[Pacemaker] Don't want to stop lsb resource on migration

Andrew Beekhof andrew at beekhof.net
Fri Apr 4 02:14:23 UTC 2014


Why not just use colocation?

On 25 Mar 2014, at 6:14 am, Bingham <knee-jerk-reaction at hotmail.com> wrote:

> Greetings,
> 
> I am getting closer to what I need. 
> 
> I am having trouble figuring out a rule for p_VIPeth1_1 that will cause it's score to be -inf if p_R_NODE1 is not running.
> Here is the rule that I am struggling with. 
> 
> location p_VIPeth1_1_loc2 p_VIPeth1_1 \
> 	rule $id="p_VIPeth1_1_loc2-rule" -inf: p_R_NODE1 eq Stopped
> 
> 
> \/ The entire config so far is below \/
> 
> root at clust1:~# crm configure show
> node $id="404ea40c-f92d-4649-869d-41beaf261d87" clust1 \
> 	attributes standby="off"
> node $id="ae33be72-ccbb-4f54-859f-fd400efeb60b" clust2 \
> 	attributes standby="off"
> primitive p_PINGDB ocf:pacemaker:ping \
> 	params host_list="192.168.254.42" name="p_PINGDB" \
> 	op monitor interval="15s" timeout="5s"
> primitive p_R_NODE1 lsb:rabbitmq-server \
> 	op monitor interval="15s" timeout="15s" \
> 	meta target-role="Started" is-managed="true"
> primitive p_R_NODE2 lsb:rabbitmq-server \
> 	op monitor interval="15s" timeout="15s" \
> 	meta target-role="Started"
> primitive p_VIPeth1_1 ocf:heartbeat:IPaddr \
> 	params ip="192.168.254.78" cidr_netmask="255.255.255.0" nic="eth1" \
> 	op monitor interval="40s" timeout="20s" \
> 	meta target-role="Started" is-managed="true"
> clone p_PINGDB_clone p_PINGDB
> location cli-standby-p_R_NODE1 p_R_NODE1 \
> 	rule $id="cli-standby-rule-p_R_NODE1" -inf: #uname eq clust2
> location p_R_NODE1_loc p_R_NODE1 -inf: clust2
> location p_R_NODE1_pref p_R_NODE1 inf: clust1
> location p_R_NODE2_loc p_R_NODE2 -inf: clust1
> location p_R_NODE2_pref p_R_NODE2 inf: clust2
> location p_VIPeth1_1_loc p_VIPeth1_1 \
> 	rule $id="p_VIPeth1_1_loc-rule" -inf: p_PINGDB lte 0
> location p_VIPeth1_1_loc2 p_VIPeth1_1 \
> 	rule $id="p_VIPeth1_1_loc2-rule" -inf: p_R_NODE1 eq Stopped
> property $id="cib-bootstrap-options" \
> 	dc-version="1.1.6-9971ebba4494012a93c03b40a2c58ec0eb60f50c" \
> 	cluster-infrastructure="Heartbeat" \
> 	stonith-enabled="false" \
> 	default-action-timeout="120" \
> 	last-lrm-refresh="1395671783"
> root at clust1:~# 
> 
> 
> Thanks, 
> Steve
> 
> 
> 
> From: andrew at beekhof.net
> Date: Wed, 19 Mar 2014 09:37:43 +1100
> To: pacemaker at oss.clusterlabs.org
> Subject: Re: [Pacemaker] Don't want to stop lsb resource on migration
> 
>  
> On 19 Mar 2014, at 6:56 am, Bingham <knee-jerk-reaction at hotmail.com> wrote:
>  
> > 
> > My problem is that I need to have rabbitmq running on both node1 and node2.
> > I also need the IP to fail over if rabbitmq were to fail on the current node.
> > 
> > The 2 rabbitmq services are communicating with each other. 
> > Data is pushed to the clients. 
> > 
> > Even though the IP may currently live on node1, data may flow through node1 then through node2 (via rabbit) and out to client.
> > 
> >                        Rnode1 -------> client1
> >                      /     /|\
> > DB---->VIP         |
> >                            \|/
> >                       Rnode2 --------> client2
> > 
> > 
> > 
> > Maybe I should not have these resources grouped together since that implies collocation infinity for IP and rabbitmq?
>  
> Correct. It also sounds like rabbitmq should be a master/slave resource
>  
> > 
> > 
> > Steve
> > 
> > 
> > From: andrew at beekhof.net
> > Date: Tue, 18 Mar 2014 11:44:34 +1100
> > To: pacemaker at oss.clusterlabs.org
> > Subject: Re: [Pacemaker] Don't want to stop lsb resource on migration
> > 
> >  
> > On 14 Mar 2014, at 1:00 am, Bingham <knee-jerk-reaction at hotmail.com> wrote:
> >  
> > > Hello,
> > > 
> > > My setup:
> > >          I have a 2 node cluster using pacemaker and heartbeat. I have 2 resources, ocf::heartbeat:IPaddr and lsb:rabbitmq-server.       
> > >          I have these 2 resources grouped together and they will fail over to the other node.
> > > 
> > > 
> > > 
> > > question:
> > >          When rabbitmq is migrated to node1 from node2 I would like to 'not' have the the </etc/init.d/rabbitmq-server stop> happen on the failed server (node1 in this example).
> >  
> > 'migrate' has special meaning here.
> > After a failure rabbitmq is moved (stopped on the old node and started on the new one), which is different from a migration.
> >  
> > Leaving rabbitmq in an unclean state on node1 would definitely not be a good idea.
> >  
> > > 
> > >           Is it possible to do this in crm? 
> > >  
> > >          I realize that I could hack the initscript's case statement for stop to just "exit 0", but I am hoping there is a way to do this in crm.
> > > 
> > >   
> > > Thanks for any help,
> > > Steve
> > >           
> > > 
> > > _______________________________________________
> > > 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
> > _______________________________________________
> > 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
> _______________________________________________
> 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 --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 841 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20140404/4bf62ce4/attachment-0003.sig>


More information about the Pacemaker mailing list