Re: Xen 4.0.1 freezes with no output

[Sven Anders <anders@anduras.de>]

Pasi Kärkkäinen schrieb:
> On Wed, Sep 08, 2010 at 11:49:07AM +0200, Sven Anders wrote:
>> Pasi Kärkkäinen schrieb:
>>> On Tue, Sep 07, 2010 at 08:50:10PM +0200, Sven Anders wrote:
>>>
>>>>> Also 2.6.32.21 is the latest pvops dom0 kernel atm.
>>> Yep, that'll give you 2.6.32.21 atm.
>> Ok, I tested the latest one. It crashes too. Same phenomenon.
> 
> Ok, I added xen-devel to CC.

>> I'm still uncertain about this error message:
>>  XENBUS: Unable to read cpu state
>>
>> Are these harmless and can I ignore them? What can be the cause?

Any comment to this ?

> How long does it take before it silently crashes?

Last time it took 26 minutes (running top with 0.5 seconds update in
parallel). Last active process was xend.

This time only 1½ minutes:

"Screenshot":

 top - 14:34:44 up 1 min,  1 user,  load average: 0.23, 0.16, 0.06
 Tasks: 158 total,   2 running, 156 sleeping,   0 stopped,   0 zombie
 Cpu(s):  0.0%us,  0.3%sy,  0.0%ni, 99.7%id,  0.0%wa,  0.0%hi,  0.0%si,  0.0%st
 Mem:   1919620k total,   190480k used,  1729140k free,     4576k buffers
 Swap:        0k total,        0k used,        0k free,    42384k cached

   PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
  6782 root      20   0  2316 1168  872 R    2  0.1   0:00.99 top
     1 root      20   0  2852 1692  548 S    0  0.1   0:01.57 init


Intrestingly in my previous test the "xend" refused to start and the system
did not crash for 1 hour and 20 minutes. Then I rebooted, because the "xend" was
not running. The "xend" gave the following error message:

  File "usr/lib/python2.5/site-packages/xen/util/pci.py", line 1226, in find_capability
    ('Looped capability chain: %s' % self.name))
PciDeviceParseError: Looped capability chain: 0000:07:00.0
[2010-09-08 14:28:26 6885] INFO (SrvDaemon:220) Xend exited with status 1.

This seems to be caused by wrong PCI data (or initialization). An "lspci" gave
the following:

 07:00.0 Ethernet controller: Intel Corporation 82574L Gigabit Network Connection (rev ff)
 08:00.0 Ethernet controller: Intel Corporation 82574L Gigabit Network Connection (rev ff)

The revision of 'ff' is bogus. The pci config data presented in /sys or /proc was
only "ff,ff,...", hence the wrong revision.
I had to switch of the hardware to normalize this.

After this the "xend" was starting and the system crashed...

Regards
 Sven

-- 
 Sven Anders <anders@anduras.de>                 () UTF-8 Ribbon Campaign
                                                 /\ Support plain text e-mail
 ANDURAS intranet security AG
 Messestraße 3 - 94036 Passau - Germany
 Web: www.anduras.de - Tel: +49 (0)851-4 90 50-0 - Fax: +49 (0)851-4 90 50-55

Rechtsform: Aktiengesellschaft - Sitz: Passau - Amtsgericht: Passau HRB 6032
Mitglieder des Vorstands: Dipl.-Inf. Sven Anders, Dipl.-Inf. Marcus Junker
Vorsitzender des Aufsichtsrats: RA Mark Peters