HP Serviceguard Linux NFS-Toolkit-A.03.03.00 User Guide (696981-001, June 2012)

To start the package, use either of the following commands:
• # cmrunpkg <package_name>
• # cmrunpkg -n <nodename> <pkgname>
Stopping the package
The Serviceguard package manager invokes the NFS toolkit script with stop as an input paramenter.
This in turn does the following:
• Unexports directory on the file system.
• Stops NFS server daemons, if lock migration is enabled.
• Stops NFS service processes (including nfs.mon).
• Stops NFS file lock migration process.
To stop the package, run the following command:
# cmhaltpkg <package_name>
Maintaining the package
Serviceguard normally monitors critical processes, but there might be situations when a planned
maintenance task (for example, changing the configuration) affects those monitored processes.
The maintenance flag can be used to signal to toolkit that it must not failover the instance to an
adoptive node in case of NFS daemon failure. However, Serviceguard still monitors other resources
used by the toolkit package. In case, any of the resources fails, Serviceguard fails over the toolkit
package to the adaptive node.
To set the package to the maintenance mode:
1. To enable the maintenance mode feature in the package configuration file, set the
MAINTENANCE_FLAG attribute to yes, and then apply the configuration file using the
cmapplyconf command.
2. To start the maintenance mode for the package, create nfs.debug file under TKIT_DIR:
# touch <<TKIT_DIR>>/nfs.debug
To bring back the package from the maintenance mode to the running state, remove the nfs.debug
file placed under TKIT_DIR, perform the following command:
# rm <<TKIT_DIR>>/nfs.debug
Deleting the package
To delete the package from the cluster, run the following command:
# cmdeleteconf -p <package_name>
This command prompts for a confirmation, before removing the package configuration from the
cluster binary file, unless you use the -f option.
Mounting NFS shares at client node
On the client node, you must NFS-mount a file system using the package name in the mount
command. The package name is associated with the relocatable IP address of the package. On
the client systems, you must use a hard mount. For auto-mounter, the timeout must be greater than
the total end-to-end recovery time for the Serviceguard NFS package, which includes failover time,
running fsck, mounting file systems, and exporting file systems on the new node. The default
value of the timeout is five minutes. If you set the timeout to zero, it does not allow you to unmount.
Managing the NFS toolkit package 17