Creates a managed redis cluster following best practices using AWS Elasticache.
| Name | Default | Description |
|---|---|---|
redis_version |
"8.0" |
Engine version. 7.x and below run Redis; 8.0 and above run Valkey. |
node_type |
"cache.t3.small" |
Node size for every node in the group. |
enforce_ssl |
true |
Use TLS (rediss). Turning this off also disables the auth token. |
multi_az |
false |
Enable Multi-AZ with automatic failover. When the primary fails, AWS promotes a replica and moves the primary endpoint within about a minute. Requires num_replicas >= 1. Not supported on cache.t2 node types. |
num_replicas |
1 |
Number of read replicas. The reader endpoint round-robins across them. 0 gives a single node: no reader endpoint, no failover. |
Changing multi_az or num_replicas on an existing deployment is an in-place modify of the replication group, not a replacement. No data is lost and the primary endpoint does not change.
Multi-AZ adds no charge beyond the nodes themselves. Each replica is one more node of node_type.
Set num_replicas: 0 for a cheap single-node cache. reader_host is "" in that case.