Couchbase rebalance keeps failing because of buckets_shutdown_wait_failed - couchbase

Question

I've updated couchbase cluster from version 2.5.1 to 3.0.3. Upgrade was performed node by node by removing one node from cluster, upgrading it and add it again.
All was working but not the DCP upgrade. I've found in error.log that we have some "Weird vbucket_move_done" for a proximic2 bucket, which isn't used for a while now.
So I've deleted this bucket and try to rebalance but it failed. It was few ours ago and rebalance keeps failing:
Rebalance exited with reason {buckets_shutdown_wait_failed,
[{'ns_1#cb6.xxx.com',
{'EXIT',
{old_buckets_shutdown_wait_failed,
["proximic2"]}}},
{'ns_1#cb2.xxx.com',
{'EXIT',
{old_buckets_shutdown_wait_failed,
["proximic2"]}}},
{'ns_1#cb1.xxx.com',
{'EXIT',
{old_buckets_shutdown_wait_failed,
["proximic2"]}}},
{'ns_1#cb0.xxx.com',
{'EXIT',
{old_buckets_shutdown_wait_failed,
["proximic2"]}}}]}
ns_orchestrator002 ns_1#cb2.xxx.com 20:42:59 - Fri Jun 19, 2015
Failed to wait deletion of some buckets on some nodes: [{'ns_1#cb6.xxx.com',
{'EXIT',
{old_buckets_shutdown_wait_failed,
["proximic2"]}}},
{'ns_1#cb2.xxx.com',
{'EXIT',
{old_buckets_shutdown_wait_failed,
["proximic2"]}}},
{'ns_1#cb1.xxx.com',
{'EXIT',
{old_buckets_shutdown_wait_failed,
["proximic2"]}}},
{'ns_1#cb0.xxx.com',
{'EXIT',
{old_buckets_shutdown_wait_failed,
["proximic2"]}}}]
I'm a bit blocked right now. I've tried to create this proximic2 bucket, but I've got information that it already exists. However this bucket is not visible in web GUI or cli. Couchbase runs on debians 7.8. All nodes are in version 3.0.3.
Marek