extern/zrepl
1
0
Fork 0
mirror of https://github.com/zrepl/zrepl.git synced 2024-11-26 02:14:44 +01:00
Code Issues Actions Packages Projects Releases Wiki Activity
dc3378e890
zrepl/cmd/sampleconf/zrepl.yml

112 lines
2.4 KiB
YAML
Raw Normal View History

config: rename 'pools' section to 'remotes'
2017-07-08 12:07:38 +02:00
remotes:
Simplify CLI by requiring explicit job names. Job names are derived from job type + user-defined name in config file CLI now has subcommands corresponding 1:1 to the config file sections: push,pull,autosnap,prune A subcommand always expects a job name, thus executes exactly one job. Dict-style syntax also used for PullACL and Sink sections. jobrun package is currently only used for autosnap, all others need to be invoked repeatedly via external tool. Plan is to re-integrate jobrun in an explicit daemon-mode (subcommand).
2017-07-06 15:36:53 +02:00
offsite_backups:
Change transport format in zrepl config & parse it.
2017-04-29 20:10:09 +02:00
transport:
ssh:
config: PushACLs, sinks are also just ClientMappings & LOCAL_TRANSPORT_IDENTITY
2017-04-30 17:55:11 +02:00
host: 192.168.122.6
Change transport format in zrepl config & parse it.
2017-04-29 20:10:09 +02:00
user: root
port: 22
config: PushACLs, sinks are also just ClientMappings & LOCAL_TRANSPORT_IDENTITY
2017-04-30 17:55:11 +02:00
identity_file: /etc/zrepl/identities/offsite_backups
Implement config parser.
2017-04-26 18:36:01 +02:00
cmd: add 'test' subcommand configbreak
2017-09-02 12:24:17 +02:00
push:
Simplify CLI by requiring explicit job names. Job names are derived from job type + user-defined name in config file CLI now has subcommands corresponding 1:1 to the config file sections: push,pull,autosnap,prune A subcommand always expects a job name, thus executes exactly one job. Dict-style syntax also used for PullACL and Sink sections. jobrun package is currently only used for autosnap, all others need to be invoked repeatedly via external tool. Plan is to re-integrate jobrun in an explicit daemon-mode (subcommand).
2017-07-06 15:36:53 +02:00
offsite:
to: offsite_backups
Implement push support. Pushing is achieved by inverting the roles on the established connection, i.e. the client tells the server what data it should pull from the client (PullMeRequest). Role inversion is achieved by moving the server loop to the serverLoop function of ByteStreamRPC, which can be called from both the Listen() function (server-side) and the PullMeRequest() client-side function. A donwside of this PullMe approach is that the replication policies become part of the rpc, because the puller must follow the policy.
2017-05-16 16:57:24 +02:00
filter: {
rework filters & mappings config defines a single datastructure that can act both as a Map and as a Filter (DatasetMapFilter) Cleanup wildcard syntax along the way (also changes semantics).
2017-08-05 21:15:37 +02:00
# like in pull_acls
"tank/var/db<": ok,
"tank/usr/home<": ok,
Implement push support. Pushing is achieved by inverting the roles on the established connection, i.e. the client tells the server what data it should pull from the client (PullMeRequest). Role inversion is achieved by moving the server loop to the serverLoop function of ByteStreamRPC, which can be called from both the Listen() function (server-side) and the PullMeRequest() client-side function. A donwside of this PullMe approach is that the replication policies become part of the rpc, because the puller must follow the policy.
2017-05-16 16:57:24 +02:00
}
Implement config parser.
2017-04-26 18:36:01 +02:00
cmd: add 'test' subcommand configbreak
2017-09-02 12:24:17 +02:00
pull:
Simplify CLI by requiring explicit job names. Job names are derived from job type + user-defined name in config file CLI now has subcommands corresponding 1:1 to the config file sections: push,pull,autosnap,prune A subcommand always expects a job name, thus executes exactly one job. Dict-style syntax also used for PullACL and Sink sections. jobrun package is currently only used for autosnap, all others need to be invoked repeatedly via external tool. Plan is to re-integrate jobrun in an explicit daemon-mode (subcommand).
2017-07-06 15:36:53 +02:00
offsite:
from: offsite_backups
Implement config parser.
