[Xen-users] domU rebooting spontaneously

[Tim Cutts <tjrc@xxxxxxxxxxxx>]

Please bear with me, I'm fairly new to Xen.  Before I describe the  
issue, here's the hardware/software of the system.
Hardware, RLX 800i blade server (800 MHz Pentium III, 1GB RAM)

Xen version:  2.0.7
dom0 OS:      Debian Sarge, given 400 M of RAM, root filesystem on / 
dev/hda
domU OS:      Debian Etch, given 400 M of RAM, root filesystem a disk  
image, also on /dev/hda, and also
	      has another image presented as /dev/sdb1, and /dev/hdc1  
presented as /dev/sdb2
/dev/sdb2 is mounted in the domU on /data, and into this I have  
loaded a benchmark suite that we usually use for evaluating hardware  
for benchmarking various bioinformatics codes.
I have run this benchmark suite successfully under a regular Debian  
Sarge kernel on the machine, and again successfully under dom0, but  
as soon as I run it in the domU VM, the VM reboots.  It seems to do  
so cleanly; I can't find any error messages - it just behaves as  
though someone typed 'reboot' in it.
The domU console doesn't show any errors, and the xend log reports  
the following:
[2006-01-12 13:27:38 xend] INFO (console:44) Console connected 10  
127.0.0.1 3424
[2006-01-12 13:27:38 xend] INFO (XendRoot:113) EVENT>  
xend.console.connect [10, '127.0.0.1', 3424]
At this point, I start the benchmark script at the domU's console.

[2006-01-12 13:31:50 xend] DEBUG (XendDomain:244) XendDomain>reap>  
domain died name=isg-vm-1 id=1
[2006-01-12 13:31:50 xend] DEBUG (XendDomain:247) XendDomain>reap>  
shutdown id=1 reason=reboot
[2006-01-12 13:31:50 xend] DEBUG (XendDomain:487)  
domain_restart_schedule> 1 reboot 0
[2006-01-12 13:31:50 xend] INFO (XendDomain:501) Scheduling restart  
for domain: name=isg-vm-1 id=1
[2006-01-12 13:31:50 xend] INFO (XendDomain:568) Destroying domain:  
name=isg-vm-1
[2006-01-12 13:31:50 xend] DEBUG (XendDomainInfo:665) Destroying vifs  
for domain 1
[2006-01-12 13:31:50 xend] DEBUG (netif:305) Destroying vif domain=1  
vif=0
[2006-01-12 13:31:50 xend] DEBUG (XendDomainInfo:674) Destroying vbds  
for domain 1
[2006-01-12 13:31:50 xend] DEBUG (blkif:552) Destroying blkif domain=1
[2006-01-12 13:31:50 xend] DEBUG (blkif:408) Destroying vbd domain=1  
idx=0
[2006-01-12 13:31:50 xend] DEBUG (blkif:408) Destroying vbd domain=1  
idx=1
[2006-01-12 13:31:50 xend] DEBUG (blkif:408) Destroying vbd domain=1  
idx=2
[2006-01-12 13:31:50 xend] DEBUG (blkif:408) Destroying vbd domain=1  
idx=3
[2006-01-12 13:31:50 xend] DEBUG (XendDomainInfo:622) Closing channel  
to domain 1
[2006-01-12 13:31:50 xend] INFO (XendRoot:113) EVENT> xend.virq 4
[2006-01-12 13:31:50 xend] DEBUG (blkif:363) Unbinding vbd (type  
file) from /dev/loop0
[2006-01-12 13:31:50 xend] DEBUG (blkif:363) Unbinding vbd (type  
file) from /dev/loop1
[2006-01-12 13:31:50 xend] DEBUG (blkif:363) Unbinding vbd (type  
file) from /dev/loop2
[2006-01-12 13:31:50 xend] DEBUG (blkif:363) Unbinding vbd (type phy)  
from hdc1
[2006-01-12 13:31:50 xend] INFO (XendRoot:113) EVENT>  
xend.domain.exit ['isg-vm-1', '1', 'reboot']
One thing that triggers this behaviour instantly is if I interrupt  
the script.  For example, if I ssh into the virual machine, start the  
benchmark, and then press Ctrl-C, this is what I see:
isg-vm-1:/data/benchmark# ./run_tests.pl
./run_tests.pl on isg-vm-1
interrupted.

Running tasks:

     1326 ARRAY(0x823e420)

isg-vm-1:/mnt/benchmark#
Broadcast message from root (console) (Thu Jan 12 16:10:22 2006):

The system is going down for reboot NOW!
Connection to isg-vm-1 closed by remote host.
Connection to isg-vm-1 closed.

So it looks like maybe something very strange may be going on with  
signals...
Tim


--
Dr Tim Cutts
Informatics Systems Group, Wellcome Trust Sanger Institute
GPG: 1024D/E3134233 FE3D 6C73 BBD6 726A A3F5  860B 3CDD 3F56 E313 4233


_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users