Overview of the IMSRSC repository

A repository is a generalized data storage facility that can be used to store any type of information. The IMS resource definition (IMSRSC) repository is a set of VSAM key sequenced data sets (KSDSs) used for storing information. The IMSRSC repository is a repository to store resource (and descriptor) definitions for IMS databases, transactions, programs, and routing codes. This repository is an alternative to the resource definition data set (RDDS) that is used with the dynamic resource definition (DRD) function.

An RDDS stores the resource and descriptor definitions for one local IMS only, and two or more system RDDS data sets must be defined for each local IMS. The repository provides the ability for an IMSplex to use a single shared repository to store resource and descriptor definitions for all the members of an IMSplex.

A repository can maintain resource and descriptor definitions for up to 64 IMS systems in an IMSplex.

The repository is managed by the Repository Server (RS) address space. RS is in turn managed by the Common Service Layer (CSL) Resource Manager (RM). Resource and descriptor definitions can be added, queried, modified, or deleted from the repository by making requests to RM using type-2 commands.

The repository can store the resource and descriptor definitions that are supported by DRD. IMS can retrieve the stored definitions to dynamically generate runtime resources and descriptors. The resource and descriptor definitions that can be stored in the repository are:
  • Application programs
  • Databases
  • Fast Path routing codes
  • Transactions
The following Multiple Systems Coupling (MSC) resource definitions can also be stored in the repository:
  • MSC physical and logical links
  • MSC logical link paths
  • Remote logical terminals (LTERMs)

The following components work together to provide the overall IMS repository function:

The following illustration shows the relationship of the repository components.

Begin figure description. The IMS control region uses SCI to communicate with OM and RM. RM communicates directly with RS and CQS. RS is connected to the repository sets of data sets. A SPOC or automation program is connected to OM. End figure description.
The RS is a started task running under z/OS® in a sysplex environment. The RS must be running to provide the following functions:
  1. Interface with the client (RM or FRPBATCH utilities) using z/OS cross-system coupling facility
  2. Communication with the RS catalog repository data sets and management of one or more IMSRSC repository data sets
  3. Management of server registrations and repository connections
  4. Audit logging and trace facilities
  5. Repository data integrity
  6. Data compression and decompression

At any time, there is a single active or master RS address space. The RS can run on any logical partition (LPAR) in the sysplex where an IMSplex is running. It does not have to be on the same LPAR as any other IMS address space.

If there are multiple RS address spaces in the sysplex, one of them is considered the master and the others are subordinates. One master RS can manage all the repositories in the sysplex or there can be one master RS for each IMSplex.

One or more subordinate RS address spaces can also be started. The subordinate RSs wait in an initialization state. When the subordinate RSs identify that the master RS has terminated, they all attempt to complete the startup process. One subordinate RS becomes the new master. The others remain as subordinate RSs.

Recommendation: Use one master RS per IMSplex to achieve maximum availability and storage efficiency.

All requests for online work associated with the stored definitions flows from IMS to RS through RM. IMS also provides RM utilities that can perform work with the repository offline.

The RS address space has the following administrative interfaces: JCL batch (FRPBATCH) and z/OS MODIFY (F) command.