When your primary instance's region is unavailable or replication lag is too high for a normal switchover, promote the disaster recovery (DR) instance to restore read/write access immediately. After promotion, the replication link is broken, and both instances run independently.
Prerequisites
Before you begin, ensure that you have:
A disaster recovery instance group with Running status
When to promote
Promote the DR instance when either of the following conditions applies:
Region outage: The primary instance's region is unavailable and you cannot wait for recovery.
High replication lag: Data synchronization latency is too high for a normal switchover.
Before promoting, check the replication performance to estimate how much unsynchronized data may be lost.
Promotion can cause permanent loss of data not yet replicated from the primary instance. Evaluate the business impact before proceeding.
Promote the disaster recovery instance
Go to the Global Distributed Cache page.
Click the instance ID of the target instance group.
In the RDS Instance List section at the bottom of the page, find the DR instance. In the Actions column, click Promote To Primary Instance.
Review the promotion effects, then click Promote To Primary Instance to confirm.
What happens after promotion
| Entity | What changes |
|---|---|
| Disaster recovery instance group | Remains with Running status but contains only the newly promoted primary instance |
| Original DR instance | Becomes the new primary instance; status transitions from Deleting Replication Link to Running; switches to read/write (RW) mode; runs as an independent ApsaraDB RDS for SQL Server instance |
| Original primary instance | Removed from the group; continues running as an independent ApsaraDB RDS for SQL Server instance with all data, connections, and access intact |
| Replication link | Broken; data synchronization charges stop |
To view the new primary instance, go to the Instances page and switch to the destination region.
To view the original primary instance, go to the Instance List page and switch to the destination region.
Post-promotion checklist
After promotion, assess your services and complete the following steps to restore your application's database connectivity:
[ ] Update the endpoint: Change the connection address in your application to the endpoint of the new primary instance.
[ ] Configure network access: Set an IP whitelist or verify network connectivity between your application and the new primary instance.
What's next
Set up a new disaster recovery system: Add a disaster recovery instance to the new primary instance to restore cross-region disaster recovery coverage.
Clean up the original primary instance: The original primary instance continues to run and incur charges. If you no longer need it, release or unsubscribe from the instance to stop billing.
Delete the disaster recovery instance group: Once the group contains only the new primary instance, you can delete the group without affecting the instance — its accounts and data are preserved. Go to the Global Distributed Cache page, click Delete in the Actions column, and then click OK.