[bugs] [illumos gate - Bug #1197] Hang after resilver finished with mpt

[illumos bugs]

Issue #1197 has been updated by Roy Sigurd Karlsbakk.


Also, iostat -en doesn't show anything alarming in particular, and none for the failed drives
----------------------------------------
Bug #1197: Hang after resilver finished with mpt
https://www.illumos.org/issues/1197

Author: Roy Sigurd Karlsbakk
Status: New
Priority: Urgent
Assignee: 
Category: driver - device drivers
Target version: 
Difficulty: Medium
Tags: needs-triage


Hi all

I just had a machine finish resilver after a drive (well, two actually) died. After resilver was finished, the Icinga (ex Nagios) check told me the pool was healthy again, so fine. But then, about 15 minutes later, Icinga complained the check timed out, and the box was unavailable. From a remote, I could see OpenIndiana spamming it with messages:

scsi: WARNING: /pci at 0,0/pci8086,340e at 7/pci1000,30a0 at 0... (mpt0):
Disconnected command timeout for Target 23

This looks familiar - I have seen similar on other servers, also just after resilver. The box is using LSI 3801 and 3081 controllers with the mpt driver. Current OS version is OpenIndiana b148.

It looks like this is the same bug I've hit earlier. I just became aware of the resilver issue when this happened within two days with two different machines (the other is 1700km from here, and I don't have a remote console for it yet - long story).

Is there anything I can do to debug this? I ran 'zpool status' from the console, and it apparently hangs there and won't go anywhere.....

Thank you for any help on this one!

roy


-- 
You have received this notification because you have either subscribed to it, or are involved in it.
To change your notification preferences, please click here: http://www.illumos.org/my/account