The correct choice is Enable a Multi-AZ Redis deployment with automatic failover because it provides continuous replication to replicas in other Availability Zones and automatically promotes a healthy replica if the primary fails. This design yields very low recovery time objectives and low recovery point objectives while keeping application latency impact to a minimum.

Enable a Multi-AZ Redis deployment with automatic failover achieves resilience by maintaining standby replicas in separate Availability Zones and by orchestrating fast, automated failover without manual intervention. This approach reduces the chance of data loss and keeps the cache available for real-time gaming traffic so that the application can continue to serve low-latency reads.

The option Schedule automatic daily snapshots during the quietest traffic window is insufficient because periodic snapshots can leave up to a day of data unprotected and restores are time consuming and disruptive to latency.

The option Add read replicas for the Redis primary in separate Availability Zones to lower the risk of data loss improves read scalability but does not by itself provide automated detection and promotion of a failed primary, and replication lag can still create gaps in protection.

The option Perform manual backups using the Redis append-only file feature adds operational overhead and does not deliver rapid, automated failover for an AZ or node failure so it does not meet the low RTO and RPO requirements.

Full AWS Practitioner Certification Question