[Users] NFS Domains down because of single node failure

[Markus Stockhausen]

> Von: Itamar Heim [iheim at redhat.com]
> Gesendet: Mittwoch, 18. September 2013 14:04
> An: Markus Stockhausen
> Cc: users; Allon Mureinik
> Betreff: Re: [Users] NFS Domains down because of single node failure
> 
> On 09/17/2013 10:01 AM, Markus Stockhausen wrote:
> > Hello,
> >
> > maybe a stupid one but ...
> >
> > When I create a (NFS) storage domain I have to provide a node host that
> > makes
> > the inital contact. All other node hosts wil directly connect to that
> > domain. So
> > no bottlenecks.
> >
> > Today I stopped one of my two nodes outside ovirt-engine. For simplicity we
> > assume the node crashed. The machine is up right now but VDSM is down.
> >
> > All domains that where setup with this host are "down" now (red arrow
> > down).
> 
> the node you used to setup the storage isn't affecting later runs.
> can the remaining node access the storage?

These drive-by errors are really hard ones. Remembering what you did
before the error and all the time your are close to restarting all cluster 
components...

Nevertheless I provided a reproduceble test case. More in
https://bugzilla.redhat.com/show_bug.cgi?id=1009610

Markus
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: InterScan_Disclaimer.txt
URL: <http://lists.ovirt.org/pipermail/users/attachments/20130918/65aacb03/attachment-0001.txt>