Observing high rebalance time up to 10hrs on 4 node cluster with 2 Billion items and 15k continuous ops

Observing high rebalance time up-to 10hrs  when multiple rebalances (in, out , swap) are made on same 4 node cluster with continuous ops running on the cluster 

It looks like rebalancing goes on hung status when data compaction kicks in  . Please take a look into this issue 

Data set up :

1. Initial data load - 2 Billion docs
2. Continuous operatons - 15 k, 5% Update , 5% Create

Rebalancing steps :

1. Step  - Rebalance-In (4-3)
2. Step  - Rebalance-Out (3-4)
3. Step - Rebalance Swap ( 4 -4 )

 
Multi Job - 
http://perf.jenkins.couchbase.com/job/Titan-reb_Multitests/45/
 
Individual Jobs (unaffected) - 
http://perf.jenkins.couchbase.com/job/Titan-reb_Multitests/42/
 
http://perf.jenkins.couchbase.com/job/Titan-reb_Multitests/37/