----- Original Message -----
> From: "Pascal Jakobi" <pascal.jakobi@gmail.com>
> To: "Federico Simoncelli" <fsimonce@redhat.com>
> Cc: users@ovirt.org

> Sent: Thursday, December 19, 2013 5:17:59 PM
> Subject: Re: [Users] AcquireHostId problem
>

> sanlock-2.8-1.fc19.x86_64
>
> Dec 18 16:46:05 lab2 systemd-wdmd[818]: Starting wdmd: [  OK  ]
> Dec 18 16:46:05 lab2 systemd[1]: Unit wdmd.service entered failed state.
> Dec 18 20:52:12 lab2 wdmd[887]: wdmd started S0 H1 G179
> Dec 18 20:52:12 lab2 wdmd[887]: /dev/watchdog0 failed to set timeout
> Dec 18 20:52:12 lab2 wdmd[887]: /dev/watchdog0 disarmed
> Dec 18 20:52:12 lab2 wdmd[887]: /dev/watchdog failed to set timeout
> Dec 18 20:52:12 lab2 wdmd[887]: /dev/watchdog disarmed
> Dec 18 20:52:12 lab2 wdmd[887]: no watchdog device, load a watchdog driver
> Dec 18 20:52:12 lab2 systemd[1]: wdmd.service: main process exited,
> code=exited, status=255/n/a
>
> I do not really understand all this. Furthermore, I could not find
> reference to it in the documentation. Am I missing sthing ? What does "load
> a wathdog driver" means in concrete ?

It seems that you have a watchdog device (probably provided by the
motherboard of your pc/server) that is not currently usable by wdmd.

The temporary workaround to get your system to work is to modprobe
the softdog kernel module and restart wdmd:

 # modprobe softdog
 # service wdmd restart

Depending on the distribution you're using there are different ways
to automatically load the module at boot, for example on fedora you
can use:

http://www.freedesktop.org/software/systemd/man/modules-load.d.html

and for centos/rhel 6 you can look at the content of:

/etc/sysconfig/modules

for an example of how to do the same (e.g. kvm.modules).

That said, in order to get your specific watchdog device to work
properly I'd need some additional information. Can you provide me
the output of these two commands?

 # lspci -vvv
 # lsmod

Thanks,
--
Federico