ElastiCache Enable Data Tiering

Opportunity Name: ElastiCache Enable Data Tiering

AWS Resource Type: ElastiCache (Redis and Valkey)

Opportunity Description:
This Finder identifies Redis and Valkey ElastiCache clusters that can benefit from enabling data tiering, a cost-saving feature that moves less frequently accessed data from expensive in-memory storage (RAM) to lower-cost SSD storage. AWS ElastiCache data tiering allows customers to significantly reduce costs for large datasets that don’t require all data to reside in memory at all times.

By enabling data tiering, you can reduce the size (and cost) of memory-based nodes and rely on SSDs for cold data access — without compromising application performance.

Criteria for identifying the opportunity:

• Cluster is running Redis version 7.0.7 or later, or Valkey version 7.2.0 or later (these are the minimums CloudFix requires; note this is higher than the Redis 6.2 minimum AWS itself imposes for data tiering).
• Cluster is not already running on an r6gd node type (i.e., data tiering is not already enabled).
• Cluster's node type generation is 6 or older (not already on a newer, e.g. r7gd-class, generation).
• Cluster is in a supported AWS region for r6gd data tiering instances.
• A suitable r6gd target instance type exists whose combined memory + SSD capacity exceeds the cluster's current memory, and whose predicted p99 CPU utilization stays at or below 80%.
• The identified target instance type must be cheaper than the current instance type based on on-demand pricing.

Potential Savings (if known):
Savings depend on the size of the dataset and node types. Data tiering enables the use of fewer and/or smaller memory nodes while storing cold data on cheaper SSDs. Customers have seen up to 60% reduction in memory costs with no noticeable performance degradation.

What happens when the Fixer is executed?
This Finder does not currently have an automatic Fixer. Enabling data tiering requires a manual upgrade to supported r6gd node types and a configuration change at the cluster level. This involves:

1. Modifying the cluster to use a compatible instance family.
2. Enabling data tiering in the Redis configuration.
3. Restarting or recreating the cluster.

CloudFix will provide detailed migration guidance, or customers can leverage AWS documentation to complete the change.

Is it possible to roll back once CloudFix implements the fixer?
Yes. You can revert to non-tiered nodes by modifying the instance type and disabling the data tiering configuration. However, this involves downtime and reconfiguration.

Can CloudFix implement the fix automatically once I accept the recommendation?
No. This fix must be applied manually or with assistance from CloudFix Professional Services.

Does this fix require downtime?
Yes. Transitioning to new instance types and enabling tiering may require restarting or recreating clusters, resulting in downtime. Plan the change during maintenance windows.

Additional Resources:

• ElastiCache Data Tiering Overview
• Supported Redis Node Types with Data Tiering
• Modifying ElastiCache Clusters