I would suggest you to have an SR opened if not done so already. Support may require switch logs to find the root cause. Have you checked for any errors on switch ports ?

When you say Host lost connectivity are you saying all the Datastores were Inactive?

If yes, did you check the switches, if there was any issue with the switch connectivity or cabling?

When you open an SR with CLARiiON Support, if there is no instance of SP Reboot, the best bet and what they would recommend you would be to check with

- VMware Support

- Switch Support

there is no SP reboot events/ switch events that states connectivity loss or reboot. 

check the esxi vmkernel.log for APD (all paths down) entries. These are the entries created when paths are lost transiently, as opposed to PDL (permanent device loss)

cat /var/log/vmkernel.log |grep -i -C 2 'apd'

That should give some insight into what happened host side so you can correlate with logs in the sp collects.

In VM kernal logs APD error is there

There is no user impact due to this LUN disconnection issue. as the error message itself says LUN disconnected for 35 Milli seconds.

How many connections (I'm assuming that you're using FC connections and not iSCSI) do you have from the host to the array? Does each host have two HBA's and there are two paths for each HBA?

If you have Mirrorview/Sync you should try to not use the same SP ports for host connections.

glen

Yes your are right, it is an FC environment. Unfortunately we have only 2 paths configured for the Host (1/SP) and 2 paths reserved for sync mirror.

Did it only happen once ?

What was the actual message(s) in the vmkernel.log ?

I'm pretty sure you can have up to 4 Front-end ports on the CX4-120. The highest numbered port on the shipped configration will be your MirrorView port and the will not change if you add more ports.

glen