Thank you for following up on this.
I did your suggested method and it's working properly now. The engine
however still seems to want to add the 172.16.1.1 IP as a host to the
cluster, but that's not doing much other than building up the log file.
On Wed, Dec 11, 2013 at 11:23 PM, Sahina Bose <sabose(a)redhat.com> wrote:
On 12/10/2013 06:16 PM, Andrew Lau wrote:
Here you go:
http://www.fpaste.org/60464/38667947/
Andrew,
Thanks for the patience in providing the logs.
When a node is identified by multiple host names, the engine is not able
to resolve it correctly. The fix is for the engine to use gluster host
uuids while resolving brick info as well (as per the bug logged)
In the meantime, for you to proceed, one way would be to use the ip
address (172.16.1.1 instead of
gsx.melb.example.net) to peer probe the
gluster host .
On Tue, Dec 10, 2013 at 4:14 PM, Sahina Bose <sabose(a)redhat.com> wrote:
>
> On 12/10/2013 04:24 AM, Andrew Lau wrote:
>
> engine.log:
http://www.fpaste.org/60336/86629496/
> vdsm.log:
http://www.fpaste.org/60337/29624138/
>
>
> Andrew,
>
> I was interested in the getCapabilities output in vdsm.log. Could you
> click on "Refresh Caps" from Host tab and attach the output of
> getCapabilities from vdsm.log?
>
>
> thanks!
>
>
>
> On Mon, Dec 9, 2013 at 5:46 PM, Sahina Bose <sabose(a)redhat.com> wrote:
>
>>
>> On 12/08/2013 01:51 PM, Andrew Lau wrote:
>>
>> The defined cluster keeps showing me the option to import the
>>
gsx.melb.example.net hosts into the cluster as hosts.
>>
>> Under the Host's Network interfaces sub tab there is another issue
>> I've got was it's picking up the keepalived floating IP address rather
than
>> the one assigned, but yes the network is listed as I managed it through
>> ovirt.
>>
>>
>> Could you attach the engine.log as well as the vdsm.log from the node?
>>
>>
>>
>> On Fri, Dec 6, 2013 at 9:22 PM, Sahina Bose <sabose(a)redhat.com> wrote:
>>
>>>
>>> On 12/06/2013 03:29 PM, Sahina Bose wrote:
>>>
>>>
>>> On 12/06/2013 10:00 AM, Kanagaraj wrote:
>>>
>>>
>>> On 12/06/2013 09:46 AM, Andrew Lau wrote:
>>>
>>> Yup - it's ovirt cluster version 3.3 with gluster 3.4.1
>>>
>>>
>>> what does "gluster volume info <vol>" and "gluster
volume status <vol>"
>>> say?
>>>
>>>
>>> There's an issue where engine cannot sync the bricks, as bricks return
>>> a different IP address than the one engine is aware of (in this case
>>>
gsx.melb.example.net while engine knows the host as
>>>
hvx.melb.example.net) . We need to fix this path to use the gluster
>>> host UUID as well.
>>>
>>> Have logged a bug to track this -
>>>
https://bugzilla.redhat.com/show_bug.cgi?id=1038988
>>>
>>>
>>>
>>> Could you also let us know if the following network is listed under
>>> your Host's Network Interfaces sub tab?
>>> 172.16.1.1 (gluster)
gsx.melb.example.net
>>>
>>> If it is, then even without any fix, it should have worked for you and
>>> we need to dig further on why it did not.
>>>
>>>
>>>
>>>
>>>
>>>
>>> On Fri, Dec 6, 2013 at 3:11 PM, Kanagaraj <kmayilsa(a)redhat.com>
wrote:
>>>
>>>>
>>>> On 12/06/2013 07:55 AM, Andrew Lau wrote:
>>>>
>>>> Because of a few issues I had with keepalived, I moved my storage
>>>> network to it's own VLAN but it seems to have broken part of the
ovirt
>>>> gluster management.
>>>>
>>>> Same scenario:
>>>> 2 Hosts
>>>>
>>>> 1x Engine, VDSM, Gluster
>>>> 1x VDSM,Gluster
>>>>
>>>> So to properly split the gluster data and ovirtmgmt I simply
>>>> assigned them two host names and two IPS.
>>>>
>>>> 172.16.0.1 (ovirtmgmt)
hvx.melb.example.net
>>>> 172.16.1.1 (gluster)
gsx.melb.example.net
>>>>
>>>> However the oVirt engine does not seem to like this, it would not
>>>> pick up the gluster volume as "running" until I did a restart
through the
>>>> UI.
>>>>
>>>> The issue (possible bug) I'm seeing is the logs are being filled
>>>> with
http://www.fpaste.org/59440/13862963/
>>>>
>>>> Volume information isn't being pulled as it thinks the
>>>>
gs01.melb.example.net is not within the cluster, where in fact it is
>>>> but registered under
hv01.melb.example.net
>>>>
>>>>
>>>> What's compatibility version of the clusters?
>>>>
>>>> From 3.3 onwards, gluster-host-uuid is used to identify a host instead
>>>> of hostname.
>>>>
>>>> Thanks,
>>>> Kanagaraj
>>>>
>>>>
>>>> Thanks,
>>>> Andrew
>>>>
>>>>
>>>> _______________________________________________
>>>> Users mailing
listUsers@ovirt.orghttp://lists.ovirt.org/mailman/listinfo/users
>>>>
>>>>
>>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Users mailing
listUsers@ovirt.orghttp://lists.ovirt.org/mailman/listinfo/users
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Users mailing
listUsers@ovirt.orghttp://lists.ovirt.org/mailman/listinfo/users
>>>
>>>
>>>
>>> _______________________________________________
>>> Users mailing list
>>> Users(a)ovirt.org
>>>
http://lists.ovirt.org/mailman/listinfo/users
>>>
>>>
>>
>>
>
>