Use of BI Master Data and Master Data-Bearing Characteristics

Definition

Master data is data that remains unchanged over a long period of time. Master data contains information that is always needed in the same way. Characteristics can bear master data in BW. With master data you are dealing with attributes, texts or hierarchies.

If characteristics have attributes, texts, or hierarchies at their disposal then we speak of master data-bearing characteristics.

The master data of a cost center contains the name, the person responsible, the relevant hierarchy area, and so on.

The master data of a supplier contains the name, address, and bank details.

The master data of a user in the SAP system contains his/her access authorizations to the system, the standard printer, start transaction, and so on.

Use

When you create a characteristic InfoObject, it is possible to assign attributes, texts, hierarchies, or a combination of this master data to the characteristic. If a characteristic bears master data, you can edit it in BW in master data maintenance.

See also Creating InfoObjects: Characteristics

Master data for a characteristic is uploaded into BW using an InfoSource with direct updating.

See also InfoSources with Direct Updating.

You can also update characteristics with master data flexibly. In this case the InfoObject becomes the data target or InfoProvider. However, you must explicitly declare the InfoObject in the InfoObject maintenance to be the InfoProvider.

See also InfoSources with Flexible Updating.

BI Master Data Types: Attributes, Texts and Hierarchies

Use

There are three different types of master data in BI:

1. Attributes

Attributes are InfoObjects that are logically subordinate to a characteristic. You cannot select attributes in the query.

You assign the attributes Person responsible for the cost center and Telephone number of the person responsible for the cost center (characteristics as attributes), as well as Size of the cost center in square meters (key figure as attribute) to a Cost Center.

2. Texts

You can create text descriptions for master data or load text descriptions for master data into BI. Texts are stored in a text table.

In the text table, the Name of the person responsible for the cost center is assigned to the master data Person responsible for the cost center.

3. Hierarchies

A hierarchy serves as a context and structure for a characteristic according to individual sort criteria. For more detailed information, see Hierarchies.

Features

Time-dependent attributes:

If the characteristic has at least one time-dependent attribute, a time interval is specified for this attribute. Since the time frame for master data on the database must always be between 01.01.1000 and 12.31.1000, the gaps are filled automatically (see Maintaining Time-Dependent Master Data).

Time-dependent texts:

If you create time-dependent texts, the system always displays the text for the key date in the query.

Time-dependent texts and attributes:

If texts and attributes are time dependent, the time intervals do not have to agree.

Language-dependent texts:

In Characteristic InfoObject Maintenance, you specify whether texts are language specific (for example, with product names: German ® Auto, English ® car) or are not language specific (for example, customer names). The system only displays texts in the selected language.

If texts are language dependent, you have to load all texts with a language indicator.

Only texts exist:

You can also create texts only for a characteristic, without maintaining attributes. When you load texts, the system automatically generates the entries in the SID table.

BI Master Data Maintenance

Use

Master data maintenance allows you to change or regenerate master data attributes or texts manually in BW. Data is always maintained per characteristic.

There are two different master data maintenance sessions:

· Creating or Changing Master Data

· Deleting Master Data at Single Record Level

Integration

You cannot run the two sessions at the same time. This means that

· if you choose the Change function in the master data maintenance screen, the deletion function is deactivated, and is only reactivated once you have saved your changes.

· if you select a master data record in the master data maintenance screen and choose the Delete function, the create and change functions are deactivated, and are only reactivated once you have finished deleting the record and clicked Save.

Functions

Creating or Changing Master Data:

You can add new master data records to a characteristic, change individual master data records, or select several master data records and assign global changes to them.

Deleting Master Data at Single Record Level:

You can delete individual records or select and delete several records.

You can only delete master data records if no transaction data exists for the master data that you want to delete, the master data is not used as attributes for an InfoObject, and there are no hierarchies for this master data.

