Backup and restoration
PolarDB-X 1.0 supports instance-level and database-level backup and restoration. Backups run in automatic or manual mode, and you can restore data to a new PolarDB-X 1.0 instance or a new ApsaraDB for RDS (RDS) instance from an existing backup set.
How it works
PolarDB-X 1.0 backup and restoration depends on log backup, which is delegated to the underlying RDS instances. The log backup policy you configure in the PolarDB-X 1.0 console is automatically synchronized to all underlying RDS instances.
When you restore data at the instance level, PolarDB-X 1.0 automatically adjusts the instance type and zone based on current inventory. After restoration completes, confirm and change the instance type and zone if needed to avoid impacting your workload.
Choose a backup method
PolarDB-X 1.0 provides two backup methods. Choose based on your consistency requirements and version.
| Backup method | Backup level | Restore options | Version requirement | Performance impact | Best for |
|---|---|---|---|---|---|
| Fast backup | Instance or database | By backup set or by point in time | All versions | Minimal | Routine business workloads |
| Consistent backup | Instance or database | By backup set only | Version 5.3.8 and later | Locks distributed transactions for seconds; SQL response time has sub-second jitter during the lock window | Financial and high-consistency transaction workloads |
Key difference: Fast backup ensures per-shard data consistency but not global consistency across shards. Consistent backup guarantees global data consistency — suitable when your workload spans multiple shards and data accuracy across all shards is critical.
Run consistent backups during off-peak hours to minimize the impact of the distributed transaction lock on response times.
Prerequisites
Before you begin, ensure that you have:
All underlying RDS instances mounted to your PolarDB-X 1.0 instance running normally
All tables with primary keys (required for consistent backup and restoration accuracy)
Log backup enabled — backup and restoration depends on it
Set an automatic backup policy
The automatic backup policy is disabled by default. Enable it before relying on scheduled backups.
Log on to the PolarDB-X 1.0 consolePolarDB-X console and select a region.
In the left-side navigation pane, click Instances, then click the instance name.
In the left-side navigation pane, choose Data Recovery > Backup and Recovery.
Click the Backup Policy tab.
Click Edit to configure the policy. If you set Backup Level to Database
Set the backup parameters and click OK.
PolarDB-X 1.0 itself does not support log backup. The log backup policy you configure here is applied to all underlying RDS instances. Do not modify this policy directly in the RDS console — doing so may break backup set validity.
Configure local binary logs
Local binary log settings control how binary logs are retained on the underlying RDS instances.
Log on to the PolarDB-X console and select a region.
In the left-side navigation pane, click Instances, then click the instance name.
In the left-side navigation pane, choose Data Recovery > Backup and Recovery.
Click the Local Log Settings tab.
Click Edit. In the Local Binlog Settings dialog box, modify the settings.
Click OK.
Local binary log settings are applied to all underlying RDS instances.
Run a manual backup
Do not run Data Definition Language (DDL) operations during backup. DDL operations during a backup can cause instance-level backup and restoration to fail.
Log on to the PolarDB-X 1.0 console and select a region.
In the left-side navigation pane, click Instances, then click the instance name.
In the left-side navigation pane, choose Data Recovery > Backup and Recovery.
Click Data Backup.
Set Backup Method (Fast Backup or Consistent Backup) and Backup Level.
Click OK to submit the backup task.
Click the Data Backup Set tab to view available backup sets. Track progress in the Task Progress list in the upper-right corner.
Restore data
Instance-level restoration creates a new PolarDB-X 1.0 instance and new underlying RDS instances from the backup data.
Instance-level restoration is billed based on the actual billing method and instance type of the restored instance. For pricing details, see Pricing.
Log on to the and select a region.
In the left-side navigation pane, click Instances, then click the instance name.
In the left-side navigation pane, choose Data Recovery > Backup and Recovery.
In the upper-right corner, click Data Recovery(Original Clone Instance).
In the dialog box, select a restoration method and configure the parameters:
By backup set: Select a valid backup set from the list.
By time: Select a specific point in time and a restoration level. Available only with fast backup.
Click Precheck to verify that a valid backup set is available. If the precheck fails, restoration cannot proceed — check that log backup is enabled and that a valid backup set exists for the target time range.
After the precheck succeeds, on the Verification for Recovery page, click Enable to start restoration.
Track progress in the Task Progress list in the upper-right corner. After restoration completes, the new instance appears on the PolarDB-X 1.0 Instance Management page.