Advanced Topics › Control and Checkpoint File Considerations

Control and Checkpoint File Considerations

Checkpoint files differ from control files in several ways:

• Control file data set names are specified on DD statements in the startup JCL procedure. Alternatively, control file data set names can be specified on ALLOCATE commands placed in the MIMINIT member to direct CA MIM to dynamically allocate the named control files. Checkpoint file names are constructed from a root name that you supply on the MIMINIT CHKPTDSN statement in the MIMINIT member. Alternatively, checkpoint files can be allocated by specific name using the ALLOCATE commands placed in the MIMINIT member.

  Note: For more information, see the CA MIM Statement and Command Reference Guide.

  Control files are identified by a ddname of MIMTBLnn, where nn is a numeric value from 00 to 99, inclusive. Checkpoint files are identified by a ddname of MIMCKPnn, where nn is a numeric value from 00 to 99, inclusive.

• Checkpoint files may never be shared between systems, but control files must be shared. Use the IFSYS and ENDIF statements in the MIMINIT member to define unique MIMINIT CHKPTDSN statements or unique ALLOCATE commands for each system in your complex.
• If the MIMINIT CHKPTDSN statement is used to define checkpoint files, then the data sets must be named sequentially, beginning at suffix 00, and increasing contiguously (01, 02, 03, and so on). However, if ALLOCATE commands placed in the MIMINIT member are used to allocate checkpoint files, the data set names can be in any sequence. Control files may be named in any sequence, whether specified through DD card images in the startup JCL procedure, or specified through ALLOCATE commands in the MIMINIT member.

Important: CA MIM does not support control files or checkpoint files on the high portion of the Extended Address Volume (EAV). If you use EAV, make sure you allocate data sets with EATTR=NO.