Has anyone else had the FC HBAs trigger a coredump on their 3PARs ? We've had it happen on 3 different arrays now (with different code levels as well), wondering if anyone else has had it, analysis emails are worded like this :
[code]This firmware trap is caused by a transient hardware error. For such an error, the appropriate recovery action is to do a "hard" reset of the HBA. That was done by the driver after taking the firmware core dump. If this is the first time, or if it happens very infrequently, no action is needed.[/code]
Yes I've had that happen a few times before. At least one of the HBAs started doing it a couple of times a week and that one was replaced.
So long as it's not frequent and you aren't seeing lots of other errors on the ports it should be okay, everything 'should' be wired with redundancy.
If it is frequent and you are planning some upgrades that might reduce redundancy (fabric work, node reboots etc) then you might want to get it replaced before just in case.
I've seen it too. It was pretty rare. Generally speaking, our normal patching and update cycles would upgrade the 3PAR before the same HBA would core dump again.
Richard Siemers The views and opinions expressed are my own and do not necessarily reflect those of my employer.