Hi,
I had to restart everything again yesterday and was confronted with the whole
iscsi issue again (ovirt can't find the volumes when you try to start a VM).
I've now figured out what the problem is, but have no idea how to fix it. I
took a look on the node and it is able to mount the volumes without problems.
The issue seems to be that ovirt is trying to find the volume by id once
it's mounted, but the id changes between reboots as it is automagically
generated by udev. The error that appears is:
Unable to find volume
/dev/disk/by-id/scsi-149455400000000000000000005000000061080000000d00f attached
to pool 192.168.50.190-iqn.2009-3.san0:services-3260.
And yes indeed that id doesn't exist it's actually in the form
/dev/disk/by-id/scsi-14945540000000000000000000?0000000??080000000d00f
or something like that. So the node mounts the volume fine but the VM can't
see it because it doesn't know which VM to use.
Has this been fixed in a later release (I'm running next from about a month
ago) or is this an ongoing problem awaiting a solution? Also, has the
provisioning from a cobbler ISO image on NFS been sorted out? Trying to work
out if I should move to the latest next.
Cheers,
Justin.