2017-04-26 18:36:01 +02:00
mapping: {
# like in sinks
}
config: PushACLs, sinks are also just ClientMappings & LOCAL_TRANSPORT_IDENTITY
2017-04-30 17:55:11 +02:00
# local replication, only allowed in pull mode
# the from name 'local' is reserved for this purpose
Simplify CLI by requiring explicit job names. Job names are derived from job type + user-defined name in config file CLI now has subcommands corresponding 1:1 to the config file sections: push,pull,autosnap,prune A subcommand always expects a job name, thus executes exactly one job. Dict-style syntax also used for PullACL and Sink sections. jobrun package is currently only used for autosnap, all others need to be invoked repeatedly via external tool. Plan is to re-integrate jobrun in an explicit daemon-mode (subcommand).
2017-07-06 15:36:53 +02:00
homemirror:
from: local
cmd/jobrun: repeat strategies as part of jobrun
2017-06-09 21:00:28 +02:00
repeat:
interval: 15m
Change transport format in zrepl config & parse it.
2017-04-29 20:10:09 +02:00
mapping: {
"tank/usr/home":"mirrorpool/foo/bar"
}
cmd: add 'test' subcommand configbreak
2017-09-02 12:24:17 +02:00
sink:
Implement config parser.
2017-04-26 18:36:01 +02:00
Simplify CLI by requiring explicit job names. Job names are derived from job type + user-defined name in config file CLI now has subcommands corresponding 1:1 to the config file sections: push,pull,autosnap,prune A subcommand always expects a job name, thus executes exactly one job. Dict-style syntax also used for PullACL and Sink sections. jobrun package is currently only used for autosnap, all others need to be invoked repeatedly via external tool. Plan is to re-integrate jobrun in an explicit daemon-mode (subcommand).
2017-07-06 15:36:53 +02:00
db1:
Implement config parser.
2017-04-26 18:36:01 +02:00
mapping: {
rework filters & mappings config defines a single datastructure that can act both as a Map and as a Filter (DatasetMapFilter) Cleanup wildcard syntax along the way (also changes semantics).
2017-08-05 21:15:37 +02:00
# direct mapping
Implement config parser.
2017-04-26 18:36:01 +02:00
"ssdpool/var/db/postgresql9.6":"zroot/backups/db1/pg_data"
}
rework filters & mappings config defines a single datastructure that can act both as a Map and as a Filter (DatasetMapFilter) Cleanup wildcard syntax along the way (also changes semantics).
2017-08-05 21:15:37 +02:00
mirror1:
Implement config parser.
2017-04-26 18:36:01 +02:00
mapping: {
rework filters & mappings config defines a single datastructure that can act both as a Map and as a Filter (DatasetMapFilter) Cleanup wildcard syntax along the way (also changes semantics).
2017-08-05 21:15:37 +02:00
# "<" subtree wildcard matches the dataset left of < and all its children
"tank/foo/bar<":"zroot/backups/mirror1"
Implement config parser.
2017-04-26 18:36:01 +02:00
}
rework filters & mappings config defines a single datastructure that can act both as a Map and as a Filter (DatasetMapFilter) Cleanup wildcard syntax along the way (also changes semantics).
2017-08-05 21:15:37 +02:00
mirror2:
# more specific path patterns win over less specific ones
# direct mappings win over subtree wildcards
# detailed rule precedence: check unit tests & docs for exact behavior
# TODO subcommand to test a mapping & filter
Implement config parser.
2017-04-26 18:36:01 +02:00
mapping: {
rework filters & mappings config defines a single datastructure that can act both as a Map and as a Filter (DatasetMapFilter) Cleanup wildcard syntax along the way (also changes semantics).
2017-08-05 21:15:37 +02:00
"tank<": "zroot/backups/mirror1/tank1",
"tank/cdn/root<": "storage/cdn/root",
"tank/legacydb": "legacypool/backups/legacydb",
Implement config parser.
2017-04-26 18:36:01 +02:00
}
config: PushACLs, sinks are also just ClientMappings & LOCAL_TRANSPORT_IDENTITY
2017-04-30 17:55:11 +02:00
cmd: add 'test' subcommand configbreak
2017-09-02 12:24:17 +02:00
pull_acl:
Simplify CLI by requiring explicit job names. Job names are derived from job type + user-defined name in config file CLI now has subcommands corresponding 1:1 to the config file sections: push,pull,autosnap,prune A subcommand always expects a job name, thus executes exactly one job. Dict-style syntax also used for PullACL and Sink sections. jobrun package is currently only used for autosnap, all others need to be invoked repeatedly via external tool. Plan is to re-integrate jobrun in an explicit daemon-mode (subcommand).
2017-07-06 15:36:53 +02:00
office_backup:
rework filters & mappings config defines a single datastructure that can act both as a Map and as a Filter (DatasetMapFilter) Cleanup wildcard syntax along the way (also changes semantics).
