 |
» |
|
|
|
Once you have configured your Serviceguard cluster, you should
verify that the various components of the cluster behave correctly
in case of a failure. In this section, the following procedures
test that the cluster responds properly in the event of a package
failure, a node failure, or a LAN failure. | | | |  | CAUTION: In testing the cluster in the following procedures,
be aware that you are causing various components of the cluster
to fail, so that you can determine that the cluster responds correctly
to failure situations. As a result, the availability of nodes and
applications may be disrupted. | | | | |
Start
the Cluster using Serviceguard Manager | |
If you have just finished configuring your cluster, it starts
automatically. If it is halted later, restart it: from the System
Management Homepage (SMH), select the cluster and choose Administration -> Run Cluster... Testing
the Package Manager | |
You can test that the package manager is operating correctly.
Perform the following procedure for each package on the cluster: Obtain the PID number of a service in the package by entering ps -ef | grep <service_cmd> |
where service_cmd is the executable specified
in the package control script with the parameter SERVICE_CMD. The service selected must not have SERVICE_RESTART specified. To kill the service_cmd PID,
enter To view the package status, enter The package should be running on the specified adoptive node. Move the package back to the primary node (see “Moving
a Failover Package ”).
Testing
the Cluster Manager | |
To test that the cluster manager is operating correctly, perform
the following steps for each node on the cluster: Turn off the power to the node SPU. To observe the cluster reforming, enter the following
command on some other configured node: You should be able to observe that the powered down node is
halted, and that its packages have been correctly switched to other
nodes. Turn on the power to the node SPU. To verify that the node is rejoining the cluster,
enter the following command on any configured node: The node should be recognized by the cluster, but its packages
should not be running. Move the packages back to original node (see “Moving
a Failover Package ”). Repeat this procedure for all nodes in the cluster
one at a time.
Testing
the Network Manager | |
To test that the network manager is operating correctly, for
each node on the cluster do the following: To identify primary and standby lan cards on the node, enter and then Disconnect the LAN connection from the Primary card. Verify that a local switch has taken place so that
the Standby card is now the Primary card. In Serviceguard Manager,
check the cluster properties. On the command line, use cmviewcl -v. Reconnect the LAN to the original Primary card,
and verify its status. In Serviceguard Manager, check the cluster
properties. On the command line, use cmviewcl -v.
|
Printable version
