
Op Op
Path string
Encrypted bool // only for Op=Add
}
func (o *FSOp) Run(ctx context.Context, e Execer) error {
switch o.Op {
case AssertExists:
return e.RunExpectSuccessNoOutput(ctx, "zfs", "get", "-H", "name", o.Path)
case AssertNotExists:
return e.RunExpectFailureNoOutput(ctx, "zfs", "get", "-H", "name", o.Path)
case Add:
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
opts := []string{"create"}
if o.Encrypted {
const passphraseFilePath = "/tmp/zreplplatformtest.encryption.passphrase"
const passphrase = "foobar2342"
err := ioutil.WriteFile(passphraseFilePath, []byte(passphrase), 0600)
if err != nil {
panic(err)
}
opts = append(opts,
"-o", "encryption=on",
"-o", "keylocation=file:///"+passphraseFilePath,
"-o", "keyformat=passphrase",
)
}
opts = append(opts, o.Path)
return e.RunExpectSuccessNoOutput(ctx, "zfs", opts...)
case Del:
return e.RunExpectSuccessNoOutput(ctx, "zfs", "destroy", o.Path)
default:
panic(o.Op)
}
}
type SnapOp struct {
Op Op
Path string
}
func (o *SnapOp) Run(ctx context.Context, e Execer) error {
switch o.Op {
case AssertExists:
return e.RunExpectSuccessNoOutput(ctx, "zfs", "get", "-H", "name", o.Path)
case AssertNotExists:
return e.RunExpectFailureNoOutput(ctx, "zfs", "get", "-H", "name", o.Path)
case Add:
return e.RunExpectSuccessNoOutput(ctx, "zfs", "snapshot", o.Path)
case Del:
return e.RunExpectSuccessNoOutput(ctx, "zfs", "destroy", o.Path)
default:
panic(o.Op)
}
}
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 BookmarkOp struct {
Op Op
Existing string
Bookmark string
}
func (o *BookmarkOp) Run(ctx context.Context, e Execer) error {
switch o.Op {
case Add:
return e.RunExpectSuccessNoOutput(ctx, "zfs", "bookmark", o.Existing, o.Bookmark)
case Del:
if o.Existing != "" {
panic("existing must be empty for destroy, got " + o.Existing)
}
return e.RunExpectSuccessNoOutput(ctx, "zfs", "destroy", o.Bookmark)
default:
panic(o.Op)
}
}
type RunOp struct {
RootDS string
Script string
}
func (o *RunOp) Run(ctx context.Context, e Execer) error {
cmd := exec.CommandContext(ctx, "/usr/bin/env", "bash", "-c", o.Script)
cmd.Env = os.Environ()
cmd.Env = append(cmd.Env, fmt.Sprintf("ROOTDS=%s", o.RootDS))
log := GetLog(ctx).WithField("script", o.Script)
log.Info("start script")
defer log.Info("script done")
output, err := cmd.CombinedOutput()
if _, ok := err.(*exec.ExitError); err != nil && !ok {
panic(err)
}
log.Printf("script output:\n%s", output)
return err
}
type LineError struct {
Line string
What string
}
func (e LineError) Error() string {
return fmt.Sprintf("%q: %s", e.Line, e.What)
}
type RunKind int
const (
PanicErr RunKind = 1 << iota
RunAll
)
func Run(ctx context.Context, rk RunKind, rootds string, stmtsStr string) {
stmt, err := parseSequence(rootds, stmtsStr)
if err != nil {
panic(err)
}
execer := NewEx(GetLog(ctx))
for _, s := range stmt {
err := s.Run(ctx, execer)
if err == nil {
continue
}
if rk == PanicErr {
panic(err)
} else if rk == RunAll {
continue
} else {
panic(rk)
}
}
}
func isNoSpace(r rune) bool {
return !unicode.IsSpace(r)
}
func splitQuotedWords(data []byte, atEOF bool) (advance int, token []byte, err error) {
begin := bytes.IndexFunc(data, isNoSpace)
if begin == -1 {
return len(data), nil, nil
}
if data[begin] == '"' {
