Managing HP Serviceguard for Linux, Sixth Edition, August 2006

Configuring Packages and Their Services
Creating a Disk Monitor Configuration
Chapter 6208
<serverPriority>
<schedClass>normal</schedClass>
<schedPriority>1</schedPriority>
</serverPriority>
<resourceConfigCategoryList>
<resourceCategory>
<categoryName>Pkg1</categoryName>
<agent>/usr/local/cmcluster/bin/cmcheckdisk</agent>
<agentTimeout>60</agentTimeout>
<pollingInterval>60</pollingInterval>
<resourceConfig>
<name>/dev/sdd1</name>
</resourceConfig>
<resourceConfig>
<name>/dev/sde1</name>
</resourceConfig>
</resourceCategory>
<resourceCategory>
<categoryName>Pkg2</categoryName>
<agent>/usr/local/cmcluster/bin/cmcheckdisk</agent>
<agentTimeout>60</agentTimeout>
<pollingInterval>60</pollingInterval>
<resourceConfig>
<name>/dev/sdf1</name>
</resourceConfig>
</resourceCategory>
</resourceConfigCategoryList>
</resourceMonitorServerConfig>
The cmconfigres create command can be issued when the cluster is
running, but packages that activate monitored disks cannot be running
because it is not possible for them to stay up if the cmresserviced
command fails. Also if the configuration is not complete, this service
fails.
After configuring monitoring, run cmresmond --start before starting
the cluster. Failure to do this will result in any disk monitor dependent
packages to fail at run time. The cmresmond --start command must be
executed on each node in order to start the disk monitoring daemon.
After initial configuration, cmresmond will start at each reboot.
To change an existing configuration, you must first delete the existing
configuration before you can create a new one. The cmconfigres create
command will not succeed if a configuration already exists. When the
server configuration is changed (via cmconfigres) and the disk monitor