07.03.2011 1:31, Todd Nine:
I would like the following distribution.
Single Node: haproxy and ec2setip
So you need a mandatory colocation constraint for them.
All other Nodes: nginx.
Multiple instances of a resource => clones.
Additionally, you may want to set negative colocation constraint between
nginx and haproxy.
All nodes: chef-client
Perhaps this should not be a part of your cluster at all. But if you
want to monitor chef-client, and restart it in case of failure, you may
set it up as a clone resource.
Essentially, I use ha proxy for load balancing. I use nginx for ssl
decryption and serving static pages, so I want that to run on every
node that isn't the ha proxy node. During a failover, I want haproxy
to be started, and ec2setip to be run on a single node, and all other
nodes to start nginx. I'm not using STONITH on purpose. If one node
takes over the IP and another is running, it does not affect my
service since none of my clustered services perform any data write.
I'm using the following configuration, and importing it with this
command.
I'm not sure, but I suspect it will not work. The cluster will not take
over the resource from a disconnected node without STONITH. You will
need to configure this behaviour explicitly.
--
Pavel Levshin
_______________________________________________
Pacemaker mailing list: Pacemaker@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://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
