Pacemaker is not started and stopped automatically with Corosync

[David Moreau Simard]

In Ubuntu 12.04 with:
# dpkg -l |egrep "pacemaker|corosync"
ii corosync 1.4.2-2ubuntu0.2 Standards-based cluster framework (daemon and modules)
ii pacemaker 1.1.6-2ubuntu3.3 HA cluster resource manager

When I start corosync, it'll start the pacemaker resources:
# ps aux |egrep "(heartbeat|corosync)"
root 27043 0.2 0.0 511172 3756 ? Ssl 19:43 0:00 /usr/sbin/corosync
root 27051 0.0 0.0 84716 3124 ? S 19:43 0:00 /usr/lib/heartbeat/stonithd
109 27052 0.1 0.1 88768 5856 ? S 19:43 0:00 /usr/lib/heartbeat/cib
root 27053 0.1 0.0 97432 3256 ? S 19:43 0:00 /usr/lib/heartbeat/lrmd
109 27054 0.0 0.0 84756 3364 ? S 19:43 0:00 /usr/lib/heartbeat/attrd
109 27055 0.0 0.0 79040 2872 ? S 19:43 0:00 /usr/lib/heartbeat/pengine
109 27056 0.0 0.0 95476 4028 ? S 19:43 0:00 /usr/lib/heartbeat/crmd

When I stop corosync, it'll stop them as well.

In Ubuntu 14.04 with:
# dpkg -l |egrep "pacemaker|corosync"
ii corosync 2.3.3-1ubuntu1 amd64 Standards-based cluster framework (daemon and modules)
ii crmsh 1.2.5+hg1034-1ubuntu4 all CRM shell for the pacemaker cluster manager
ii libcorosync-common4 2.3.3-1ubuntu1 amd64 Standards-based cluster framework, common library
ii pacemaker 1.1.10+git20130802-1ubuntu2.3 amd64 HA cluster resource manager
ii pacemaker-cli-utils 1.1.10+git20130802-1ubuntu2.3 amd64 Command line interface utilities for Pacemaker

When I start corosync, it will NOT start the pacemaker resources. I need to start pacemaker manually (service pacemaker start or /etc/init.d/pacemaker start).

This results in nothing working until I figured that out. Yielding errors such as:
---
# crm status
Could not establish cib_ro connection: Connection refused (111)
ERROR: crm_mon exited with code 107 and said: Connection to cluster failed: Transport endpoint is not connected
---
# crm_mon
Attempting connection to the cluster...Could not establish cib_ro connection: Connection refused (111)
---

In my testing, both the precise and trusty releases had Pacemaker in the service.d directory as such:
---
service {
  name: pacemaker
  ver: 0
}
---

Is this a bug or is it expected that Pacemaker is no longer started and stopped by corosync ?

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[Gui||aume (michaudg)]

Same problem here. temporarily fix with this :

service pacemaker start
update-rc.d pacemaker defaults

[Baohua Yang (yangbaohua)]

same problem in ubuntu 14.04

[Elbandi (ea333)]

look the important section:
http://clusterlabs.org/doc/fr/Pacemaker/1.1-pcs/html/Clusters_from_Scratch/_configuring_corosync.html

[Louis Bouchard (louis)]

Hello,

Well, elbandi have found the answer to your question : pacemaker now needs to be started as a separate service. To quote the document :

"Pacemaker used to obtain membership and quorum from a custom Corosync plugin. This plugin also had the capability to start Pacemaker automatically when Corosync was started.
Neither behavior is possible with Corosync 2.0 and beyond as support for plugins was removed. Instead, Pacemaker must be started as a separate service."

Looks like the services are correctly setup as pacemaker starts normally at boot time. If you want to manually stop your cluster, you will have to issue :

$ service corosync stop
$ service pacemaker stop