[Pacemaker] Problems with Pacemaker 1.1.8 on F17

[Lars Kellogg-Stedman]

Howdy all,

I'm trying to run Pacemaker 1.1.8 (from
http://clusterlabs.org/rpm-next/fedora-17/clusterlabs.repo) in a
two-node configuration under Fedora 17.  I'm running into a few
different problems.

I'm trying to create resources using the systemd: manager.  If I run
"pcs resource standards" I see:

  ocf
  lsb
  service
  systemd
  stonith

But if I try to run something like "pcs resource create httpd
systemd:httpd" I get:

  Error: Unable to create resource 'systemd:httpd', it is not
  installed on this system (use --force to override)

The unit file exists:

  # ls -l /lib/systemd/system/httpd.service
  -rw-r--r-- 1 root root 434 Jan 29 07:32 /lib/systemd/system/httpd.service

It looks like I can use the "--force" option to create the resource
anyway (...okay), but the systemd: resource agent doesn't appear to
track services correctly, so if, for example, httpd fails to start:

  # systemctl status httpd
  httpd.service - The Apache HTTP Server (prefork MPM)
            Loaded: loaded (/usr/lib/systemd/system/httpd.service; disabled)
            Active: failed (Result: exit-code) since Thu, 21 Feb 2013 16:57:17 -0500; 41s ago
           Process: 7236 ExecStart=/usr/sbin/httpd $OPTIONS -k start (code=exited, status=1/FAILURE)
            CGroup: name=systemd:/system/httpd.service

Pacemaker thinks it's running happily:

  # pcs resource show
  httpd  (systemd:httpd):        Started 

Along with all of this, corosync (corosync-2.3.0-1.fc17.x86_64) keeps
falling over and I end up with pages of this in my logs:

  Feb 21 16:51:33 host1 corosync[13399]: [TOTEM ] Retransmit List: 96

At this point, I can't stop pacemaker without a "kill -9", and I need
to restart corosync to get it communicating again.

-- 
Lars Kellogg-Stedman <lars at oddbit.com>