Are My RDS DB Instances Still Available During Storage Scale-up and Instance Class Change?

Currently, you can scale up storage space and change the vCPU or memory of a DB instance.

• When storage space is being scaled up, RDS DB instances are still available and services are not affected. However, you cannot delete or reboot DB instances that are being scaled.
• During the change of the vCPU or memory, the network is intermittently disconnected for one or two times in seconds. To prevent service interruption, perform the operation during off-peak hours. Changing an instance class takes 5 to 15 minutes.

  After you change the instance class of a DB instance, the DB instance will be rebooted and the cache in the memory will be automatically cleared. A failover is triggered during the change. The change process involves changing the instance class of the standby instance, rebooting the standby instance, performing a failover, changing the instance class of the new standby instance, and then rebooting the new standby instance. Workloads can be interrupted during the failover. The length of the interruption depends on how long the failover will take.