Setting Up Chunking

This topic discusses how to:

Establish chunking in the engine metadata.
Set up chunking criteria.

Pages Used to Set Up Chunking

Page Name	Definition Name	Navigation	Usage
Engine Metadata - State Variables	PF_META_ENG_TBL2	select EPM Foundation, then select Job Processing, then select Setup Engines and Jobstreams, then select Engine Metadata, then select State Variables	Specify records to be chunked during the resolve process.
Engine Metadata - Source TableMaps	PF_META_ENG_TBL4	select EPM Foundation, then select Job Processing, then select Setup Engines and Jobstreams, then select Engine Metadata, then select Source TableMaps	Specify source tablemaps to be chunked during the resolve process.
Technical Scenario	PF_CHUNK_DFN	select EPM Foundation, then select Business Metadata, then select Business Framework, then select Technical Scenarios	Set up the object type values that the Resolver uses to chunk the record or tablemap you are resolving.

Engine Metadata Pages

Use the Engine Metadata - State Variables page (PF_META_ENG_TBL2) to specify records to be chunked during the resolve process.

Navigation

select EPM Foundation, then select Job Processing, then select Setup Engines and Jobstreams, then select Engine Metadata, then select State Variables

Use the Engine Metadata - Source TableMaps page (PF_META_ENG_TBL4 ) to specify source tablemaps to be chunked during the resolve process.

Navigation

select EPM Foundation, then select Job Processing, then select Setup Engines and Jobstreams, then select Engine Metadata, then select Source TableMaps

To set up chunking you access the Engine Metadata - State Variables or Engine Metadata - Source TableMaps pages to set up application engine parameters. On the State Variables page, you specify the records to be chunked during the resolve process. Alternatively, on the Source TableMaps page you can specify the tablemaps to be chunked. The Working with Engine Metadatatopic details the pages in the Engine Metadata component on which you define this setup.

As the next step, go to the Technical Scenarios page to set up the chunking criteria. Technical scenarios enable you to set up the object type values that the Resolver uses to chunk the records and tablemaps you are resolving. You establish the records or tablemaps to resolve on the Engine Metadata - State Variables and Engine Metadata - Source TableMaps pages.

Technical Scenario Page

Use the Technical Scenario page (PF_CHUNK_DFN) to set up the object type values that the Resolver uses to chunk the record or tablemap you are resolving.

Navigation

select EPM Foundation, then select Business Metadata, then select Business Framework, then select Technical Scenarios

Image: Technical Scenarios page

This example illustrates the fields and controls on the Technical Scenarios page. You can find definitions for the fields and controls later on this page.

Scenario ID

Select the scenario ID to which to link this technical scenario. When a jobstream runs for this scenario, the chunking is invoked.

Technical Scenario ID

Enter the technical scenario ID to identify the chunking selection. You may enter as many chunk codes as you like for each scenario. Each chunk code may have different chunk criteria.

Chunk By

Select the object type to use for chunking. You must select the object type you selected in the Engine Metadata component.

Note: You can apply only one chunking criteria to a tablemap. The system does not support duplicate object types.

Constraint Code

Select the constraint that filters the values for your objects at run time.

DataMap Column

Select one column from the datamap of the constraint. This should be, but is not restricted to, the column that matches the object type you have selected. For example, if your chunk object type is PRODUCT ID, then your datamap column from the constraint may be PRODUCT_ID or PRODUCT_TYPE. You may copy chunking criteria to another effective date for the same scenario only. You need to be very careful about defining chunk criteria. You need to make sure to define your chunks to cover a complete set of data without any duplicates.

Note: If you delete a scenario with chunking criteria, the chunking criteria are automatically deleted.

You can enter multiple technical scenarios to process data in multiple chunks. Remember, that all technical scenarios should select a mutually exclusive data set. The data is also not necessarily processed in the order defined.

Warning! If you enter an invalid or duplicate constraint code, the jobstream abends at run time.

You have established the chunk objects and the chunking criteria. Chunking initiates when there is a technical scenario associated with the run scenario.

For chunking to be successful, you must ensure that:

The object type selected in the Engine Metadata component matches the one in the technical scenario used by the engine. This initiates the chunking process.
The column you select for the chunking criteria on the Technical Scenarios page must be in the record or any record of a tablemap you have selected for chunking. This completes the chunking process.

Note: You can only resolve (chunk) a record once in an engine. The system resolves tablemaps first. Any records resolved as part of the tablemap are not resolved again in the state record.