Difficulty: Easy
Description
We identified several ElastiCache clusters in your account that are idle — they have received no read or write commands for the past 7 days. Deleting these unused clusters eliminates unnecessary compute costs with no impact on active workloads.
Selection Criteria
The ElastiCache cluster has received no read or write commands for the last 7 days
The cluster is in "Available" state.
Expected Saving
Deleting an idle ElastiCache cluster eliminates 100% of its compute costs.
Example Instance | Engine | Estimated Monthly Cost |
cache.t3.medium | Redis / Memcached | ~$25/month |
cache.r6g.large | Redis | ~$172/month |
cache.m6g.xlarge | Redis | ~$208/month |
⚠️ Reserved Instances : deleting the cluster does not cancel an existing RI — charges continue until expiration. Factor RI commitments into the decision.
Operational Impact
⚠️ CRITICAL : deletion is irreversible. Verify that no application is still using the cluster before proceeding.
Redis without persistence (no RDB/AOF backup) : all in-memory data is permanently lost.
Memcached : no persistence by design — data is always lost upon deletion.
Redis with backup enabled (RDB/AOF) : a snapshot can be taken before deletion, allowing restoration to a new cluster if needed.
⚠️ Reserved Instance : the RI commitment continues billing even after the cluster is deleted — verify RI expiry date before acting.
References