Skip to content

Latest commit

 

History

History
163 lines (116 loc) · 30.4 KB

cluster_spec.md

File metadata and controls

163 lines (116 loc) · 30.4 KB

Cluster Specification

Stolon has a declarative model where you specify the desired cluster state, this is called cluster specification and it's saved inside the cluster data in the store.

A cluster needs to be initialized providing a cluster specification. This can be achieved using stolonctl init.

A cluster specification is updatable using stolonctl update.

Some options in a running cluster specification can be changed to update the desired state. Sometimes a cluster state can be updated only in some directions, this means that some options cannot be updated on a running cluster but will require a new cluster initialization.

Cluster Specification Format.

Name Description Required Type Default
sleepInterval interval to wait before next check (for keepers and sentinels). no string (duration) 5s
requestTimeout time after which any request to external resources (store, postgres queries etc...) will fail. no string (duration) 10s
failInterval interval after the first fail to declare a keeper as not healthy. no string (duration) 20s
proxyCheckInterval interval to wait before next proxy check. no string (duration) 5s
proxyTimeout interval where a proxy check must successfully complete or the proxy will close all connections to the master. no string (duration) 15s
deadKeeperRemovalInterval interval after which a dead keeper will be removed from the cluster data no string (duration) 48h
maxStandbys max number of standbys. This needs to be greater enough to cover both standby managed by stolon and additional standbys configured by the user. Its value affect different postgres parameters like max_replication_slots and max_wal_senders. Setting this to a number lower than the sum of stolon managed standbys and user managed standbys will have unpredicatable effects due to problems creating replication slots or replication problems due to exhausted wal senders. no uint16 20
maxStandbysPerSender max number of standbys for every sender. A sender can be a master or another standby (with cascading replication). no uint16 3
maxStandbyLag maximum lag (from the last reported master state, in bytes) that an asynchronous standby can have to be elected in place of a failed master. no uint32 1MiB
synchronousReplication use synchronous replication between the master and its standbys no bool false
minSynchronousStandbys minimum number of required synchronous standbys when synchronous replication is enabled (only set this to a value != 1 when using PostgreSQL >= 9.6). Notice that you are also allowed to set it to 0, in which case the primary will keep accepting commits even when there is no healthy sync standby. Setting minSynchronousStandbys to 0 is not generally advised - please see syncrepl.md no uint16 1
maxSynchronousStandbys maximum number of required synchronous standbys when synchronous replication is enabled (only set this to a value > 1 when using PostgreSQL >= 9.6) no uint16 1
additionalWalSenders number of additional wal_senders in addition to the ones internally defined by stolon, useful to provide enough wal senders for external standbys (changing this value requires an instance restart) no uint16 5
additionalMasterReplicationSlots a list of additional physical replication slots to be created on the master postgres instance. They will be prefixed with stolon_ (like internal replication slots used for standby replication) to make them "namespaced" from other replication slots. Replication slots starting with stolon_ and not defined here (and not used for standby replication) will be dropped from the master instance. no []string null
usePgrewind try to use pg_rewind for faster instance resyncronization. no bool false
initMode The cluster initialization mode. Can be new or existing. new means that a new db cluster will be created on a random keeper and the other keepers will sync with it. existing means that a keeper (that needs to have an already created db cluster) will be choosed as the initial master and the other keepers will sync with it. In this case the existingConfig object needs to be populated. yes string
existingConfig configuration for initMode of type "existing" if initMode is "existing" ExistingConfig
mergePgParameters merge pgParameters of the initialized db cluster, useful the retain initdb generated parameters when InitMode is new, retain current parameters when initMode is existing or pitr. no bool true
role cluster role (master or standby) no bool master
defaultSUReplAccessMode mode for the default hba rules used for replication by standby keepers (the su and repl auth methods will be the one provided in the keeper command line options). Values can be all or strict. all allow access from all ips, strict restrict master access to standby servers ips. no string all
newConfig configuration for initMode of type "new" if initMode is "new" NewConfig
pitrConfig configuration for initMode of type "pitr" if initMode is "pitr" PITRConfig
standbyConfig standby config when the cluster is a standby cluster if role is "standby" StandbyConfig
pgParameters a map containing the postgres server parameters and their values. The parameters value don't have to be quoted and single quotes don't have to be doubled since this is already done by the keeper when writing the postgresql.conf file no map[string]string
pgHBA a list containing additional pg_hba.conf entries. They will be added to the pg_hba.conf generated by stolon. NOTE: these lines aren't validated so if some of them are wrong postgres will refuse to start or, on reload, will log a warning and ignore the updated pg_hba.conf file no []string null. Will use the default behaviour of accepting connections from all hosts for all dbs and users with md5 password authentication
automaticPgRestart restart postgres automatically after changing the pgParameters that requires restart. Refer pending_restart in pg_settings no bool false
dbWaitReadyTimeout Time to wait for the database to become ready after starting. Increase this value if your Postgres takes longer to boot, e.g. because it has to recover a lot of WAL. no string (duration) 60s
syncTimeout Time to wait for a database recovery (including the replay of WAL files in case of Point-In-Time-Recovery) no string (duration) 0 (no timeout, waits until recovery has finished)

