CA IDMS DLI Transparency Software Components › The Load Utility

The Load Utility

The load utility consists of the IDMSDLLD module. It accepts data unloaded from a DL/I database (via IBM's HD unload utility) and stores it in a CA IDMS/DB database. The CA IDMS/DB database must be prepared and initialized before running the load utility.

To execute, the load utility also requires:

• An IPSB load module. The IPSB translates the DL/I segment and data structure definitions to equivalent CA IDMS/DB record and set definitions. The load utility uses the DL/I-to-CA IDMS/DB equivalencies when storing the data in the CA IDMS/DB database. The IPSB definition must reflect the special considerations for a load IPSB (IPSB used with the load utility).
• CA IDMS/DB schema, subschema, and DMCL modules. The CA IDMS/DB modules constitute the runtime environment for the CA IDMS/DB database.

The process of loading the DL/I data can involve up to six steps. If the DL/I data does not include logical relationships, the only step required is the actual database load (Step 2). The steps in the load process are:

1. Preload CALC processing ── Calculates CA IDMS/DB preload database pages for DL/I root segments to speed up the actual load (Step 2). Included in this step is a sort of the preload CALC data.
2. Database load ── Stores the DL/I data in the CA IDMS/DB database. If logical relationships are found, the load utility writes the logical child records and their related logical parents to a workfile for additional processing. If the DL/I data comes from multiple databases (DBDs), a separate workfile is produced for each source database.
3. Workfile sort/merge ── Merges multiple workfiles from Step 2 and sorts the resulting file to arrange logical child records under their logical parents.
4. Prefix (concatenated key) resolution ── Processes the sorted workfile and generates correct prefixes (concatenated keys) for the logical child records.
5. Workfile hierarchical sort ── Sorts the workfile with resolved prefixes so that the logical child records are in their original DBD hierarchical sequences.
6. Prefix update ── Updates the logical child records in the CA IDMS/DB database with the generated prefixes. The prefixes are needed to establish the CA IDMS/DB set pointers for the logical child (member) sets and their logical parent (owner) sets.

Only Steps 1, 2, 4, and 6 invoke the IDMSDLLD module. Steps 3 and 5 (the sorts) take place outside of the load utility and CA IDMS DLI Transparency. They require use of the user's native sort/merge facility.

The IDMSDLLD Steps 1, 2, 4, and 6 produce reports that show the results of the processing and a count of the records involved.