end := begin + 1
for end < len(data) {
endCandidate := bytes.Index(data[end:], []byte(`"`))
if endCandidate == -1 {
return 0, nil, nil
}
end += endCandidate
if data[end-1] != '\\' {
// unescaped quote, end of this string
// remove backslash-escapes
withBackslash := data[begin+1 : end]
withoutBackslash := bytes.Replace(withBackslash, []byte("\\\""), []byte("\""), -1)
return end + 1, withoutBackslash, nil
} else {
// continue to next quote
end += 1
}
}
} else {
endOffset := bytes.IndexFunc(data[begin:], unicode.IsSpace)
var end int
if endOffset == -1 {
if !atEOF {
return 0, nil, nil
} else {
end = len(data)
}
} else {
end = begin + endOffset
}
return end, data[begin:end], nil
}
return 0, nil, fmt.Errorf("unexpected")
}
func parseSequence(rootds, stmtsStr string) (stmts []Stmt, err error) {
scan := bufio.NewScanner(strings.NewReader(stmtsStr))
nextLine:
for scan.Scan() {
if len(bytes.TrimSpace(scan.Bytes())) == 0 {
continue
}
comps := bufio.NewScanner(bytes.NewReader(scan.Bytes()))
comps.Split(splitQuotedWords)
expectMoreTokens := func() error {
if !comps.Scan() {
return &LineError{scan.Text(), "unexpected EOL"}
}
return nil
}
// Op
if err := expectMoreTokens(); err != nil {
return nil, err
}
var op Op
switch comps.Text() {
case string(RunCmd):
script := strings.TrimPrefix(strings.TrimSpace(scan.Text()), string(RunCmd))
stmts = append(stmts, &RunOp{RootDS: rootds, Script: script})
continue nextLine
case string(DestroyRoot):
if comps.Scan() {
return nil, &LineError{scan.Text(), "unexpected tokens at EOL"}
}
stmts = append(stmts, &DestroyRootOp{rootds})
continue nextLine
case string(CreateRoot):
if comps.Scan() {
return nil, &LineError{scan.Text(), "unexpected tokens at EOL"}
}
stmts = append(stmts, &FSOp{Op: Add, Path: rootds})
continue nextLine
case string(Add):
op = Add
case string(Del):
op = Del
case string(AssertExists):
op = AssertExists
case string(AssertNotExists):
op = AssertNotExists
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
case string(Comment):
op = Comment
continue
default:
return nil, &LineError{scan.Text(), fmt.Sprintf("invalid op %q", comps.Text())}
}
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
// FS / SNAP / BOOKMARK
if err := expectMoreTokens(); err != nil {
return nil, err
}
if strings.ContainsAny(comps.Text(), "@") {
stmts = append(stmts, &SnapOp{Op: op, Path: fmt.Sprintf("%s/%s", rootds, comps.Text())})
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
} else if strings.ContainsAny(comps.Text(), "#") {
bookmark := fmt.Sprintf("%s/%s", rootds, comps.Text())
if err := expectMoreTokens(); err != nil {
return nil, err
}
existing := fmt.Sprintf("%s/%s", rootds, comps.Text())
stmts = append(stmts, &BookmarkOp{Op: op, Existing: existing, Bookmark: bookmark})
} else {
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
// FS
fs := comps.Text()
var encrypted bool = false
if op == Add {
if comps.Scan() {
t := comps.Text()
switch t {
case "encrypted":
encrypted = true
default:
panic(fmt.Sprintf("unexpected token %q", t))
}
}
}
stmts = append(stmts, &FSOp{
Op: op,
Path: fmt.Sprintf("%s/%s", rootds, fs),
Encrypted: encrypted,
})
}
if comps.Scan() {
return nil, &LineError{scan.Text(), "unexpected tokens at EOL"}
}
}
return stmts, nil
}