Configuration Best Practices › Practices for Improved Productivity and System Availability › Use Change Tracking

Use Change Tracking

Use change tracking in each CA IDMS central version by referencing a SYSTRK file in the execution JCL of the central version.

A SYSTRK file contains a description of the database environment most recently in use by the central version. During startup, an image of the current DMCL is written to SYSTRK along with information about database and journal files defined in the JCL.

Business Value:

The use of change tracking improves DBA productivity and provides for enhanced system availability by eliminating the need for manual intervention if a failure occurs during certain maintenance activities.

Change tracking lets you change the database environment of a central version in a fault-tolerant manner. Specifically, it permits the DBA to perform the following dynamic actions:

• Vary the data set name of a journal or database file within a central version without introducing the need for manual intervention in case of a central version failure.
• Vary a new version of a DMCL without introducing the potential for a warmstart failure.
• Vary the status of an area or segment permanently on a central version regardless of subsequent page range changes.
• Change the journal files in use by a central version and coordinate those changes with the associated archive journal jobs.

If the central version fails, the runtime database definition is restored from SYSTRK files during the restart, ensuring that the files being updated at the time of failure are the ones recovered by the warmstart unless explicitly overridden by changes in the JCL used to restart the central version.

Additional Consideration:

Change tracking requires the use of SYSTRK files for recording the current state of a central version’s database definition.

To implement change tracking for a central version, take the following steps:

1. Create a model SYSTRK file whose dataset name establishes the pattern for the individual SYSTRK files.
2. Create and format two to four SYSTRK files whose data set names are the same as that of the model SYSTRK file suffixed with a unique digit from 1 to 9. A minimum of two SYSTRK files are required due to internal mirroring, which is used to provide fault tolerance and recoverability in case of file damage.
3. Alter the central version execution JCL to reference the model SYSTRK file. Use a ddname established by the SYSIDMS parameter SYSTRK_DDNAME_PREFIX. The default value for this parameter is SYSTRK. SYSIDMS is a parameter file added to the JCL stream of batch jobs running in local mode or under the central version.
4. Change the JCL for the associated archive journal job to also reference the model SYSTRK file and to remove references to the disk journal files.
5. Optionally, change the JCL of other local mode jobs to reference the model SYSTRK file and remove explicit references to database files.

More Information:

For more information about the use of change tracking and how to allocate and format SYSTRK files, see the CA IDMS System Operations Guide.