Introducing WAL Failover

CockroachDB release 24.3 introduces write-ahead log (WAL) failover to minimize the impact of transient disk stalls commonly seen in cloud storage. WAL failover uses a secondary disk as a fail-to for the write-ahead log, mitigating latency introduced by disk stalls. This feature is highly recommended for deployments on AWS, Azure, or GCP that use network storage devices or local SSDs and for on-premise installations where disk stalls are occurring. WAL failover helps CockroachDB maintain its reputation as a resilient multi-cloud distributed SQL database in the market.