SAP-C02 Question 12
Single answerAn organization is running a multi-tier application on AWS. The application has multiple EC2 instances in an Auto Scaling group for the frontend, a fleet of instances for the backend, and an RDS database for the data layer. The organization wants to ensure high availability, seamless failover of the database, and minimal downtime during maintenance activities. Which solution is most suitable to meet these requirements?
- A
Use a Multi-AZ RDS deployment with automatic failover.
- B
Use RDS Read Replicas in the same region for high availability.
- C
Deploy the RDS database in a single Availability Zone with scheduled backups.
- D
Use DynamoDB with on-demand capacity to replace the RDS database.
Show answer and explanation
Correct answer: A
Explanation
In a multi-tier application, ensuring high availability and seamless failover is critical for the database layer. AWS RDS Multi-AZ deployments are specifically designed for such use cases. They maintain a synchronous standby replica in a different Availability Zone and automatically failover in the event of failures, maintenance, or zone outages. This ensures minimal downtime and no manual intervention during failover events. Other options, like Read Replicas or single-AZ setups, do not meet the high availability requirements.
- A. Correct.
Correct: Multi-AZ RDS deployment provides high availability with automatic failover support. If the primary database instance fails or maintenance occurs, the standby instance in another Availability Zone takes over automatically.
- B. Incorrect.
Incorrect: RDS Read Replicas are designed for read scalability, not high availability. Failover does not happen automatically with Read Replicas.
- C. Incorrect.
Incorrect: Deploying the database in a single Availability Zone does not provide high availability or fault tolerance. Scheduled backups only help in disaster recovery, not seamless failover.
- D. Incorrect.
Incorrect: DynamoDB is a NoSQL database and may not be a suitable replacement for an RDS database in all scenarios. Furthermore, it does not inherently address traditional relational database failover requirements.