Build remote copies of S3-compatible object data to support disaster recovery and business continuity. Link independent RustFS sites or replicate selected bucket data. Plan DNS, traffic routing, and application failover separately.
Directional bucket replication copies objects from a versioned source bucket to one or more versioned targets. Filter by prefix or tags and use separate rules to keep selected datasets aligned without linking every site.
Synchronous mode waits for a replication attempt before returning; asynchronous mode transfers objects in the background. Monitor replication status and failures: source write success alone does not guarantee a completed remote copy.
Limit replication with prefix and tag filters. Objects that match no enabled rule stay local. Existing-object replication and delete propagation are configured separately; verify backfill completion before relying on a destination for recovery.
Configure multiple destinations with separate filters and replication modes. S3-compatible targets require enabled versioning, suitable permissions, and compatible replication operations. Validate each destination before production use.
Site replication links independent RustFS deployments so buckets, object versions, delete markers, and supported IAM data stay aligned across sites. It is broader than a single bucket rule and remains asynchronous.
Participating sites exchange replicated state so applications can write at more than one location while object versions and supported identity data stay aligned.
Configure multiple remote sites in one site-replication relationship. Each site needs a stable S3 API endpoint, reachability, and bucket versioning.
Set up and manage site replication from the Web Console or the rc CLI. Keep administration credentials on a secured host.