10. MODIFICATION › 10.1 Parameter Modification › 10.1.3 Implement Incremental Update

10.1.3 Implement Incremental Update

To implement incremental update in the CA MICS VAX/VMS
Analyzer, follow the checklist provided below:

See Section 7.3.1 of this guide for a detailed description of
the statement syntax.


   ****************************************************
   *                                                  *
   *           IMPLEMENT INCREMENTAL UPDATE           *
   *                                                  *
   ****************************************************

___ 1.  Edit prefix.MICS.PARMS(cccOPS), where (ccc) is the
        component identifier:

        o  Specify the following:

           INCRUPDATE  YES

        o  If you want to store the incremental update
           database files on tape between incremental
           updates, specify this:

           INCRDB  TAPE  #gdgs

        o  If you want to allocate the incremental update
           database files during the first incremental update
           of the day and delete these data sets at the end
           of the DAILY job step, specify this:

           INCRDB  DYNAM

        o  If you specified INCRDB TAPE or INCRDB DYNAM, then
           you must also specify this:

           INCRDETAIL  data_set_allocation_parameters
           INCRDAYS    data_set_allocation_parameters

        o  If you want the incremental update job for this
           product to get input measurement data from the
           output of the SPLITSMF job, specify this:

           INCRSPLIT  USE  data_set_allocation_parameters

        o  For additional information on related topic,
           review the documentation for this product on
           INCRCKPT, INCRDETAIL, INCRDAYS, or INCRSPLIT
           parameters to override default data set allocation
           parameters.

___ 2.  Submit the job in prefix.MICS.CNTL(cccPGEN).

___ 3.  Edit prefix.MICS.PARMS(JCLGENU) so that it contains
        two or more lines reading:

        DAILY
        INCRccc cccIUALC cccIUGDG

___ 4.  Submit the job in prefix.MICS.CNTL(JCLGENU).  Ensure
        that there are no error messages in MICSLOG or
        SYSTSPRT, that the MICSLOG contains the normal
        termination message, BAS10999I, and that the job
        completes with a condition code of zero.

___ 5.  Edit the job in prefix.MICS.CNTL(cccIUALC).

        o  Inspect and/or specify data set allocation
           parameters for the incremental update database and
           checkpoint files.  If you specified INCRDB TAPE or
           INCRDB DYNAM, the cccIUALC job will only allocate
           the incremental update checkpoint data set.

        o  Submit the job.  Ensure that there are no error
           messages in MICSLOG or SASLOG, and that the job
           completes with a condition code of zero.

___ 6.  If you specified INCRDB TAPE, submit the job in
        prefix.MICS.CNTL(cccIUGDG) to define generation group
        indexes for the incremental update DETAIL and DAYS
        tape data sets.  Examine SASLOG, MICSLOG, and
        SYSPRINT to verify that the generation group indexes
        were correctly defined.

        Note:  You may see error messages for the DLTX (or
        DELETE) statements.  This is not a problem.  cccIUGDG
        deletes each index prior to defining it, and an error
        message is issued if the index does not yet exist
        (e.g., if this is the first time you ran the cccIUGDG
        job).

___ 7.  The following operational job(s) have changed:

        DAILY  INCRccc

        If your site has implemented the operational CA MICS
        processes in a scheduling product, the JCL may have
        to be refreshed in that product.  See the scheduling
        product's administrator for the exact processes
        involved in updating that product's representation of
        the CA MICS jobs.

___ 8.  Implement operational procedures for gathering input
        measurement data and executing incremental updates
        (INCRccc) during the day.

        You may also need to modify operational procedures
        for the DAILY job to ensure that processing is
        limited to input measurement data that has not been
        input to one of the day's incremental update
        executions.

SPECIAL CONSIDERATIONS:
-----------------------

There are a number of files in the VAX/VMS Analyzer database
that contain metrics which are the result of "delta"
calculations done from one interval to the next.  In these
files, the first interval encountered on any given processing
run is used to initialize the values for the subsequent
"delta" calculations.  Because of this, that first time
interval in most cases will NOT be represented in the output
file.

This architecture is acceptable as it only comes into play
once for each DAILY processing cycle.  When the
Incremental Update feature is activated, this condition may
become more noticeable, depending on the number of increments
in any given DAILY processing cycle.

It is important to schedule your incremental cutoffs at off
peak times to avoid the reporting impacts of a missing
interval.

The following VAX/VMS Analyzer files have been identified
as being susceptible to the condition described above:

    DEMDSK - VMS MONITOR Disk File
    DEMMOD - VMS MONITOR Modes File
    DEMPRO - VMS MONITOR Process File
    DEMSCS - VMS MONITOR SCS File
    DEMSPR - VMS System Profile File
    DESPRX - VMS Process Activity File

Another important area of consideration is the processing of
DEXSUS data.  DEXSUS data is created by running, once a day,
a script of commands on a VAX machine, which takes a snapshot
of the system environment.  Since this is a once a day
activity, the DEXSUS data will only be processed in the DAILY
processing step DAY074.  The DEXSUS data will not be
processed in the INCRDEX job even though the DD for the data
is present.