DBS-C01 Question #61: Real Exam Question with Answer & Explanation

Question

A Database Specialist is troubleshooting complaints from an application's users who are experiencing performance issues when saving data in an Amazon ElastiCache for Redis cluster with cluster mode disabled. The Database Specialist finds that the performance issues are occurring during the cluster's backup window. The cluster runs in a replication group containing three nodes. Memory on the nodes is fully utilized. Organizational policies prohibit the Database Specialist from changing the backup window time. How could the Database Specialist address the performance concern? (Select TWO.)

Options

• AAdd an additional node to the cluster in the same Availability Zone as the primary.
• BConfigure the backup job to take a snapshot of a read replica.
• CIncrease the local instance storage size for the cluster nodes.
• DIncrease the reserved-memory-percent parameter value.
• EConfigure the backup process to flush the cache before taking the backup.

Explanation

Explanation

During ElastiCache Redis backups, a fork() system call is used, which requires additional memory overhead - when memory is fully utilized, this causes performance degradation on the node performing the backup. Option B is correct because redirecting the backup to a read replica instead of the primary node offloads the memory pressure and fork() overhead away from the node handling write operations, protecting application performance. Option D is correct because increasing the reserved-memory-percent parameter explicitly reserves memory for Redis background processes (including the fork() operation during snapshots), preventing memory exhaustion during backups.

Why the Distractors Are Wrong

• A – Adding a node to the same AZ as the primary doesn't reduce the memory pressure on the node performing the backup; it doesn't change which node backs up or how much memory is available.
• C – ElastiCache is an in-memory store; local instance storage size is irrelevant to Redis memory management during backups.
• E – Flushing the cache before a backup would destroy all cached data, which defeats the entire purpose of the backup and the cache itself.

💡 Memory Tip

Think "Fork needs room" - Redis backups fork a process that needs free memory. You solve this by either reserving memory (D) or moving the fork to a replica (B), keeping your primary node free to serve writes.

No community discussion yet for this question.