SAA-C03 Design Resilient Architectures Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Store the database credentials in AWS Secrets Manager and rotate them automatically on a schedule.

  Why it's wrong here

  Automatic credential rotation improves security posture by limiting the lifetime of secrets, but it has no effect on database availability. If the primary DB instance fails, the application still cannot read data, so this option does not satisfy the resilience requirement described.

• ✓

  Convert the DB instance to an Amazon RDS Multi-AZ DB instance deployment so a standby is maintained in another Availability Zone.

  Why this is correct

  Amazon RDS Multi-AZ maintains a synchronously replicated standby in a different Availability Zone and automatically fails over by promoting the standby and updating the DNS endpoint, so the application reconnects without operator action. This directly addresses database failure and minimizes application downtime.

• ✗

  Create a read replica in the same Availability Zone and point read traffic to it using a custom application load-balancing layer.

  Why it's wrong here

  A read replica in the same Availability Zone shares the failure domain of the primary, so an AZ event can remove both. Read replicas use asynchronous replication and must be manually promoted, and application-level routing adds complexity without meeting the requirement to minimize outage time.

• ✗

  Enable automated backups with a longer retention period and increase the DB instance class size.

  Why it's wrong here

  Automated backups enable point-in-time restore, not failover, and restoring creates a new instance that takes time and changes the endpoint. Increasing the instance class improves capacity but does not protect against an Availability Zone or instance failure, so downtime would still be significant.

Go deeper

Related to this question