ExistingConfig

Name Description Required Type Default
keeperUID the keeperUID to use as the initial master db cluster. yes string

NewConfig

Name Description Required Type Default
locale Defines the locale to be used when initializing a new postgres db cluster (initdb --locale option). This option isn't validated by stolon so initdb will fail if a wrong option is provided. no string
encoding Defines the encoding to be used when initializing a new postgres db cluster (initdb --encoding option). This option isn't validated by stolon so initdb will fail if a wrong option is provided. no string
dataChecksums Defines if data checksums should be enabled when initializing a new postgres db cluster (initdb --data-checksums option). This option isn't validated by stolon so initdb will fail if a wrong option is provided. no bool

PITRConfig

Name Description Required Type Default
dataRestoreCommand defines the command to execute for restoring the db cluster data. %d is replaced with the full path to the db cluster datadir. Use %% to embed an actual % character. Must return a 0 exit code only on success. yes string
archiveRecoverySettings archive recovery configuration yes ArchiveRecoverySettings
recoveryTargetSettings recovery target configuration no RecoveryTargetSettings

StandbyConfig

Name Description Required Type Default
standbySettings standby configuration no StandbySettings
archiveRecoverySettings archive recovery configuration no ArchiveRecoverySettings

ArchiveRecoverySettings

Name Description Required Type Default
restoreCommand defines the command to execute for restoring the archives. See the related postgresql doc yes string

RecoveryTargetSettings

These parameters are the same as defined in postgresql recovery target settings doc

Name Description Required Type Default
recoveryTarget See recovery_target in the related postgresql doc no string
recoveryTargetLsn See recovery_target_lsn in the related postgresql doc no string
recoveryTargetName See recovery_target_name in the related postgresql doc no string
recoveryTargetTime See recovery_target_time in the related postgresql doc no string
recoveryTargetXid See recovery_target_xid in the related postgresql doc no string
recoveryTargetTimeline See recovery_target_timeline in the related postgresql doc no string

StandbySettings

Name Description Required Type Default
primaryConnInfo connection string to connect to the primary server (its value will be placed in the primary_conninfo parameter of the instance recovery.conf file. See the related postgresql doc yes string
primarySlotName optional replication slot to use (its value will be placed in the primary_slot_name parameter of the instance recovery.conf file. See the related postgresql doc no string
recoveryMinApplyDelay delay recovery for a fixed period of time (its value will be placed in the recovery_min_apply_delay parameter of the instance recovery.conf file. See the related postgresql doc no string

Special Types

duration types (as described in https://golang.org/pkg/time/#ParseDuration) are signed sequence of decimal numbers, each with optional fraction and a unit suffix, such as "300ms", "-1.5h" or "2h45m". Valid time units are "ns", "us" (or "µs"), "ms", "s", "m", "h".

Cluster Specification management

It's possible to replace the whole current cluster specification or patch only some parts of it (see https://tools.ietf.org/html/rfc7386). Currently the specification must be provided in json format.

Cluster Specification patching

stolonctl --cluster-name=mycluster update --patch '{ "synchronousReplication" : true }'

You can also pass the cluster specification or a patch as a file using the -f option:

stolonctl --cluster-name=mycluster update --patch -f spec.json

Using - as the file name means stdin:

echo '{ "synchronousReplication" : true }' | stolonctl --cluster-name=mycluster update --patch -f -

Cluster Specification replace

This command will replace the whole cluster specification. The unspecificed options will be populated with their defaults.

stolonctl --cluster-name=mycluster update '{ "requestTimeout": "10s", "sleepInterval": "10s" }'

Examples

Set some postgres parameters:

You can patch the cluster specification providing postgres parameters. For example, if you want to set log_min_duration_statement = 1s you can do:

stolonctl --cluster-name=mycluster update --patch '{ "pgParameters" : {"log_min_duration_statement" : "1s" } }'

Remove some postgres parameters

To remove a postgres parameter just patch the cluster spec setting the parameter's value to null:

stolonctl --cluster-name=mycluster update --patch '{ "pgParameters" : {"log_min_duration_statement" : null } }'

Remove all the postgres parameters

To remove all the postgres parameters just patch the cluster spec setting pgParameters value to null:

stolonctl --cluster-name=mycluster update --patch '{ "pgParameters" : null }'