[upgrade] Rebalance failed after add vulcan nodes to 5.0.0 cluster

Create a 2 nodes cluster with 5.0.0 version (57 and 34)

Install Couchbase server 5.5.0-2432 on 2 nodes (143 and 108)

Add 2 nodes 143 and 108 to 5.0.0 cluser.

Rebalance.  Rebalance failed with errors
[user:error,2018-04-08T08:16:48.929-07:00,ns_1@172.23.105.108:<0.3356.0>:ns_orchestrator:do_log_rebalance_completion:1122]Rebalance exited with reason {mover_crashed,
{unexpected_exit,
{'EXIT',<0.5142.0>,
{{noproc,
{gen_server,call,
[compaction_new_daemon,
{inhibit_view_compaction,
<<"standard_bucket1">>,<0.3706.0>},
infinity]}},
{gen_server,call,
[{'janitor_agent-standard_bucket1',
'ns_1@172.23.107.143'},
{if_rebalance,<0.3706.0>,
{dcp_takeover,'ns_1@172.23.106.57',504}},
infinity]}}}}}
 

I saw some failure with this error in upgrade job

Test to run newupgradetests.MultiNodesUpgradeTests.online_upgrade_rebalance_in_out,initial_version=5.0.0-3519,items=10000,upgrade_version=5.5.0-2432,skip_init_check_cbserver=true