Why it matters
The engine, size and data stay the same, so it is a price cut with no change for your application.How nOps finds it
nOps reads the last 7 days of on-demand ElastiCache node billing. m-family nodes are compared with m6g and m7g, r-family with r6g and r7g, and t3 with t4g, all of the same size, at public list prices in the same region. The saving is the node’s on-demand cost, scaled to a month, times the list-price difference to the cheapest target. Nodes not running in the last 2 days, nodes mostly covered by reserved nodes or credits, and savings under $5 a month are left out.Before you start
An engine version that supports Graviton nodes. Permission to modify the cluster.How to implement
- Check that your engine and version support the target node type.
- Change the node type (ElastiCache console: your cluster → Modify → Node type, or
aws elasticache modify-replication-group --replication-group-id GROUP --cache-node-type NEW_TYPE --apply-immediately). - Make the change in a quiet period and watch memory and evictions afterwards.
- Update the node type in Terraform or CloudFormation so a later deployment does not change it back.
Risks and considerations
- Changing the node type replaces the nodes. For Valkey and Redis OSS replication groups this is online; for Memcached it means a new cluster.
- Reserved nodes bought for the Intel type do not cover the Graviton type.
- To roll back, modify the node type back.
FAQs
Do I need to change my application?
Do I need to change my application?
No. The cache engine and protocol are the same on Graviton.