[Users] host remains in status unassigned

[Roy Golan]

On Sun 14 Jul 2013 08:12:21 PM IDT, Gianluca Cecchi wrote:
>
> Il giorno 14/lug/2013 17:35, "Joop" <jvdwege at xs4all.nl
> <mailto:jvdwege at xs4all.nl>> ha scritto:
> >
> > Gianluca Cecchi wrote:
> >>
> >> Hello,
> >>
> >> oVirt 3.2.1 on F18
> >> One mgmt server + one host.
> >>
> >> At some point the host had some problem and was rebooted
> >> at reboot ther was a problem because I got
> >>
> >> detected unhandled Python exception in '/usr/share/vdsm/vdsm'
> >>
> >> watching at
> >> http://www.ovirt.org/Troubleshooting
> >>
> >> I verified that for some reason vdsmd.log was root.root.
> >> So I set correct permissions and restarted the node.
> >>
> >>
> >
> > OK, I have had this too and managed to get out of it uptil now,
> maybe my experience will help you, may you have already tried what I'm
> going to suggest.
> > Put the host in maintenance, login with ssh on the host and restart
> the vdsmd service, activate the host.
> >
> > I think I amended vdsms logrotate.conf with a user statement so that
> it wouldn't make root owner of vdsmd.log. There is a race condition
> that has been fixed I think in 3.2.2 and onwards.
> >
> > Hope this helps you,
> >
> > Joop
> >
> I'm going to check but it seems to me that I could only select "host
> has rebooted" that failed because "unassigned" is not a good starting
> point to set it.
> Not sure I could put in maintenance. ..
> Gianluca
>
Unfortunatly you can't. "unassigned" puts the host in transition and 
its status should be transited automatically by the engine. Its a bug 
that it didn't
move to any other state - Non-operational/up/non-responsive etc.

Can you help me with a stacktrace of your engine with the output of 
"jstack -l <jboss-pid>?



>
>
> _______________________________________________
> Users mailing list
> Users at ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users