Step 1 Preload CALC Processing
>Preload CALC processing is an optional step that precedes the actual database load.
Preload CALC processing is an optional step that precedes the actual database load. Its intent is to improve the performance of load processing and is especially recommended, as follows:
- There are large amounts of DL/I data.
- There are logical relationships in the DL/I database.
- Space requirements need to be determined for the logical workfile(s) that will be generated by the load (Step 2).
Preload CALC processing performs the following operations:
- Accessing the IPSB load module
- Accessing the subschema module named in the IPSB
- Reading the DL/I input data
- Generating database page numbers for the DL/I root segments
- Updating the DL/I data with the database page numbers and writing it out to the DL/I output file
- Printing a report on the updated DL/I data
The following diagram shows the operations performed by preload CALC processing.
To execute the preload CALC processing step, use the JCL in CA IDMS DLI Transparency JCL.
The report produced by the preload CALC processing step lists the following:
- The DBDNAME for each DL/I database included in the input DL/I data
- The name and level for each DL/I segment, by database
- An indication if a segment is a logical child (LC) or logical parent (LP)
- The number of segment occurrences (records) found, by database
- The number of logical records found, by database
Following is an example of a report:
*** CA IDMS/DLI TRANSPARENCY DATABASE LOAD PROCESS=CALC,IPSB=ITEMPART DBDNAME=ITEMDBDL SEGMENT COUNT LEVEL RECORD ITEM 1086 01 ITEM LC DETAIL 3542 02 DETAIL TOTAL: 4628 RECORDS READ 3542 LOGICAL RECORDS 0 LOGICAL RECORDS WRITTEN *** CALC PROCESSING COMPLETE ___________________________________________________________________ *** CA IDMS/DLI TRANSPARENCY DATABASE LOAD PROCESS=CALC,IPSB=ITEMPART DBDNAME=PARTDBDL SEGMENT COUNT LEVEL RECORD LP PART 789 01 PART TOTAL: 789 RECORDS READ 789 LOGICAL RECORDS 0 LOGICAL RECORDS WRITTEN *** CALC PROCESSING COMPLETE
Preload Sorting (step 1, part 2)
Use Your Own Sort/Merge Utility
To further optimize the CALC-processed data for loading, you can sort it using your own sort/merge facility. As input to the sort/merge facility, supply the DL/I output file produced by the preload CALC processing. The output file will contain the CALC-processed data in sorted form. You can then use the sorted output file as input to the database load (Step 2).
The preload sort is not strictly required, but it should be performed to produce the most effective ordering of the CALC-processed data.
To perform the preload sort, you must use your own sort/merge facility.
What the Preload Sort Does
The preload sort performs the following operations:
- Accessing the CALC DL/I data produced by the preload CALC processing (Step 1, Part1)
- Sorting the data so that root segments (CALC records) are in descending database page sequence (the optimum CA IDMS/DB database load order)
To execute the preload sort processing step, use the JCL (Step 1, Part 2) in CA IDMS DLI Transparency JCL.