Question

A financial services company runs a session-store on a self-designed Amazon ElastiCache for Redis (cluster mode disabled) deployment with a single primary node in one Availability Zone. During a recent AZ impairment, the primary became unreachable and the application lost all active user sessions, forcing every user to re-authenticate. The team must minimize both recovery time and data loss for future AZ failures while keeping the change fully managed by AWS. What should the DevOps engineer implement?

Accepted Answer

Enable Multi-AZ with automatic failover by adding one or more read replicas in different Availability Zones to the existing replication group.

Answer

Configure an EventBridge rule on the ElastiCache node-failure event that triggers a Lambda function to create a new primary node from the latest daily backup.

Answer

Schedule daily ElastiCache backups to S3 and write an SSM Automation runbook that restores the snapshot into a new cluster when CloudWatch reports the primary unhealthy.

Answer

Migrate the session data to a single-node Memcached cluster spread across multiple Availability Zones to gain built-in AZ redundancy.