Scenario QCF01: Configuring a full RTE with a high-availability hub monitoring server
If your environment meets the requirements, you should create a high-availability (HA) hub monitoring server. An HA hub is not defined to run on any specific system; it retains the same parameter values on any system in the sysplex. So it can be started on any LPAR using the same DVIPA address. Using an HA hub ensures that monitoring can continue with minimal disruption if the current system goes down, is shut down for maintenance, or suffers any other problem. You can use the IBM-provided template to create this RTE or you can configure it from scratch.
Before you begin
- It must be configured in a sysplex.
- It must have a dynamic virtual IP address (DVIPA), so it can respond to the same IP address on any LPAR in the sysplex. The DVIPA address must be defined on any LPAR that is a candidate for the HA hub.
- Its runtime libraries must be stored on shared DASD, so it can start on any LPAR in the sysplex without requiring replication of the libraries.
- It must be configured in its own runtime environment (RTE), without any monitoring agents configured in the RTE.
- Its runtime environment cannot use system variables.
Before you begin configuring the RTE, make sure that an application-instance-specific (private) DVIPA address has been created and defined to DNS as a new name (for example, OMEGAHUB). If you intend to enable the self-describing agent feature on the monitoring server, or to configure any products that require a z/OS® UNIX System Services file system, make sure that make sure that the required Hierarchical File System (HFS) or z/OS File System has been created or set aside and that it is mounted before you begin configuration. The file system must have access to a Java™ runtime environment running under IBM®'s 31-bit or 64-bit Java SDK Version 5 (or higher) on an HFS or zFS file system.