Re: [Xen-API] vhd_complete errors and disk corruption

[Denis Cardon <denis.cardon@xxxxxxxxxxxxxxxxxxxxxx>]

Hi Chris,

> Has anyone had any thoughts on this, or can anyone suggest a better
> place for me to ask such questions?
I've had a few vhd chain corruption with xcp 1.0 and xcp 1.1. The last 
one I had happened at the same time as a oomkiller in the dom0. There 
has been a patch in the blktap libvhd 
(https://github.com/xen-org/blktap/commit/62de80d899f0eb83bf92de94f9fab310c4a7231d) 
that fix a memory leak in the coalescing process.
If you can find a oomkiller  message in your logs, you may have been hit 
by the same coalescing bug.  The only option I found is to try to avoid 
snapshot as much as possible. I guess the patch is included in the xcp 
1.5 release.
Cheers,

Denis


> Thanks,
> Chris
>
> On Tue, Apr 24, 2012 at 3:54 PM, Chris Percol<chris.percol@xxxxxxxxx>  wrote:
> > Hi,
> > We're encountering the following error intermittently on XCP 1.0.  It
> > seems to be restricted to particular VMs.  We've had 2 out of around
> > 30 affected by this issue, which seems to end up causing disk
> > corruption.
> >
> >
> > Apr 24 12:55:25 x5 tapdisk[5051]: ERROR: errno -5 at vhd_complete:
> > /dev/mapper/VG_XenStorage--86d18095--7819--25b5--e4d4--fff89212bd56-VHD--248ac421--1683--4ab2--b5c6--49fb62849cb6:
> > op: 3, lsec: 15908864, secs: 1, nbytes: 512, blk: 3884, blk_offset:
> > 1576183
> >
> > I'd appreciate any suggestions on how to debug this further.
> >
> > Thanks,
> > Chris
> _______________________________________________
> xen-api mailing list
> xen-api@xxxxxxxxxxxxx
> http://lists.xen.org/cgi-bin/mailman/listinfo/xen-api

--
Denis Cardon
Tranquil IT Systems
44 bvd des pas enchantés
44230 Saint Sébastien sur Loire
tel : +33 (0) 2.40.97.57.57
http://www.tranquil-it-systems.fr


_______________________________________________
Xen-api mailing list
Xen-api@xxxxxxxxxxxxx
http://lists.xen.org/cgi-bin/mailman/listinfo/xen-api