IMSRSC repository and RS catalog repository data sets

The IMSRSC repository uses a set of VSAM key-sequenced data sets (KSDSs) that you define to store the resource and descriptor definitions for the members of an IMSplex in a common repository.

In addition, the Repository Server (RS) has its own repository, called the RS catalog repository, that is used to maintain the definitions for the repositories for the IMS repository function. The RS catalog repository manages the association of repository names with repository data sets. An RS address space and the RS catalog repository are required to manage an IMSRSC repository.

You must create two pairs of VSAM KSDSs for each repository and for each RS catalog repository:
  • A primary repository index data set (RID) and a primary repository member data set (RMD). The RID stores the names and keys of all members within a repository. The RMD stores the member data indexed by the RID. These data sets are required.
  • A secondary RID and a secondary RMD. The secondary RID and RMD provide a duplex copy of the data in the primary RID and RMD. These data sets are required.
Recommendation: In addition to the required data sets, create a third pair of data sets (a spare RID and a spare RMD) to allow data set sparing. The spare RID and RMD are optional. They are empty data sets that are available to replace either the primary or secondary RID/RMD pair if a repository write failure occurs. If a repository write failure occurs on one of the data set pairs, the valid data set pair is copied to the spare and the failed copy is marked as discarded. This strategy reduces the potential for a repository outage because of a data set error condition.

You can create a spare RID and RMD for the IMSRSC repository, but not for the RS catalog repository.

The properties of each secondary and spare data set must be the same as the corresponding primary data set.

The IVP program that builds a sample IMS system includes support for dynamic resource definition (DRD) and the repository. You can use the IVP to define and allocate VSAM KSDSs that can be used as repository data sets. For information about the IVP sample jobs, see IMS Version 15.5 Installation.

Guidelines and restrictions for IMSRSC repository and RS catalog repository data sets

Take the following information into consideration when defining the repository data sets:
  • The REUSE attribute is required, because it allows the RS to automatically recover corrupt repository data sets at repository open time.
  • You can use only share options (1,3) or (2,3). The share options for the INDEX and DATA components of a repository data set must match.
  • The SPANNED attribute is not supported. Repository data sets must be non-spanned KSDSs.
  • The recommended CI size is 8 KB for both the INDEX and DATA components of all of the repository data sets.
  • The RS uses a single local shared resource (LSR) pool with a default 8 KB buffer size. For this buffer pool to be used optimally, make sure that the repository data sets have a matching 8 KB control interval size.
Recommendation: Define the primary, secondary and spare data set pairs on different volumes, to ensure availability. Ensure that the size of the secondary RID and RMD is greater than the size of the primary RID and RMD and that the size of the spare RID and RMD is greater than the size of the secondary RID and RMD.