Hi we have a host part of the cluster with one dual port HBA for FC connection to EMC , on a single port there is a regular dead path shown up which changes someother LUN each time a rescan is performed the other port works fine without any issue , had a check on the SFP , cabling and driver and everything looks good , any suggestion
to add up we are facing latency on this host just because of this port were DAVG is reaching as high as few thousands
Out of interest, what scsi sense codes are generated in /var/log/VMkernel.log. (Interpreting SCSI sense codes in VMware ESXi and ESX (289902) | VMware KB)
Interpreting those might help track down where any fault sits.
VMware ESXi SCSI Sense Code Decoder | Virten.net
However, as I'm assuming from the description that there are no issues on any other HBAs in the environment, it sounds like it sits with the HBA Are you able to replace the HBA?
Simon