5.1 Configuration of PFM - RM for Platform in a cluster system
This section describes the configuration in which PFM - RM for Platform is run in a cluster system. For an overview of cluster systems and details about the system configuration for running the Performance Management system in a cluster system, see the chapter that describes system construction and operations in a cluster system in the JP1/Performance Management User's Guide.
If you run PFM - RM for Platform in a cluster system, you can improve availability because the system operation can be maintained by failover in the event of a problem.
To run PFM - RM for Platform in a cluster system, you must configure an environment in which the same instance of PFM - RM for Platform can be run in both the executing node and the standby node. Also you must store all data, such as data files, configuration files, and log files, on a shared disk.
When PFM - RM for Platform is run in a cluster system, it is configured as shown below.
|
|
![[Figure]](GRAPHICS/ZU030010.GIF)
|
|
![[Figure]](GRAPHICS/ZU030020.GIF)
PFM - RM for Platform in a cluster system runs in a logical host environment and monitors other hosts as monitoring targets. Therefore, you must configure the system in such a manner that the same host names can be used to connect to the monitored hosts from each host.
PFM - RM for Platform stores necessary information on the shared disk, such as definition and performance information, and inherits this information in the event of a failover. If a single logical host contains multiple Performance Management programs, all of them use the same shared directory.
You can run multiple PFM - RM for Platforms at the same node. If there are multiple cluster configurations (active-active configuration), run PFM - RM for Platform in each logical host environment. You can operate each PFM - RM for Platform independently and have them perform failover separately.