Dell Unity: SP Panic When Both Management Cables are Removed/Inserted Too Fast in the Same Period

Unity SP may panic sometimes when customers swap management cables too fast.

If MGMT components like ECOM are running on SPA, then we pull out the cable from SPA, components would be failover to SPB. But before the failover completed, if the SPB cable is pulled out and the SPA cable is plugged in back. Then this issue might be triggered.

//spb/EMC/C4Core/logs/start_c4.log
Thu Jul 26 09:47:04 UTC 2018 carrier.sh: changing Connectivity state to 0
Thu Jul 26 09:48:05 UTC 2018 carrier.sh: changing Connectivity state to 1
Thu Jul 26 09:48:14 UTC 2018 peservice.sh: /EMC/fastvp is NOT  mounted, cannot start PEService.exe <========

Like start_c4.log above, after cable was pulled out from SPB at 2018-07-26 09:48:05 , some components were stopped while some not stopped at all, then cable was plugged in back.
Then non-stopped components never restarted, while those stopped components started to start like c4fastvpa. And the components which c4fastvpa depends on like c4loga and metricsluna never stopped in this process.
Ungraceful failover messed up the start-up sequence of components and may make SP panic.

This mechanism is working by design. Customer should not swap network cable too fast. Suggest to wait 5-10 mins before any change on another cable.