Ensuring Data Integrity

It is critical to the success of an IBP implementation that data integrity is protected. IBP has

many built-in checks. Additionally, IBP can be configured to do further checks. During a data

load, when CPI is finished processing, IBP performs these checks.

If your source system is a single system such as SAP ERP, much of the most critical data

integrity is built in. Data integrity checks become most critical when data is received from

external sources, since the external systems may not be 100% in sync with your internal

systems. Even if all data is received from a single SAP ECC system, there can be records with

integrity errors due to timing differences between master data loads and transactional data

loads. If IBP is not configured to perform these integrity checks, CPI-DS should be used to find

errors and generate error reports.

When data integrity checks do not occur, faulty data can be loaded, aggregated values many

not equal the more granular data, and plans generated may be inaccurate. The CPI-DS

configuration specialist needs to understand what integrity checks are built into the IBP

system to deliver the best possible data to IBP. In addition, knowledge of these rules

facilitates the loading of data into IBP in the correct sequence.

Additional integrity checks can be built into the IBP data model. Attribute checks can be

added to check for:

● Attributes and keys of a simple master data type can be checked against another master

data type (this is similar to the concept of check table in ABAP) . For example: The Base UOM (UOMID) on product could be checked against UOMTOID on UOMTO and PRODUCTIONSOURCE simple master data should have its keys

checked against LOCATION (for LOCID) and PRODUCT (for PRDID).

● Attributes of a compound master data type can be checked against another master data

type. For example: The LOCATIONPRODUCT compound master data type should have its

subnetwork ID (PLUNITID) checked against PLUNITID on PLANNINGUNIT.

In addition, data loads are always checked for:

● Missing values and duplicate records

● Missing values occur when a key figure load does NOT have all its base planning level roots

populated

● Duplicate records occur when two records in a load have the same planning level root combination

- For example: Key figure ACTUALSTOCKTRANS is being loaded and its planning level is MTHPRDCUST. 

If two records in a single data load are updating CUSTID = ABC, PRDID = 123, and DATE = 09/30/2018, 

both records are reported as errors.

● When records fail these tests, the records do not update IBP.

Example to configure Attribute check

In this example, we are loading a compound master data for LOCATIONPRODUCT. The IBP

configuration specialist has added a check to ensure that the PLUNITIDs (planning unit) on the

LOCATIONPRODUCT records exist in IBP before loading. This check must be configured

because PLUNITID is not a key to LOCATIONPRODUCT. Because of this configuration, record

3 is rejected by the IBP Data Integration Job.

To configure this check, the master data type must be changed. For example, to implement

the PLUNIT check, you must perform the following steps:

1. Go to LOCATIONPRODUCT master data type configuration editing.

2. Select Attribute Checks .

3. Turn on Attribute Checks .

4. Select New Attribute Check

5. Specify that PLUITID should be checked against attribute PLUNIT on the PLANNINGUNIT

master data type.

You can see that the attribute check is active.

Example to configure Key Figure Check

In this example, we are loading key figure data. 

The IBP configuration specialist has added a check to ensure that a key figure cannot be loaded 

if the product location combination is not already in the LOCATIONPRODUCT master in IBP before loading. 

Because of this configuration, record 3 is rejected by the IBP Data Integration Job.

The data integration log reports:

Mandatory non-root "MATTRLOCATIONPRODUCT" in source MD

T##LOCATIONPRODUCT root WKPRODLOC PLEVEL "LOCID";"PRDID" not found.

The steps to implement a key figure load check against compound master data includes creating an attribute, 

adjusting the planning area, and adjusting the planning levels containing the compound master data type. 

To implement a key figure load check against the LOCATIONPRODUCT, 

the following steps are required using LOCATIONPRODUCT as an example:

1. Create attribute MATTRLOCATIONPRODUCT.

● Description: Mandatory Attribute Location Product

● Type: NVARCHAR

● Length: 1

2. Add MATTRLOCATIONPRODUCT to the T##LOCATIONPRODUCT master data type and activate.

3. Add MATTRLOCATIONPRODUCT to the planning area. Mark the attribute as having an

attribute category of Required.

A key configuration step is the setting of the attribute’s category on the planning area configuration tab. 

The attribute category specifies whether master data has to exist for the attribute when new planning records are added in IBP using either the Excel add-in or during data integration. 

By default, all attributes have the category NULL (optional). When optional, a check is not carried out.

4. Add MATTRLOCATIONPRODUCT to planning levels that use product and location as

shown in the following figure.

5. Activate the planning area.