[system test] a large number of errors Port server goxdcr on node 'babysitter_of_ns_1@127.0.0.1' exited with status 1. Restarting. Messages: net/http.(*persistConn).writeLoop(0xc208044420)

steps:
1. 3 nodes in cluster, 4 buckets. run data loader more then a day
2. rebalance in at SRC cluster
rebalance in at DEST cluster
3. Graceful Fail Over(rebalance) for node in SRC cluster, add back(Delta Recovery)
4. click failover, Hard Fail Over for node in SRC cluster A, add back(Full Recovery) and rebalance
6) remove node in SRC cluster(4-1) - stuck

please note, in the previous run was opened MB-14205