A relational database must remain available for writes.
Fail over automatically to a standby in another Availability Zone without changing the application's endpoint.
Which architecture is most appropriate?
Use an RDS Multi-AZ deployment with a synchronous standby and managed failover.
Detailed explanation
An Availability Zone failure makes the database unavailable.
An Availability Zone failure makes the database unavailable.
S3 replication does not provide database connection failover.
S3 replication does not provide database connection failover.
RDS synchronizes to a standby in another Availability Zone and fails over when needed.
RDS synchronizes to a standby in another Availability Zone and fails over when needed.
A read replica is primarily for read scaling and is not the synchronous standby used for automatic failover.
A read replica is primarily for read scaling and is not the synchronous standby used for automatic failover.
Check official documentation
Compare the following points with the official documentation.
Amazon RDS Multi-AZ deploymentの公式説明を確認する。
What to confirm
Multi-AZとRead Replicaの目的を可用性要件から区別できる。
Key points
- RDS Multi-AZ
- Standby instance
- Failover
Notes
- Environment: AWS公式RDSドキュメントの確認
- Confirm current AWS service behavior in the official documentation.
Foundation review
Availability versus read scaling
Multi-AZ protects availability, whereas a read replica primarily scales read traffic.
Stable endpoint
RDS manages the endpoint transition so the application need not hard-code a new address.