Nagios XI upgrade in cluster environment

[fpernet]

Hi All,

I have a NagiosXI in a DRBD+Heartbeat+Pacemaker cluster configuration (one active and one passive node). NagiosXI is 5.3.2 and two nodes are VirtualApplicances (OVF with CentOS).
I would like to know if any special operations must be performed in this environment, in order to update my NagiosXI (5.3.2 -> 5.3.3).
Of course all parts of Nagios (/usr/local/nagiosxi, /usr/local/nagios, ..., postgresql, mysql) is moved from one node to the other, so i am not very anxious about that.
But i think that sometimes, during the installation, some core components (CentoS) are also certainly updated and i MUST do that on both nodes.
I guess that the best approach could be:
- update all components with yum on both nodes
- update NagiosXI on one node, then switch NagiosXI to the other node, then redo the update script ...

Is there someone with experience in such case or does one of you have any idea of some special steps ?

Many thanks in advance for your bright ideas

Francois

[tgriep]

Sometime when doing a yum upgrade, it downgrades some of the required components that XI uses so I suggest to run the Manual Upgrade on both systems by following the procedure in the link below.
https://assets.nagios.com/downloads/nag ... nstall.pdf
If you have any more questions about this, let us know.

[fpernet]

Thx for the answer ... Yes of course, i meant manual upgrade ...
Really no other considerations before to try ?? No one in this forum, is in this case (active / passive cluster) ?
Ok. I'll keep you inform when i'll do the upgrade

[rkennedy]

It's hard to say, as our standard install doesn't take heartbeat / DRBD into account. We generally point these sort of solutions to our partner, Linbit - https://www.nagios.com/news/2015/10/pre ... nagios-xi/ - they may have more information available for you.

[fpernet]

Ok, i understand and I can ask LINBIT for sure.
But at the NagiosXI point of vue, is there any problem in running twice the same manual upgrade ?
In this case, it is supposed to perform everything the first time (on the first node), then only do the differential the second time (second node) ? don't you think ?
And at the databases (postgresql, mysql) point of view, same thing ?
Many thanks for your guidelines...
Francois

[eloyd]

I think what @rkennedy is saying is that since Nagios does not currently provide or support a clustered Nagios solution, that Nagios cannot answer your question with 100% accuracy.

In general, however, we have not had a problem doing this for our customers that use a similar approach. But again, this is not a Nagios-supported solution at this time.

[avandemore]

This is our HA documentation:

https://assets.nagios.com/downloads/nag ... ios-XI.pdf

Anything beyond that is left as an exercise for the user. HA options are so many and so specialized we cannot unfortunately support them directly. However in large part most of our documentation still applies depending on your configuration.