9. PROCESSING › 9.3 Daily Processing Flow for HSM and HSB (DAY095) › 9.3.3 Phase 2 - Sort and Build Files

9.3.3 Phase 2 - Sort and Build Files

This phase begins by checking to see if multiple inventories
(unique SYSID and ENDTS) were present in the input.

If YES, then a flag is set for passing to the DAYS
summarization routines (#fffDSUM) which signals that the data
from the DETAIL timespan will need to be SORTED.

If NO, then the DAYS summarization routines may be able to
skip the SORTing of the DETAIL data into the DAYS timespan
sequence. This status check is an indicator of possible
performance improvements; refer to Appendix A for an
explanation of message HSM00300I.

In general, phase 2 processing aims to accomplish the
following objectives:

1. PROC SORT with the NODUP option on all WORK files

   a. Sufficient granularity is introduced by adding such
      extra fields such as MIGRATTS (time of migrate or time
      of backup) to arrange any duplicate observations next
      to each other so PROC SORT can purge the duplicates
      automatically.

2. Final processing prior to updating DETAIL timespan.   Each
   file has slightly different requirements.

   a. Backup versions are summarized to the DSNAME level
      (aggregate all versions of a given data set name using
      normal CA MICS summarization macros).

      Once reduced to a single observation per data set, then
      the following exit points are called:

         - DSNODEs are parsed from the full DSNAME
         - HSMACRT is %INCLUDEd to assign account codes
         - The cost routine generated by CA MICS Accounting
           and Chargeback is called to price the current
           inventory of backup versions (journal file for
           backup resources will only be created if
           ACCOUNTING DETAIL is in effect)
         - The USRSBAC exit is called to allow ultimate user
           control over content and existence of this
           observation.

      Once built in this format, the HSBBAC00 file is sorted
      again into the final sequence dictated by the DETAIL
      timespan. If the DETAIL timespan is inactive to save
      space, it is left on a WORK file for later absorption
      into the DAYS timespan.

   b. Migrated data set observations have already been
      assigned account codes in phase 1. Therefore:

         - The cost routine generated by CA MICS Accounting
           and Chargeback is called to price the migrated
           data set (again, only if ACCOUNTING DETAIL is in
           effect).
         - The USRSMIG exit is called to allow ultimate user
           control over content and existence of this
           observation.

      The HSMMIG00 file is already sorted into the final
      sequence dictated by the DETAIL timespan and thus does
      not need to be sorted again.  If the DETAIL timespan
      is inactive to save space, it is left on a WORK file
      for later absorption into the DAYS timespan.

   c. DASD planning observations do not have account codes
      and are already in their final DETAIL sequence and so:

         - The USRSVOL exit is called to allow ultimate user
           control over content and existence of this
           observation.

   d. Tape planning observations do not have account codes
      and are already in their final DETAIL sequence and so:

         - The USRSTAP exit is called to allow ultimate user
           control over content and existence of this
           observation.