Aiven for Valkey™ clustering now generally available

Aiven for Valkey™ clustering is now generally available (GA). Clustering distributes your data across multiple shards so your service can handle larger datasets and higher throughput than a single node can support. It includes high availability with automatic failover, replicas on each shard, and automatic resharding when you change the number of primary nodes.

Clustered services are backed up automatically. Each shard is backed up independently, so a restored cluster isn't point-in-time consistent across shards, and you can only restore to a plan with the same shard count. You can't convert an existing service to or from a cluster plan, or migrate data into a cluster from an external Redis or Valkey server. Applications that connect with a standalone-mode client need to switch to a cluster-aware Valkey or Redis client.

To get started, choose a cluster-N plan when you create an Aiven for Valkey service, where N is the per-node memory in GiB (for example cluster-4 or cluster-8). Set the shard_count and replicas configuration options to size your cluster. replicas defaults to 0, so set it to at least 1 if you want automatic failover. Resharding requires Valkey 9.x or later. For more details, see Aiven for Valkey clustering.