Failover of node taking ~15 sec when the node down is orchestrator node and timeout is 5 sec

1. Create a 3 node cluster and have a bucket in it (Sample bucket will do)
2. Enable autofailover with 5 seconds timeout
3. Stop the couchbase server on the orchestrator node.
4. Wait for the failover to occur.

We expect the failover to occur in ~5 seconds. But the failover is triggered after 15 sec after failure detection.