Why it matters
It is a price cut that needs no change of node type or cluster size, and usually no application change.How nOps finds it
nOps reads the last 7 complete days of on-demand MemoryDB node billing on Redis OSS for each cluster. The saving is the node cost, scaled to a month, times the difference between the public Valkey and Redis OSS node-hour prices for the same node type and region. Data-written and snapshot charges are not counted. Clusters not running recently, clusters mostly on reserved nodes or credits and savings under $5 a month are left out.Before you start
Confirm your client libraries and any Redis-specific features work with Valkey, ideally on a test cluster.How to implement
- Test your application against a Valkey cluster first.
- Upgrade the engine (MemoryDB console: Clusters → your cluster → Modify → Engine → Valkey, or
aws memorydb update-cluster --cluster-name CLUSTER --engine valkey --engine-version VERSION). - Apply it in a maintenance window and watch the cluster afterwards.
- Update the engine in Terraform or CloudFormation so a later deployment does not switch it back.
Risks and considerations
- Plan the upgrade as one-way: switching a MemoryDB cluster back from Valkey to Redis OSS is not a simple in-place change.
- The upgrade replaces nodes one at a time; use a maintenance window.
- Existing reserved nodes keep applying after the switch.
FAQs
Do I need to change my application?
Do I need to change my application?
Usually not. Valkey uses the same commands and clients as Redis OSS. Test first.