Kernel panic 0.6.0 RC5 on Centos 6

[scotttcollins]

Hi,

I am running ZFS on Linux 0.6.0 RC5 with Centos 6 and kernel 2.6.32-71.29.1.el6.x86_64 on a 4 CPU server with 8 GB RAM. Last Friday, the server had a kernel panic caused by ZFS and the screen output is below.

[root]~>
Message from syslogd@edge-nfs00 at Nov 25 00:26:34 ...
kernel:Oops: 0002 [#1] SMP

Message from syslogd@edge-nfs00 at Nov 25 00:26:34 ...
kernel:last sysfs file: /sys/devices/system/cpu/cpu3/cache/index1/shared_cpu_map

Message from syslogd@edge-nfs00 at Nov 25 00:26:34 ...
kernel:Stack:

Message from syslogd@edge-nfs00 at Nov 25 00:26:34 ...
kernel:Call Trace:

Message from syslogd@edge-nfs00 at Nov 25 00:26:34 ...
kernel:Code: 89 e9 48 83 c0 01 48 c1 e0 04 4a 8b 14 20 48 c1 e2 03 e8 f0 8b fe ff 85 c0 0f 84 68 01 00 00 48 8b 85 28 ff ff ff be d0 00 00 00 <48> c7 00 00 00 00 00 48 8b 95 20 ff ff ff 48 8b bd 60 ff ff ff

Message from syslogd@edge-nfs00 at Nov 25 00:26:34 ...
kernel:CR2: 0000000000000000

Message from syslogd@edge-nfs00 at Nov 25 00:26:34 ...
kernel:Kernel panic - not syncing: Fatal exception
Write failed: Broken pipe

[behlendorf]

Do you still have the console output logged somewhere? At a minimum we need to see the stack track which was dumped at the time of the failure, without it there's not much we can do.

[scotttcollins]

Unfortunately, another admin went to the console and tried things that
cleared the console otherwise I would have sent it along. Sorry. I
thought I would log what I have in any event ...

I am going to try to recreate what the system was doing at the time of the
panic as well and will let you know if it is repeatable. I plan to do that
tomorrow night.

-Scott

On Mon, Nov 28, 2011 at 1:10 PM, Brian Behlendorf <
[email protected]

wrote:

Do you still have the console output logged somewhere? At a minimum we
need to see the stack track which was dumped at the time of the failure,
without it there's not much we can do.

---

Reply to this email directly or view it on GitHub:
#468 (comment)

[behlendorf]

You should probably also update to spl/zfs-0.6.0-rc6 which did have a fix for a NULL dereference which could be this issue.

[scotttcollins]

I will, thank you ...

Cheers,
-Scott

On Mon, Nov 28, 2011 at 1:51 PM, Brian Behlendorf <
[email protected]

wrote:

You should probably also update to spl/zfs-0.6.0-rc6 which did have a fix
for a NULL dereference which could be this issue.

---

Reply to this email directly or view it on GitHub:
#468 (comment)

[behlendorf]

Closing the issue since it's stale and we've fixed quite a few issues which could have caused this. Without a stack trace its impossible to say.