Details
Description
Any OS, version 918 and above
1. Create 3 node cluster
2. Add 1 default bucket with replica=1
3. Add 500K items
4. Wait till all queues drain and replication is complete
5. Graceful failover one node
6. When Step 5 is executing stop graceful failover using UI button
7. Start Graceful failover again
Step 7 never happens since we are allowed only rebalance after we stop graceful failover once. By this I mean graceful over to continue is not possible- just as rebalance. We will have to pick the node again to graceful failover as in Step 5.
https://s3.amazonaws.com/bugdb/jira/MB-11680/log_failover_issue.tar.gz, logs are for version 943