E.9. The cluster.xml File

The cluster configuration file, /etc/cluster.xml, contains detailed information about the cluster members and services. Do not manually edit the configuration file. Instead, use the Cluster Configuration Tool to modify the cluster configuration.

When you run Cluster Configuration Tool, cluster-specific information is entered in a hierarchical XML format. The following is a description of each area of configuration, from daemon and shared storage to cluster members and services. Note that the back slash (\) represents a continuation of a single line.

<?xml version="1.0"?>
<cluconfig version="3.0">
  <clumembd broadcast="no" interval="500000" loglevel="4" multicast="yes" \ 
	    multicast_ipaddress="225.0.0.11" thread="yes" tko_count="20"/>
  <cluquorumd loglevel="6" pinginterval="" tiebreaker_ip=""/>
  <clurmtabd loglevel="4" pollinterval="4"/>
  <clusvcmgrd loglevel="4"/>
  <clulockd loglevel="4"/>
  <cluster config_viewnumber="18" key="7a497d303feefeef0f8be9b72697aaed" name="Octane"/>

The above fields contain versioning information and cluster daemon operation parameters such as logging levels, networking addresses, and more. For more information on configuring cluster daemon parameters, refer to Section 3.6 Configuring Cluster Daemons.

<cluster config_viewnumber="18" key="7a497d303feefeef0f8be9b72697aaed" \
	  name="Test_cluster"/>
<sharedstate driver="libsharedraw.so" rawprimary="/dev/raw/raw1" \
	  rawshadow="/dev/raw/raw2" type="raw"/>

The fields above define the cluster quorum and shared cluster configuration parameters. The driver and raw partition information about the shared primary and backup partitions are also specified in these fields. For more information about configuring the shared partitions, refer to Section 3.5 Editing the rawdevices File.

  <members>
    <member id="0" name="clu1" watchdog="yes">
      <powercontroller id="0" ipaddress="192.168.65.51" password="apc" \
	  port="1:1" type="apcmaster" user="apc"/>
    </member>
    <member id="1" name="clu2" watchdog="yes">
      <powercontroller id="0" ipaddress="192.168.65.52" password="baytech" \
	  port="1" type="baytech" user="admin"/>
    </member>
    <member id="2" name="clu3" watchdog="yes">
      <powercontroller id="0" ipaddress="192.168.65.53" password="baytech" \
	  port="2" type="baytech" user="admin"/>
    </member>
    <member id="3" name="clu4" watchdog="yes">
      <powercontroller id="0" ipaddress="192.168.65.54" password="wti" \
	  port="blue" type="wti_nps" user=""/>
    </member>
    </members>

The fields above define the cluster and its individual members. Each member field contains identification and configuration information, including cluster names, addresses, power controllers and types, and authentication details. For more information on configuring cluster members, refer to Section 3.7 Adding and Deleting Members.

  <services>
    <service checkinterval="0" failoverdomain="None" id="0" name="test" \
	userscript="None">
      <service_ipaddresses/>
    </service>
    <service checkinterval="0" failoverdomain="foodomain" id="1" name="test2" \
	userscript="None">
      <service_ipaddresses/>
    </service>
  </services>

The fields above define the services controlled by the cluster system, such as NFS, Samba, and HTTP. The parameters in these fields include service names, failover domain names, service status check intervals, and location of service init scripts (if applicable). For more information about configuring clustered services, refer to Section 3.10 Adding a Service to the Cluster.

  <failoverdomains>
    <failoverdomain id="0" name="fonfs" ordered="yes" restricted="yes">
      <failoverdomainnode id="0" name="clu2"/>
      <failoverdomainnode id="1" name="clu3"/>
    </failoverdomain>
    <failoverdomain id="1" name="fosamba" ordered="no" restricted="no">
      <failoverdomainnode id="0" name="clu1"/>
      <failoverdomainnode id="1" name="clu3"/>
    </failoverdomain>
  </failoverdomains>

The fields above define the failover domains that control the priority and order in which the cluster members queue in the event of failover. Parameters in these fields include failover domain name, restricted and ordered toggling, and node order by member name. For more information about configuring failover domains for cluster systems, refer to Section 3.9 Configuring a Failover Domain.