[ClusterLabs] 32 nodes pacemaker cluster setup issue

[Klaus Wenninger]

On 5/19/21 1:45 PM, Klaus Wenninger wrote:
> On 5/19/21 12:54 PM, S Sathish S wrote:
>>
>> Hi Klaus,
>>
>> pacemaker/corosync we generated our own build from clusterlab source 
>> code.
>>
>> [root at node1 ~]# cat /etc/redhat-release
>>
>> Red Hat Enterprise Linux Server release 7.4 (Maipo)
>>
>> [root at node1 ~]# uname -r
>>
>> 3.10.0-693.82.1.el7.x86_64
>>
>> [root at node1 ~]# rpm -qa | grep -iE 'pacemaker|corosync|pcs'
>>
>> pcs-0.9.169-1.el7.x86_64
>>
>> pacemaker-2.0.2-2.el7.x86_64
>>
>> corosync-2.4.4-2.el7.x86_64
>>
>>
> hmm ... corosync-2.4.4 should have corosync_move_to_root_cgroup
> and it should be set to do that per default which should prevent
> the issues I'm aware of.
> Do you start with '-R' option that would disable the behavior?
> Looking to the code it could depend on the version of libqb as well
> or some other build issue. But if you are building from source you
> can check what is happening in
> corosync_move_to_root_cgroup & corosync_set_rr_scheduler
> in main.c of corosync.
You might be missing the selinux-policy required to set the priority
as well. Easy check would be if disabling selinux temporarily
removes the log-entry.
>
> Wouldn't necessarily solve your issues btw. just something
> I stumbled over ...
>
> Klaus
>>
>> Thanks and Regards,
>>
>> S Sathish S
>>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/users/attachments/20210519/1155dd79/attachment.htm>