Hard failover immediately fails over nodes from clusters.
During failover, replica vBuckets are promoted to active vBuckets and active vBuckets on the failed over node are transitioned to replica vBuckets. After a node is failed over, it can be added back to the cluster via delta or full recovery. When a node is added back to the cluster after a failover has been performed, the replica vBuckets on the failed over node are resynchronized and promoted back to active. Topology changes (such as adding, removing, or failing over a server) after a failover, initiates a different type of rebalance operation.
Hard failover is typically used when the node is in a bad state. Auto-failover is a hard failover. The following conditions are usually present when hard failover is used:
- The server node is not healthy.
- Each active vBucket does not have an equivalent replica bucket.
- No replica buckets are present.