kernel panic on ZFS root when writing to RAIDZ

One way would be to tell us what you were doing at the time of the crash, what OS version and platform you have and if you have any special settings.

You know, we aren't magicians; we cannot guess your hardware and software setup.
 
Fair enough da1.

I was running chown -R $user /nas/backup my RAIDZ share (on a supermicro aoc-sat2-mv8 SATA controller) when it went down. On 8.1, it's crashed like this for as much as bash-tabbing files on the RAIDZ.
 
saoirse87 said:
On 8.1, it's crashed like this for as much as bash-tabbing files on the RAIDZ.
Click to expand...
Do all of these crashes happen on the same ZFS pool? I've got 60+TB of files on 3 ZFS pools here and haven't seen anything like that, so I'd suspect something unusual in your pool.

You might try downloading the mfsBSD ISO image w/ ZFS v28 from here and try executing the failing # chown in that environment. If it works or you get an error message instead of a panic, that would be informative.

IMPORTANT: You should not upgrade your pool(s) to v28 with that image or they won't mount on current 8.x releases. But the v28 code can work with older on-disk versions.
 
saoirse87 said:
Fair enough da1.

I was running chown -R $user /nas/backup my RAIDZ share (on a supermicro aoc-sat2-mv8 SATA controller) when it went down. On 8.1, it's crashed like this for as much as bash-tabbing files on the RAIDZ.
Click to expand...

Hmm ... interesting; did you try a scrub ?

One more thing ... try configuring a dump device, maybe you will get more info from the dump if it crashes again.
 
Terry_Kennedy said:
Do all of these crashes happen on the same ZFS pool? I've got 60+TB of files on 3 ZFS pools here and haven't seen anything like that, so I'd suspect something unusual in your pool.
Click to expand...

[Same ZFS pool and OpenSolaris started complaining about a disk going bad. I'm ordering a 640gb tonight to zpool replace.

Terry_Kennedy said:
You might try downloading the mfsBSD ISO image w/ ZFS v28 from here and try executing the failing # chown in that environment. If it works or you get an error message instead of a panic, that would be informative.
Click to expand...

Will do. Will update.

da1 said:
One more thing ... try configuring a dump device, maybe you will get more info from the dump if it crashes again.
Click to expand...

[cmd=]ls -la /dev/dumpdev -> /dev/gpt/swap0[/cmd]

What kind of a dump diagnostic is this? I need to read more about it, admittedly.
 
bump. how do I debug the crash guys?
Code: 
lrwxr-xr-x  1 root  wheel  14 Mar  5 12:04 /dev/dumpdev -> /dev/gpt/swap0
what next ?!?!
 
Did you change anything with memory recently? I have an old mobo that is very picky about having 4 sticks of ram, and last night after I swapped to 8gb (4x2gb) of ram, it wasn't stable with all 4 sticks. I bumped the voltage and it booted just fine, but once I tried to do anything on my zfs pool, it had a kernel panic just like that.
 
You must log in or register to reply here.
Back
Top