CA Dispatch uses the VSGMU210-Sysgen Archive Control screen (Option 9.A from the main menu) defined retention under two specific circumstances:
1. The first time a report ever gets achived. (After which, we create a database record that will contain the retention for that report).
2. If the database retention record we created for the report (Seen on the VARMU100 screen, option A.1.S from the main menu) contains a value of 'SD' in the DAYS or GENS fields on this screen. The value of 'SD' on the VARMU100 screen tells CA Dispatch to use the "System Default" which is actually the VSGMU210 SYSGEN defined retention.
Should it become necessary for you to update archive retentions, you can update them MANUALLY by modifying the appropriate reports VARMU100 screen and defining the desired retention value. Or, for reports that have a DAYS or GENS value of "SD" on their VARMU100 screen, by modifying the VSGMU210-Sysgen Archive Control screens retention value.
*Note* Changes to either the VARMU100 or the VAGMU210 screen are effective immediately and will be invoked when the nightly DSEXTMIG/DSEXPSAR archive batch maintenance job is run to delete expired entries.
If you have a lot of VARMU100 screen database records that need to be updated, we provide a BATCH utility known as the DATABASE LOAD/UNLOAD utility that you may find useful. The archive retentions for reports are located in the "TYPE J" records, as recognized by this batch database update utility.
To assist you with executing the Database load/Unload utility to update the TYPE J archive retention records, below is a basic "how to" outline. However, you may need to read the documentation and review the associated jobs JCL FIRST, so that the following information makes more sense...
Basics are as follows:
To UNLOAD the TYPE J records...
1. Create a dataset in TSO as FB,255,255
2. Go into this dataset and code the unload control card based on the
format for unloading the type J record in the "UNLOAD RECORD FORMAT"
section of the documentation. Specifically, the System
Programmers guide documents the format of the UNLOAD control card
for a TYPE J record in the section entitled "Type J - Archive Report
Example: UJM * * *
In the above example, all type J records would be unloaded. To filter
what records will be unloaded, code the actual REPORT or JOBNAME in
place of the asterisks at the appropriate offset.
3. Save the unload file.
4. Update the DSEXLOD/DSEXLODL JCL "CONVERT" symbolic (which is always your
input) to point to the file you created that has the control card. If
this is the FIRST time you are running the job, also set ALLOC='LT' so
that the UNLOAD file is created. Submit the job.
5. The UNLOAD file will have the actual TYPE J records in it. Edit this
file and make your changes. Save the file.
* The Offset/Columns containing archive RETENTIONS are: 21 (GENS) and 24 (DAYS)
(See the documentation section entitled "Type J - Archive Report Definition - Add, Modify, Delete").
To RELOAD the TYPE J records...
6. To RELOAD the records, simply change the "CONVERT" symbolic to point to
this UNLOAD file. It is your input file for the reload. Also, set
ALLOC='LE' because at this point, the files have already been created.
* KEY POINTS *
-The CONVERT symbolic in your submit JCL always points to the "input" file for the utility.
-When "UNLOADING" the CONVERT symbolic points to the file that has the unload CONTROL CARD. This run will create an UNLOAD dataset that has all of the records in it.
-When "LOADING" the CONVERT symbolic points to the UNLOAD file that has all of the updated records in it.