This site best when viewed with a modern standards-compliant browser. We recommend Firefox Get Firefox!.

Linux-HA project logo
Providing Open Source High-Availability Software for Linux and other OSes since 1999.

USA Flag UK Flag

Japanese Flag

Homepage

About Us

Contact Us

Legal Info

How To Contribute

Security Issues

18 August 2008 Heartbeat release 2.1.4 is now out Download it and install it!

11 October 2007 NEW educational HA/DR Blog hosted by Alan Robertson

9 April 2007 Check out the Cool Heartbeat Screencasts: Installation, Intro to the GUI Part of the Heartbeat Education project

Last site update:
2008-08-28 08:24:05

System Prerequisites

The following services must be started in this order before starting the ClusterResourceManagerDaemon:

  1. The HeartbeatProgram must be running

    1. heartbeat needs to be started with the -M switch to disable resource management
    2. all nodes should have been added to ha.cf using the node directive

  2. The ClusterConcensusMembership must be running
    add the line: respawn hacluster /full/path/to/ccm to ha.cf or start it by hand

  3. The LocalResourceManager must be running
    add the line: respawn hacluster /full/path/to/lrmd to ha.cf or start it by hand

If the HeartbeatProgram is not running, the ClusterResourceManagerDaemon will exit with status somestatus indicating that it should NOT be started again. The only situation in which this should happen is when the CRM is being started by hand.

If either of the last two components are not running, the ClusterResourceManagerDaemon (CRMd) will exit with status someotherstatus indicating that it should be started again later once they are running.

If any of the systems above fail once the CRMd has been started, the CRMd will exit with the same code as it would in the startup case.

Eventually the script that starts the HeartbeatProgram may also start everything else in the correct order. Possibly this will only consist of the CRMd and the HeartbeatProgram with all other components being started directly from ha.cf

Configuration Prerequisites

At this point you should configure part of the ClusterInformationBase by hand. The configuration should be located in /var/lib/heartbeat/crm/cib.xml depending on your OS flavour. Right now the minium information required is the node entry for the machine that will be first powered up and elected as DC. This requirement will be removed at a future point. Optionally you may also choose to configure all the nodes, resources, constraints and general CRM parameters. See crm/crm-1.0.dtd in CVS for details on the required format.

Minimum configuration sample:

<?xml version="1.0"?>
<!DOCTYPE cib SYSTEM "crm-1.0.dtd">
<cib>
 <configuration>
  <nodes>
    <node id="node1" uname="node1" type="member"/>
  </nodes>
  <resources/>
  <constraints/>
  <crm_config/>
 </configuration>
 <status>
 </status>
</cib>

Startup

The CRMd can now be started by hand. It requires no command-line arguments. Afterwards, further configuration can be done with:

  1. the text editor of your choice provided that the CRM system is not running at all
  2. the crmadmin tool if the CRM system is running
  3. the GUI tool(s) once they're written

See ClusterResourceManagerDaemon for further details of what happens inside the CRMd on startup and shutdown as well as how corner cases such as starting the first CRMd and stopping the last CRMd are handled.