SGShardedBackup

Kind: SGShardedBackup

listKind: SGShardedBackupList

plural: sgbackups

singular: sgbackup

shortNames sgbkp

The SGShardedBackup custom resource represents a backup of the sharded Postgres cluster. Backups are created automatically by a cron job configured using the settings in the backup configuration or manually by creating a SGShardedBackup.

Example:


apiVersion: stackgres.io/v1
kind: SGShardedBackup
metadata:
  name: backup
spec:
  sgShardedCluster: stackgres
  managedLifecycle: true
status:
  sgBackups:
  - backup-coord
  - backup-worker0
  - backup-worker1
  process:
    status: Completed
    jobPod: backup-backup-q79zq
    timing:
      start: "2020-01-22T10:17:24.983902Z"
      stored: "2020-01-22T10:17:27.183Z"
      end: "2020-01-22T10:17:27.165204Z"
  backupInformation:
    postgresVersion: "11.6"
    size:
      compressed: 6691164
      uncompressed: 24037844

See also Backups section.

The SGShardedBackup represents a manual or automatically generated sharded backup of an SGShardedCluster configured with an SGObjectStorage.

When a SGShardedBackup is created a Job will perform a full backup for the coordinator and each worker that will be stored in a SGBackup and will update the status of the SGShardedBackup with the all the information required to restore it and some stats (or a failure message in case something unexpected happened). After an SGShardedBackup is created the same Job performs a reconciliation of the sharded backups by applying the retention window that has been configured in the SGObjectStorage and removing the sharded backups with managed lifecycle and the WAL files older than the ones that fit in the retention window. The reconciliation also removes backups (excluding WAL files) that do not belongs to any SGShardedBackup. If the target storage of the SGObjectStorage is changed deletion of an SGShardedBackup backups with managed lifecycle and the WAL files older than the ones that fit in the retention window and of backups that do not belongs to any SGShardedBackup will not be performed anymore on the previous storage, only on the new target storage.

A manual or automatically generated sharded backup of an SGShardedCluster configured with an SGObjectStorage.

When an SGShardedBackup is created a Job will perform a full sharded backup of the database and update the status of the SGShardedBackup with all the information required to restore it and some stats (or a failure message in case something unexpected happened). After an SGShardedBackup is created the same Job performs a reconciliation of the sharded backups by applying the retention window that has been configured in the SGObjectStorage and removing the sharded backups with managed lifecycle and the WAL files older than the ones that fit in the retention window. The reconciliation also removes sharded backups (excluding WAL files) that do not belong to any SGBackup. If the target storage of the SGObjectStorage is changed deletion of an SGBackup sharded backups with managed lifecycle and the WAL files older than the ones that fit in the retention window and of sharded backups that do not belong to any SGBackup will not be performed anymore on the previous storage, only on the new target storage.

Property
 Description

Workaround for hugo bug not rendering first table row

apiVersion
string		 stackgres.io/v1
Constraints: required, immutable
kind
string		 SGShardedBackup
Constraints: required, immutable
metadata
object		 Refer to the Kubernetes API documentation for the fields of the metadata field.
Constraints: required, updatable
spec
object


Constraints: required, updatable
status
object


Constraints: optional, updatable

SGShardedBackup.spec

↩ Parent

Property
 Description

Workaround for hugo bug not rendering first table row

managedLifecycle
boolean
Indicate if this sharded backup is permanent and should not be removed by the automated retention policy. Default is false.

Constraints: optional, updatable
maxRetries
integer
The maximum number of retries the backup operation is allowed to do after a failure.

A value of 0 (zero) means no retries are made. Defaults to: 3.


Constraints: optional, updatable
reconciliationTimeout
integer
Allow to set a timeout for the reconciliation process that takes place after the backup.

If not set defaults to 300 (5 minutes). If set to 0 it will disable timeout.

Failure of reconciliation will not make the backup fail and will be re-tried the next time a SGBackup or scheduled backup Job takes place.


Constraints: optional, updatable
Default: 300
retryDelay
string
An ISO 8601 duration in the format PnDTnHnMn.nS, that specifies the initial delay used by the exponential backoff applied when a command run by the backup script is retried and in general by unbounded retries and in some cases without exponential backoff.

Defaults to: PT1S.


Constraints: optional, updatable
retryLimit
integer
The maximum number of times a command run by the backup script is retried before failing.

Defaults to: 10.


Constraints: optional, updatable
retryMaxDelay
string
An ISO 8601 duration in the format PnDTnHnMn.nS, that specifies the maximum delay the exponential backoff is allowed to reach when a command run by the backup script is retried.

Defaults to: PT1M.


Constraints: optional, updatable
sgShardedCluster
string
The name of the SGShardedCluster from which this sharded backup is/will be taken.

If this is a copy of an existing completed sharded backup in a different namespace the value must be prefixed with the namespace of the source backup and a dot . (e.g. <sharded cluster namespace>.<sharded cluster name>) or have the same value if the source sharded backup is also a copy.


Constraints: optional, updatable
timeout
integer
Allow to set a timeout for the backup creation.

If not set it will be disabled and the backup operation will continue until the backup completes or fails. If set to 0 is the same as not being set.

Make sure to set a reasonably high value in order to allow for any unexpected delays during backup creation (network low bandwidth, disk low throughput and so forth).


Constraints: optional, updatable

SGShardedBackup.status

↩ Parent

Property
 Description

Workaround for hugo bug not rendering first table row

backupInformation
object


Constraints: optional, updatable
process
object


Constraints: optional, updatable
sgBackups
[]string
The list of SGBackups that compose the SGShardedBackup used to restore the sharded cluster.

Constraints: optional, updatable

SGShardedBackup.status.backupInformation

↩ Parent

Property
 Description

Workaround for hugo bug not rendering first table row

postgresVersion
string
Postgres version of the server where the sharded backup is taken from.

Constraints: optional, updatable
size
object


Constraints: optional, updatable

SGShardedBackup.status.backupInformation.size

↩ Parent

Property
 Description

Workaround for hugo bug not rendering first table row

compressed
integer
Size (in bytes) of the compressed sharded backup.

Constraints: optional, updatable
Format: int64
uncompressed
integer
Size (in bytes) of the uncompressed sharded backup.

Constraints: optional, updatable
Format: int64

SGShardedBackup.status.process

↩ Parent

Property
 Description

Workaround for hugo bug not rendering first table row

failure
string
If the status is failed this field will contain a message indicating the failure reason.

Constraints: optional, updatable
jobPod
string
Name of the pod assigned to the sharded backup. StackGres utilizes internally a locking mechanism based on the pod name of the job that creates the sharded backup.

Constraints: optional, updatable
status
string
Status of the sharded backup.

Constraints: optional, updatable
timing
object


Constraints: optional, updatable

SGShardedBackup.status.process.timing

↩ Parent

Property
 Description

Workaround for hugo bug not rendering first table row

end
string
End time of sharded backup.

Constraints: optional, updatable
start
string
Start time of sharded backup.

Constraints: optional, updatable
stored
string
Time at which the sharded backup is safely stored in the object storage.

Constraints: optional, updatable