<div dir="ltr"><div dir="ltr"><br></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Mon, Sep 4, 2023 at 1:50 PM Andrei Borzenkov <<a href="mailto:arvidjaar@gmail.com">arvidjaar@gmail.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">On Mon, Sep 4, 2023 at 2:18 PM Klaus Wenninger <<a href="mailto:kwenning@redhat.com" target="_blank">kwenning@redhat.com</a>> wrote:<br>
><br>
><br>
><br>
> On Mon, Sep 4, 2023 at 12:45 PM David Dolan <<a href="mailto:daithidolan@gmail.com" target="_blank">daithidolan@gmail.com</a>> wrote:<br>
>><br>
>> Hi Klaus,<br>
>><br>
>> With default quorum options I've performed the following on my 3 node cluster<br>
>><br>
>> Bring down cluster services on one node - the running services migrate to another node<br>
>> Wait 3 minutes<br>
>> Bring down cluster services on one of the two remaining nodes - the surviving node in the cluster is then fenced<br>
>><br>
>> Instead of the surviving node being fenced, I hoped that the services would migrate and run on that remaining node.<br>
>><br>
>> Just looking for confirmation that my understanding is ok and if I'm missing something?<br>
><br>
><br>
> As said I've never used it ...<br>
> Well when down to 2 nodes LMS per definition is getting into trouble as after another<br>
> outage any of them is gonna be alone. In case of an ordered shutdown this could<br>
> possibly be circumvented though. So I guess your fist attempt to enable auto-tie-breaker<br>
> was the right idea. Like this you will have further service at least on one of the nodes.<br>
> So I guess what you were seeing is the right - and unfortunately only possible - behavior.<br>
<br>
I still do not see where fencing comes from. Pacemaker requests<br>
fencing of the missing nodes. It also may request self-fencing, but<br>
not in the default settings. It is rather hard to tell what happens<br>
without logs from the last remaining node.<br>
<br>
That said, the default action is to stop all resources, so the end<br>
result is not very different :)<br></blockquote><div><br></div><div>But you are of course right. The expected behaviour would be that</div><div>the leftover node stops the resources.</div><div>But maybe we're missing something here. Hard to tell without</div><div>the exact configuration including fencing.</div><div>Again, as already said, I don't know anything about the LMS</div><div>implementation with corosync. In theory there were both arguments</div><div>to either suicide (but that would have to be done by pacemaker) or</div><div>to automatically switch to some 2-node-mode once the remaining</div><div>partition is reduced to just 2 followed by a fence-race (when done</div><div>without the precautions otherwise used for 2-node-clusters).</div><div>But I guess in this case it is none of those 2.</div><div><br></div><div>Klaus </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
_______________________________________________<br>
Manage your subscription:<br>
<a href="https://lists.clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">https://lists.clusterlabs.org/mailman/listinfo/users</a><br>
<br>
ClusterLabs home: <a href="https://www.clusterlabs.org/" rel="noreferrer" target="_blank">https://www.clusterlabs.org/</a><br>
</blockquote></div></div>