[Pacemaker] moving from corosync.conf to cluster.conf

Parveen Jain parveenjain at live.com
Tue Feb 25 08:56:56 UTC 2014


Hi Andrew, Thanks for responding so quickly for this.Actually I lost all of my logs as I re Installed my machine for trying some other combination.
Is it possible that I just upgrade the O/S to RHEL6.5 and its clustering software.And without moving to maintenance mode I try to reconfigure my instances using "pcs" commands ?

From: andrew at beekhof.net
Date: Mon, 24 Feb 2014 20:54:51 +1100
To: pacemaker at oss.clusterlabs.org
Subject: Re: [Pacemaker] moving from corosync.conf to cluster.conf

 
On 24 Feb 2014, at 7:18 pm, Parveen Jain <parveenjain at live.com> wrote:
 
> Hi All,
>   Following was my problem:
> 1)      My RHEL 6.3 is using CRM shell and corosync.conf.
> 2)      Wanted to move to RHEL6.5 and hence the underlying cluster. Also wanted to move to recommended way of using cluster with CMAN using “cluster.conf”.
> 
> (Before executing this document I first upgraded my O/S to RHEL6.5.)
> To solve this problem , found following link:
> http://floriancrouzat.net/2013/04/rhel-6-4-pacemaker-1-1-8-adding-cman-support-and-getting-rid-of-the-plugin/
> (Also got another link which has more or less same thing explain in a documented way:
> http://clusterlabs.org/doc/en-US/Pacemaker/1.1-plugin/html/Clusters_from_Scratch/_configuring_cman.html)
> 
> This document is written well and I could execute all of the commands successfully without disturbing my cluster, but I found some issue after executing this last line:
> crm configure property maintenance-mode=false
> After executing this command, I lost my VIP.
 
That is unexpected.
 
> I need to execute this upgrade on my production environment and can’t afford to go “offline” in this way.
 
Sure. Can you send us a crm_report that starts just before you set maintenance-mode=false and ends after the VIP was lost?
(No need to reproduce, just run give it the times from when you already had it happen.)
 
> 
> Few are my questions to execute this upgrade in a more convenient way:
> 1)      What can be the reason of this VIP removal (if someone has already tested with this approach)?
> 2)      Is it right approach, if I do directly assign my VIP (after moving my cluster to maintenance mode) to one of my Ethernet interface and then run the commands given in this document.
> 3)      Is not there some command line option to upgrade from corosync.conf to cluster.conf directly(ideally I felt that we should be able to generate an ).
> 4)      How will I move to the usage of “pcs commands”. This article is given with using “ccs” commands. Shall I assume that once I finish with “ccs”, I can do rest of my configuration using “pcs” commands.
 
Right, once cman is running, everything is done with pcs.
It sounds like you're used to crmsh, there is a conversion document available at:
 
   https://github.com/ClusterLabs/pacemaker/blob/master/doc/pcs-crmsh-quick-ref.md
 
> 
> Regards,
> Parveen Jain
> _______________________________________________
> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
> 
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org
 

_______________________________________________
Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
http://oss.clusterlabs.org/mailman/listinfo/pacemaker

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: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20140225/435db982/attachment.htm>


More information about the Pacemaker mailing list