[ClusterLabs] 2-Node Cluster - fencing with just one node running ?

[Reid Wahl]

On Thu, Aug 4, 2022 at 7:40 AM Lentes, Bernd
<bernd.lentes at helmholtz-muenchen.de> wrote:
>
>
> ----- On 4 Aug, 2022, at 15:14, arvidjaar arvidjaar at gmail.com wrote:
>
> > On 04.08.2022 16:06, Lentes, Bernd wrote:
> >>
> >> ----- On 4 Aug, 2022, at 00:27, Reid Wahl nwahl at redhat.com wrote:
> >>
> >> Would do you mean by "banned" ? "crm resource ban ..." ?
> >> Is that something different than a location constraint ?
> > "crm resource ban" creates location constraint, but not every location
> > constraint is created by "crm resource ban".
>
> OK.
>
> It seems that the cluster realizes that something went wrong.
> It wants to shutdown ha-idg-1:
> Aug 03 01:19:12 [19367] ha-idg-1    pengine:  warning: pe_fence_node:   Cluster node ha-idg-1 will be fenced: vm-mausdb failed there
> Aug 03 01:19:12 [19367] ha-idg-1    pengine:     info: native_stop_constraints: fence_ilo_ha-idg-2_stop_0 is implicit after ha-idg-1 is fenced
> Aug 03 01:19:12 [19367] ha-idg-1    pengine:   notice: LogNodeActions:   * Fence (Off) ha-idg-1 'vm-mausdb failed there'
> Aug 03 01:19:14 [19367] ha-idg-1    pengine:  warning: pe_fence_node:   Cluster node ha-idg-1 will be fenced: vm-mausdb failed there
> Aug 03 01:19:15 [19368] ha-idg-1       crmd:   notice: te_fence_node:   Requesting fencing (Off) of node ha-idg-1 | action=8 timeout=60000
>  ...
>
> It shuts down ha-idg-2:
> 2022-08-03T01:19:51.866200+02:00 ha-idg-2 systemd-logind[1535]: Power key pressed.
> 2022-08-03T01:19:52.048335+02:00 ha-idg-2 systemd-logind[1535]: System is powering down.
> 2022-08-03T01:19:52.051815+02:00 ha-idg-2 systemd[1]: Stopped target resource-agents dependencies.
>  ...

Yes, but it thought it was shutting down ha-idg-1.

>
> Then it stops cluster software on ha-idg-1:
> Aug 03 01:19:58 [19361] ha-idg-1 pacemakerd:  warning: pcmk_child_exit: Shutting cluster down because crmd[19368] had fatal failure
> Aug 03 01:19:58 [19361] ha-idg-1 pacemakerd:   notice: pcmk_shutdown_worker:    Shutting down Pacemaker
> Aug 03 01:19:58 [19361] ha-idg-1 pacemakerd:   notice: stop_child:      Stopping pengine | sent signal 15 to process 19367
>  ...

Node ha-idg-1 received a notification from the fencer that said "hey,
we just fenced ha-idg-1!" Then it said "oh no, that's me! I'll shut
myself down now."

That can be helpful if we're using fabric fencing. That's not supposed
to happen with power fencing. The shutdown on ha-idg-1 didn't hurt
anything, but it should have gotten powered off (instead of powering
off ha-idg-2.

>
> Bernd
> _______________________________________________
> Manage your subscription:
> https://lists.clusterlabs.org/mailman/listinfo/users
>
> ClusterLabs home: https://www.clusterlabs.org/



-- 
Regards,

Reid Wahl (He/Him)
Senior Software Engineer, Red Hat
RHEL High Availability - Pacemaker