Reworked my Ansible playbook to install, configure and start Corosync and Pacemaker. Only tested it on CentOS 6.2 64 bit. The playbook now makes use of the built-it Ansible modules to enable services at boot time and to configure SELinux.

Ansible playbook for corosync

# This playbook installs and configure corosync and pacemaker on a set of nodes part of a given group.
---

# See the file /etc/ansible/hosts where the group is actually defined. There might be a way to define groups in a file 'closer' to this playbook.
- hosts: clusternodes
  vars:
    mcastport : 5405
  tasks:
  
  # It would be better to use Ansible to change the IP tables config to allow corosync/pacemaker on the nodes part of the cluster.
  - name: ensure iptables is stopped and is not running at boot time.
    action: service name=iptables state=stopped enabled=no

  - name: ensure python-selinux is installed on all nodes.
    action: yum name=libselinux-python state=installed  

  - name: set SELinux to permissive
    action: selinux policy=targeted state=permissive

  - name: ensure SELinux is turned off
    action: command /usr/sbin/setenforce 0

  - name: ensure corosync is installed.
    action: yum name=corosync state=installed

  - name: write the corosync config files
    action: template src=corosync.conf dest=/etc/corosync/corosync.conf

  - name: add the pcmk file in the folder service.d for corosync configuration
    action: copy src=pcmk dest=/etc/corosync/service.d/pcmk

  - name: ensure Pacemaker is installed.
    action: yum name=pacemaker state=installed

  - name: ensure corosync is running (and automatically at boot time)
    action: service name=corosync state=started enabled=yes

  - name: ensure pacemaker is running automatically at boot time
    action: command /sbin/chkconfig pacemaker on

  - name: ensure Pacemaker is running (and also enabled to start at boot time). For some reason, Pacemaker will fail to start when the steps below are done synchronously.
    action: service name=pacemaker state=started enabled=yes
    async: 60
    poll:  5

Ansible template for Corosync

# Please read the corosync.conf.5 manual page
compatibility: whitetank

totem {
  version: 2
	secauth: off
	interface {
		{% for host in hostvars %}
  		member {
			memberaddr: {{ hostvars[host]['ansible_eth1']['ipv4']['address']}}
		}
		{% endfor %}
		ringnumber: 0 
		bindnetaddr: {{ bindnetaddr }}
		mcastport: {{ mcastport }}
		ttl: 1
	}
	transport: udpu
}

logging {
	fileline: off
	to_logfile: yes
	to_syslog: yes
	debug: on
	logfile: /var/log/corosync.log
	debug: off
	timestamp: on
	logger_subsys {
		subsys: AMF
		debug: off
	}
}

Configure Corosync to use Pacemaker

service{
  name:	pacemaker
	ver:	1
}

Where does the {{ bindnetaddr }} variable in this example come from? Manually, I'd use ipcalc to find that value but I'm curious if you've found a better way.