[ClusterLabs] fence_vbox Unable to connect/login to fencing device

ArekW arkaduis at gmail.com
Tue Jul 11 09:13:16 UTC 2017


Hi,
I may be wrong but it doesn't seem to be timeout problem because the log
repeats the same way every few minutes and it contains "Unable to connect"
and just after that there is list of vms etc so It has connected
successfully.
I described a active-active failover problem in separate mail. When a node
is poweroff the cluster enters UNCLEAN status and whole thing hungs. Could
it be related to stonith problem? I'm out of ideas what is wrong because I
seems to work manually but seems not to work as a fence process.
How can I increase the login_timeout (Is it for stonith?)
Thanks
Arek

2017-07-10 13:10 GMT+02:00 Marek Grac <mgrac at redhat.com>:

>
>
> On Fri, Jul 7, 2017 at 1:45 PM, ArekW <arkaduis at gmail.com> wrote:
>
>> The reason for --force is:
>> Error: missing required option(s): 'ipaddr, login, plug' for resource
>> type: stonith:fence_vbox (use --force to override)
>>
>
> It looks like you use unreleased upstream of fence agents without a
> similary new version of pcs (with the commit 7f85340b7aa4e8c016720012cf42c3
> 04e68dd1fe)
>
>
>>
>> I have selinux disabled on both nodes:
>> [root at nfsnode1 ~]# cat /etc/sysconfig/selinux
>> SELINUX=disabled
>>
>> pcs stonith update vbox-fencing verbose=true
>> Error: resource option(s): 'verbose', are not recognized for resource
>> type: 'stonith::fence_vbox' (use --force to override)
>>
>
> It shoulbe fixed in commit b47558331ba6615aa5720484301d644cc8e973fd (Jun
> 12)
>
>
>>
>>
>
>>
>> Jul  7 13:37:49 nfsnode1 fence_vbox: Unable to connect/login to fencing
>> device
>> Jul  7 13:37:49 nfsnode1 stonith-ng[2045]: warning: fence_vbox[4765]
>> stderr: [ Running command: /usr/bin/ssh -4  AW23321 at 10.0.2.2 -i
>> /root/.ssh/id_rsa -p 22 -t '/bin/bash -c "PS1=\\[EXPECT\\]#\  /bin/bash
>> --noprofile --norc"' ]
>>
>
> ok, so sometimes it works and sometimes not. It looks like that our
> timeouts are set quite strict for your environment. Try to increase
> login_timeout from default 30s higher.
>
> m,
>
> _______________________________________________
> Users mailing list: Users at clusterlabs.org
> http://lists.clusterlabs.org/mailman/listinfo/users
>
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.clusterlabs.org/pipermail/users/attachments/20170711/6f9807e4/attachment-0002.html>


More information about the Users mailing list