I found the solution, add colocation constraint like this:

pcs constraint colocation add vip-slave with Slave pgsql-master 100
pcs constraint colocation add vip-slave with Master pgsql-master 50

On Fri, May 13, 2016 at 6:08 PM, wd <[email protected]> wrote:

> hi,
>
> I'm running a master/slave 2 node PostgreSQL cluster, and 2 virtual ip
> resource, one is for master, one is for slave.
>
> I have add colocation constraint for both virtual ip and master/slave
> resource, to ensure virtual ip for master is run at master, and virtual ip
> for slave is run at slave. And also add constraint for PostgreSQL promote.
>
> pcs constraint colocation add vip-master with Master pgsql-master
> pcs constraint colocation add vip-slave with Slave pgsql-master
>
> pcs constraint order promote pgsql-master then start vip-master
> symmetrical=false score=INFINITY
> pcs constraint order demote pgsql-master then stop vip-master
> symmetrical=false score=0
>
> If I kill PostgreSQL master process, pacemaker will promote slave to new
> master, and move master-vip to the new master. But the slave-vip will stop.
> So how to let the slave-vip also move to the new master at this situation?
> And if I kill slave PostgreSQL, how to let the slave-vip move to the master?
>
_______________________________________________
Users mailing list: [email protected]
http://clusterlabs.org/mailman/listinfo/users

Project Home: http://www.clusterlabs.org
Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
Bugs: http://bugs.clusterlabs.org

Reply via email to