RDS provides the high availability. Select the primary/standby mode.
Also called an unplanned handover. If the primary DB instance fails, the system will automatically switch to the standby DB instance within 5 minutes. No human intervention is required. The connection IP address remains unchanged. DB instances cannot be accessed during the failover. You need to configure automatic reconnections between applications and RDS DB instances to ensure near-continuous availability.
Also called a planned handover. When a DB instance is running properly, you can manually perform a primary/standby switchover if it is needed.
Alternatively, click in the DB instance topology on the Basic Information page to manually trigger a primary/standby switchover.
A primary/standby switchover may cause service interruption for some seconds or minutes (depending on the replication delay). If the replication delay is too long, a small amount of data may be lost. To prevent traffic congestion, perform a switchover during off-peak hours.
If the replication status is Available and the replication delay is greater than 300s, the primary/standby switchover task cannot be delivered.