[Xen-users] (open iscsi) initiator crashes

[Dominik Klein <dk@xxxxxxxxxxxxxxxx>]

I think I have found the reason for this:

The setup runs just fine until I set the xen dom0 to only use one of the 
four CPUs in my machine (actually 2 HT CPUs).
So with
(dom0-cpus 0)
in /etc/xen/xend-config.sxp
this works. The while-loop actually ran fine for 2 days straight.

With
(dom0-cpus 1)
it crashes as described within a few minutes.

I will cc this to the xen-list.
Full thread here: 
http://groups.google.com/group/open-iscsi/browse_thread/thread/495b17fa2ab52e3c/8acf2cc82a384646?lnk=gst&q=crashes&rnum=8
I'll be happy to supply additional information when needed.

Regards
Dominik

Dominik Klein schrieb:
>>> Recompile open-iscsi with
>>>
>>> make clean
>>> make DEBUG_SCSI=1 DEBUG_TCP=1
>>>
>>> We might get lucky and see something in the trace. I would also try to
>>> do some of the things Ming listed.
>> I did this but it does not seem to produce any more output!?
>> I checked the ttys and all files in /var/log
>> Where is the debug output supposed to be displayed?
>
> Okay, actually
> make DEBUG_SCSI=1 DEBUG_TCP=1
> and
> make DEBUG_SCSI=1 DEBUG_TCP=1 install (not just make install)
> was necessary.
> I attached the output in a compressed file.
>
> Here's what I did:
>
> linux:~ # uname -a
> Linux ACD-xen01 2.6.16.33-xen #1 SMP Tue Apr 3 10:29:14 CEST 2007 
i686 i686 i386 GNU/Linux
> linux:/mnt # df -h
> Dateisystem          Größe Benut  Verf Ben% Eingehängt auf
> /dev/sda3             9,9G  7,4G  2,0G  80% /
> udev                  257M  204K  256M   1% /dev
> /dev/sda4              22G  6,9G   14G  35% /home
> /dev/sdb1              20G  173M   19G   1% /mnt
> linux:/mnt # ls -la
> insgesamt 24
> drwxr-xr-x  3 root root  4096  3. Apr 11:36 ./
> drwxr-xr-x 22 root root  4096  3. Apr 10:44 ../
> drwx------  2 root root 16384 30. Mär 16:27 lost+found/
> linux:/mnt # logger dk teststart
> linux:/mnt # while :
>  > do
>  > dd if=/dev/zero of=largefile bs=1024k count=2000
>  > sleep 5
>  > logger dk loop done
>  > done
>
> The end of the log files is when the machine actually "stalled". It 
was not reachable via ssh, all ttys are dead, no keypress recognized. 
But still (as I wrote earlier), it was pingable.
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users