I saw that, but my entire environment is running on a single host (Even the data domain is on the same host).  How would it be possible for the host to basically stop talking with itself?  Isn't that a little peculiar?

On Thu, Aug 24, 2017 at 4:09 PM, Douglas Landgraf <dlandgra@redhat.com> wrote:

On Thu, Aug 24, 2017 at 3:32 PM, Wesley Stewart <wstewart3@gmail.com> wrote:
> I have noticed this a couple times now.  digging through the logs, it looks
> like the host decided to become unresponsive:
>
> 2017-08-24 12:09:05,365-04 INFO
> [org.ovirt.engine.core.bll.VdsEventListener]
> (org.ovirt.thread.pool-6-thread-5) [56defee2]
> ResourceManager::vdsNotResponding entered for Host
> 'a1bf54c1-2890-4aae-a23d-c83ea2c664d2', 'Host IP'


Looks like the network is down in the host:

Failed to refresh VDS, network error, continuing,
vds='OVIRT-Host'(a1bf54c1-2890-4aae-a23d-c83ea2c664d2):
java.net.SocketException: Network is unreachable



>
> Caused by: java.net.ConnectException: Connection refused
>
> There is a lot more, but I thought dropping it into a pastebin would be
> better:
> https://pastebin.com/drPNnsGB
>
> It looks like the host stopped responding somehow but I can't quite tell
> how.  Any advice or input would be greatly appreciated!
>


Few questions:

Are you able ping/communicate from host to engine and vice-versa?
In the node, do you have the ovirtmgmt interface up?
what's the output of systemctl status vdsmd -l  ?
Did you change the firewall settings?
Which version are you running? Did you upgrade?

In the hypervisor, the log /var/log/vdsm/vdsm.log might help us too.

Finally, check in the hypervisor if your vms are really down, ps aux |
grep qemu-kvm  should return your vms process


--
Cheers
Douglas

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