Keyboard shortcuts

Press ← or → to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

Backup Policy

Backup policies are used to unify the management of backup times, recoverable versions, and storage space. It is recommended to define a timeline before configuring retention quantities and automatic cleanup rules, and periodically verify that restore points are available.

Define timeline, configure scheduling

Determine the frequency of backups, execution windows, and recovery point intervals based on the frequency of data changes and acceptable data loss windows. The execution window should avoid business peaks and allow sufficient time for database export, compression, encryption, and upload.

Each time the data range, backup destination, or scheduling plan is adjusted, the configuration version, reason for the change, and effective time should be recorded. After the configuration changes, first execute dry-run, and then create a new backup point.

Configure the number of versions, reserve the restore point

Set the number of reserved versions for local and remote, respectively, and reserve at least one historical restore point that has completed recovery verification. The reserved quantity shall cover the daily misoperation recovery, recent failure recovery and longer term business traceability requirements.

Backup files, configuration snapshots, and validation reports should correspond to the same timeline; when restoring, select a restore point that matches the target data time, not just guess the version by file name.

Automatic cleaning to avoid wasted space

Historical files stored locally and on objects should be cleaned up in accordance with their respective retention policies. Before cleaning, execute dry-run to confirm the target path, project identification, prefix, and quantity to be deleted; use --apply after confirming that it is correct.

backup run --upload does not automatically clean up remote history files, remote cleanup requiresppk cleanup object-storageto be performed separately. Alarms should be retained and storage quotas checked when a cleanup fails, and backup or recovery issues cannot be masked by shortening the retention period.

Detailed scheduling and retention rules are available in Scheduling and Retention Policies and Configure Timeline and Versioning Policies.