[CentOS] Xen VMs: regular filesystem corruption

Kai Schaetzl maillists at conactive.com
Fri Oct 12 12:31:59 UTC 2007


It seems that Xen VMs are subject to file system corruption under certain 
conditions. I find that often when I reboot or shutdown a VM the 
filesystem is corrupted after that and the VM no longer usable. The xm 
create then fails with some Python errors and an eventual 

Runtime Error: unable to read filesystem
No handlers could be found for logger "xend".
Error: boot loader didn't return any data.

and I have to throw it away as there doesn't seem a way to fix that (or is 
there?).
I use files for storing the VMs. There is enough space on the drive. The 
problem seems to happen when I reboot the machine from within itself, not 
each time, but almost each time. When I shut it down with CTRL+ALT+DEL or 
with the Shutdown button it doesn't seem to happen.
I also find that sometimes a good VM will start up and crash early with a 
kernel panic, but start up just fine when I try a second or third time. 
Most often this happened after a new installation on the first reboot.

Anyone else seen this or known how to fix that?

Kai

-- 
Kai Schätzl, Berlin, Germany
Get your web at Conactive Internet Services: http://www.conactive.com






More information about the CentOS mailing list