[System test]: Client time out when data loading with durability=MAJORITY

Build: 6.5.0-4349

Test Steps:

• Create 6 node cluster all kv
• Create 3 buckets with replica 1,1,2
• load data
• as with no majority and with majority (bucket having replica 1 and 2) – >for mad-hatter
• without durability for alice

• Rebalance in kv node
• Rebalance out kv node
• Swap kv nodes
• Update replica of bucket as 2
• Failover -> full recovery 
• Failover -> delta recovery
• Graceful Failover and rebalance out 1 KV
• Rebalance-in KV which were out due to failover
• reset bucket replica (1,1,2)
• Rebalance-in 2 kv
• Rebalance-out 2 kv
• Swap rebalance 2 kv
• Rebalance in 1 kv and out 2 kv
• STOP and START memcached
• Add back all the nodes for cluster reset

We use Gideon as dataloader https://github.com/couchbaselabs/gideon

While running data load I am seeing following error when Durability = MAJORITY 

WARNING:root:cluster timed trying to handle mset

WARNING:root:cluster timed trying to handle mset

WARNING:root:cluster timed trying to handle mset

WARNING:root:cluster timed trying to handle mset

WARNING:root:cluster timed trying to handle mset  

consumer logs: consumer_2.log consumer_1.log consumer_none.log