2017-08-05 21:15:37 +02:00
filter: {
# valid filter results (right hand side): ok, omit
# default is to omit
# rule precedence is same as for mappings
"tank<": omit,
"tank/usr/home": ok,
config: PushACLs, sinks are also just ClientMappings & LOCAL_TRANSPORT_IDENTITY
2017-04-30 17:55:11 +02:00
}
implement snapshot pruning feature
2017-06-22 19:04:48 +02:00
prune:
Simplify CLI by requiring explicit job names. Job names are derived from job type + user-defined name in config file CLI now has subcommands corresponding 1:1 to the config file sections: push,pull,autosnap,prune A subcommand always expects a job name, thus executes exactly one job. Dict-style syntax also used for PullACL and Sink sections. jobrun package is currently only used for autosnap, all others need to be invoked repeatedly via external tool. Plan is to re-integrate jobrun in an explicit daemon-mode (subcommand).
2017-07-06 15:36:53 +02:00
clean_backups:
implement snapshot pruning feature
2017-06-22 19:04:48 +02:00
policy: grid
Remove months as a possible time interval unit as it is too volatile. Thanks to @erdgeist for pointing that out. refs #2
2017-07-09 00:38:16 +02:00
grid: 6x10m | 24x1h | 7x1d | 5 x 1w | 4 x 5w
implement snapshot pruning feature
2017-06-22 19:04:48 +02:00
dataset_filter: {
rework filters & mappings config defines a single datastructure that can act both as a Map and as a Filter (DatasetMapFilter) Cleanup wildcard syntax along the way (also changes semantics).
2017-08-05 21:15:37 +02:00
"tank/backups/legacyscript<": omit,
"tank/backups<": ok,
implement snapshot pruning feature
2017-06-22 19:04:48 +02:00
}
snapshot_filter: {
prefix: zrepl_
}
implement automatic snapshotting feature
2017-07-01 20:28:46 +02:00
Simplify CLI by requiring explicit job names. Job names are derived from job type + user-defined name in config file CLI now has subcommands corresponding 1:1 to the config file sections: push,pull,autosnap,prune A subcommand always expects a job name, thus executes exactly one job. Dict-style syntax also used for PullACL and Sink sections. jobrun package is currently only used for autosnap, all others need to be invoked repeatedly via external tool. Plan is to re-integrate jobrun in an explicit daemon-mode (subcommand).
2017-07-06 15:36:53 +02:00
hfbak_prune: # cleans up after hfbak autosnap job
implement automatic snapshotting feature
2017-07-01 20:28:46 +02:00
policy: grid
Remove months as a possible time interval unit as it is too volatile. Thanks to @erdgeist for pointing that out. refs #2
2017-07-09 00:38:16 +02:00
grid: 1x1m(keep=all)
implement automatic snapshotting feature
2017-07-01 20:28:46 +02:00
dataset_filter: {
rework filters & mappings config defines a single datastructure that can act both as a Map and as a Filter (DatasetMapFilter) Cleanup wildcard syntax along the way (also changes semantics).
2017-08-05 21:15:37 +02:00
"pool1<": ok
implement automatic snapshotting feature
2017-07-01 20:28:46 +02:00
}
snapshot_filter: {
prefix: zrepl_hfbak_
}
cmd: add repeat config option to Prune
2017-09-01 16:45:43 +02:00
repeat:
interval: 10s
implement automatic snapshotting feature
2017-07-01 20:28:46 +02:00
autosnap:
Simplify CLI by requiring explicit job names. Job names are derived from job type + user-defined name in config file CLI now has subcommands corresponding 1:1 to the config file sections: push,pull,autosnap,prune A subcommand always expects a job name, thus executes exactly one job. Dict-style syntax also used for PullACL and Sink sections. jobrun package is currently only used for autosnap, all others need to be invoked repeatedly via external tool. Plan is to re-integrate jobrun in an explicit daemon-mode (subcommand).
2017-07-06 15:36:53 +02:00
hfbak:
implement automatic snapshotting feature
2017-07-01 20:28:46 +02:00
prefix: zrepl_hfbak_
interval: 1s
dataset_filter: {
rework filters & mappings config defines a single datastructure that can act both as a Map and as a Filter (DatasetMapFilter) Cleanup wildcard syntax along the way (also changes semantics).
2017-08-05 21:15:37 +02:00
"pool1<": ok
implement automatic snapshotting feature
2017-07-01 20:28:46 +02:00
}
# prune: hfbak_prune
# future versions may inline the retention policy here, but for now,
# pruning has to be triggered manually (it's safe to run autosnap + prune in parallel)
Reference in New Issue Copy Permalink