On my macbook I
- created a two node MH cluster
- added travel-sample bucket
- added memcached bucket
- added ephemeral bucket
- created a CC node
- added CC node to MH cluster
- gracefail failover of one MH node
- rebalance to get rid of MH node and move load to CC node
Rebalance fails (see RebalanceFailure1 screenshot).
Retried rebalance and it fails again (see RebalanceFailure2 screenshot).