[DRBD-user] Problems using image files with drbd, maybe unsupported?

Bart Coninckx bart.coninckx at telenet.be
Wed Oct 6 14:39:52 CEST 2010


On Wednesday 06 October 2010 14:09:44 Martin Fandel wrote:
> Thats exactly the point of my question :). How can I make the system stable
> for unplanned failovers? Power loss for example? In this case, HA has no
> chance for cleaning up the resources. :/
> 
> -------- Original-Nachricht --------
> 
> > Datum: Wed, 6 Oct 2010 11:30:19 +0200
> > Von: Bart Coninckx <bart.coninckx at telenet.be>
> > An: "Martin Fandel" <pingu.freak at gmx.de>
> > CC: drbd-user at lists.linbit.com
> > Betreff: Re: [DRBD-user] Problems using image files with drbd,	maybe
> > unsupported?
> > 
> > On Wednesday 06 October 2010 10:51:54 Martin Fandel wrote:
> > > Hi,
> > > 
> > > hearbeat does the same as I do in the following steps:
> > > 
> > > drbd disconnect all
> > > drbd primary all
> > > mount /dev/drbd0 /var/lib/xen/images
> > > 
> > > However, in a heartbeat environment, the xen images are also damaged.
> > > If the primary node fails, all xen images located in
> > > /var/lib/xen/images
> > 
> > are
> > 
> > > inconsistent and not bootable.
> > > 
> > > I can avoid this by using fsck on the images. But I don't know how use
> > 
> > fsck
> > 
> > > on image files. Does anybody know this?
> > > 
> > > Thanks!
> > > 
> > > Regards,
> > > 
> > > Martin
> > > 
> > > -------- Original-Nachricht --------
> > > 
> > > > Datum: Tue, 5 Oct 2010 21:59:06 +0200
> > > > Von: Bart Coninckx <bart.coninckx at telenet.be>
> > > > An: drbd-user at lists.linbit.com
> > > > CC: "Martin Fandel" <pingu.freak at gmx.de>
> > > > Betreff: Re: [DRBD-user] Problems using image files with drbd, maybe
> > > > unsupported?
> > > > 
> > > > On Tuesday 05 October 2010 12:26:08 Martin Fandel wrote:
> > > > > Hi,
> > > > > 
> > > > > I'm not sure if I understand the documentation correct. I've
> > 
> > installed
> > 
> > > > > SLES10 with DRBD and Xen. I added an drbd-device /dev/drbd0 for the
> > 
> > xen
> > 
> > > > > image files. The drbd device /dev/drbd0 is created with ext3 and is
> > > > > mounted to /var/lib/xen/images. Under this directory, 10 vms are
> > > > > installed. If I do an switch to the backup machine, xen can't start
> > 
> > the
> > 
> > > > > image files / the guest. The kernel says:
> > > > > 
> > > > > [87585.519596] EXT3-fs: INFO: recovery required on readonly
> > 
> > filesystem.
> > 
> > > > > [87585.519602] EXT3-fs: write access unavailable, cannot proceed.
> > > > > 
> > > > > Is it possible, that image files with drbd are not supported? I've
> > 
> > also
> > 
> > > > > found an tutorial on google. In the tutorial is a notice, that the
> > > > > image files must be dumped to dedicated drbd partitions?
> > > > > 
> > > > > Thanks!
> > > > > 
> > > > > Regards,
> > > > > 
> > > > > Martin
> > > > 
> > > > You need to install the heartbeat rpms for SLES10 and configure DRBD
> > > > resources
> > > > so the DRBD resource becomes active after failover.
> > > > 
> > > > B.
> > 
> > I probably should have mentionned that a resource for a ext3 volume
> > should also be made, so that with planned failovers you get a clean
> > result. However,
> > this is not the best thing to happen for running Xen guests, because they
> > will
> > sense their storage has just been swept from under their feet. So you
> > also need to add Xen resources. Personally with SLES 10 I would run
> > guests on iSCSI
> > storage hosted by DRBD and Heartbeat.
> > 
> > B.
> > 
> > _______________________________________________
> > drbd-user mailing list
> > drbd-user at lists.linbit.com
> > http://lists.linbit.com/mailman/listinfo/drbd-user

Also referring to Lars' answer: it should be possible to have in the largest 
cases of unexpected failovers a functioning ext3 volume. So something goes 
wrong there. If you are using a CRM based config it's useful to have more 
config details. 

Rgds,

B.


More information about the drbd-user mailing list