How to: Convert an existing static hub monitoring environment to a high-availability hub environment

To convert an existing hub environment to a high-availability (HA) hub runtime environment (RTE), you create a new RTE with an HA hub and then either create a new RTE with a remote monitoring server or convert the existing hub RTE to a remote and configure it and any other remotes to connect to the new HA. This topic explains how to perform this task using PARMGEN.

About this task

When converting from a standard, static z/OS hub to an HA hub, you create a new runtime environment (either full or sharing) for the HA hub. The old static hub is either decommissioned if no longer needed or replaced by a remote monitoring server. The newly converted remote will take over all non-hub responsibilities of the old hub, such as collecting data for OMEGAMON for z/OS product or hosting agents. The new HA hub handles only hub-type functions. To preserve any customization (situations, managed systems list, and the like), the Enterprise Information Base from the static hub can be given to the new HA hub. Any application support such as the catalog and attribute files for distributed agents that were manually added to the old, static hub will need to be copied from the old hub to the new HA hub.

Procedure

Create a new runtime environment and configure a high-availability hub in it (see Scenario QCF01: Configuring a full RTE with a high-availability hub monitoring server).

Before shutting down the old hub for the last time, create a member in RKANSQLU with the following statements:

DELETE FROM O4SRV.TSITDESC WHERE SITNAME LIKE 'UADVISOR_KM5_V*'; 
DELETE FROM O4SRV.SITDB WHERE RULENAME LIKE 'KM5.V*';

Execute the commands by using a Modify command:
```
F <hub stc>,CTDS START SPUFIL <name of member>
```
These situations are seeded using the configuration software and must not be in the database of the HA hub because they can trigger data collection attempts.
Back up the old hub runtime environment, preferably with the hub down.
Delete the following data sets from the new HA hub runtime environment and replace them with the old hub data sets. :
- %HILEV%.%RTENAME%.RKDS*
- %HILEV%.%RTENAME%.RKCF* if applicable
- %HILEV%.%RTENAME%.RKCP* if applicable
This gives the new hub the entire EIB of the old one as well as the MQ Configuration Database (if applicable) and the CICSplex Service Level parameters (if applicable).
Take one of the following paths:
- Build a new runtime environment for the LPAR where the old hub used to run. Define the monitoring server as remote and configure it to report to the new HA hub (see, for example, Scenario QCF04: Configuring a full RTE with a static hub monitoring server).
- Modify the existing runtime environment to convert the hub to a remote (see, for example, Scenario RTE04: Converting a hub monitoring server to a remote).
Reconfigure any portal server or agent connecting directly to the hub, such as Warehouse Proxy Agents. For the portal server, do not rebuild the portal server database but use the same one that was used for the old hub (respond No to the pop-up). Reconfigure any remote monitoring servers to report to the new HA hub (see Scenario RTE04: Converting a hub monitoring server to a remote for step-by-step instructions).
After a portal server is connected to the new HA hub, use the Managed Systems List (MSL) editor to review all custom MSLs that include monitoring server node IDs (if any) and use the situation editor to review all situations that are distributed to remote monitoring servers directly or to Managed Systems Lists that include monitoring server Node IDs. Also, review all Historical Data Collection configuration.