Changing an instance or upgrading its version triggers an instance switchover in ApsaraDB RDS for MySQL. The switchover causes a brief disconnection and temporary service unavailability. Schedule these operations during off-peak hours and make sure your application has an automatic reconnection mechanism.
Downtime by instance type
The duration of service unavailability depends on your instance type:
| Instance type | Switchover behavior | Maximum downtime |
|---|---|---|
| Cluster Edition or high-availability series | Primary/standby switchover | Up to 15 seconds |
| Basic Edition | Downtime depends on data volume | Longer than high-availability series |
If an instance fails, the switchover time may increase.
Behavior when local resources are insufficient
If local resources on the current host are insufficient during an instance change, RDS automatically creates a new instance on a host with sufficient resources and synchronizes data from the original instance. You can continue using the original instance during the synchronization. After synchronization completes, RDS performs an instance switchover, causing up to 15 seconds of service unavailability.