Creating and Changing BI Master Data

Prerequisites

If master data is maintained for a master data-bearing characteristic, you can change it in BW or create additional master data records.

Procedure

1. You get to the master data maintenance via

InfoObject Tree ® InfoObject ® Context Menu (right mouse button) ® Maintain Master Data.

or

InfoSource Tree ® Application Component ® Master Data InfoSource ® Context Menu (right mouse button) ® Maintain Master Data.

or in the InfoObject maintenance by choosing Maintain Master Data

A selection screen appears for restricting the master data. You may want to edit here.

2. Taking into account the options from F4 Help, select the relevant data.

You get to the list header for the selection. The list header is also displayed if no hits have been found for your selection, so that you can enter new master records for particular criteria.

3. Make your changes with the help of the relevant maintenance function.

Creating new master records

Choose Create to add new master records. New records are tagged onto the end of the list.

Changing single records

A double-click on a data record takes you to the individual maintenance. Make the relevant changes in the following Change dialog box.

Mass changes

Select several records, and choose Change. The Change dialog box appears in which the attributes and also texts are offered. Enter the relevant entries. These are then transferred to all the selected records.

4. Choose Save.

If a newly created record already exists in the database but does not appear in the processing list (because you have not selected it in the selection screen) there is no check. Instead, the old records are overwritten.

If you change master data in BW, you must adjust the respective source system accordingly. Otherwise the changes will be overwritten in BW the next time you upload data.

Master data, which you have created in BW, remains even after you have uploaded data from the source system.

Note the exception for time-dependent master data.

Maintaining Time-Dependent BI Master Data

Use

The maintenance of the master data is more complex with time-dependent master data, as the validity period of a text is not necessarily in concordance with that of an attribute master record.

The InfoObject User master record has the time-dependent attribute Personnel number, and the time-dependent text User name. If the user name changes (after marriage, for example), the personal number still remains the same.

Prerequisites

In the InfoObject Maintenance, make sure that the relevant InfoObject is flagged as ‘time-dependent’.

Procedure

To maintain texts with time-dependent master data, proceed as follows:

...

1. Select the master data that want, and select one of the three text pushbuttons.

If you choose Display text, a list appears containing all the texts for this characteristic value. By double clicking, you can select a text. A dialog box appears with the selected text for the characteristic value.

If you choose Change text, a list appears containing all the texts for this characteristic value. By double clicking, you can select a text. A dialog box appears with the selected text for the characteristic value, which you can then edit.

If you choose Create text, a dialog box appears in which you can enter a new text for the characteristic value.

The texts always refer to the selected characteristic value.

2. Choose Save.

When you select time-dependent master data with attributes, the list displays the texts that are valid until the end of the validity period of the characteristic value. When you change and enter new texts, the lists are updated.

Master data must exist between the period of 01.01.1000 and 12.31.1000 in the database. When you create data, gaps are automatically filled. When you change or initially create master data, in some cases, you must adjust the validity periods of the adjoining records accordingly.

If a newly created record already exists in the database but does not appear in the processing list (because you have not selected it in the selection screen) there is no check. Instead, the old records are overwritten.

If you change master data in BW, you must adjust the respective source system accordingly. Otherwise the changes will be overwritten in BW the next time you upload data.

Master data, which you have created in BW, remains even after you have uploaded data from the source system.

Time-Dependent BI Master Data from Different Systems

Use

You have the option of uploading time-dependent characteristic attributes from different systems, even if the time intervals of the attributes are different.

Functions

If you load time-dependent characteristic attributes from different source systems, these are written in the master data table, even if the time intervals are different.

From source system 1, load attribute A with the values 10, 20, 30 and 40. From source system B, load attribute B with the values 15, 25, 35 and 45. The time intervals of the last two values are different.

The system inserts another row into the master data table:

dateto	datefrom	Person Responsible	Cost Center
01.01.1999	28.02.1999	Mrs Steward	Vehicles
01.03.1999	31.05.1999	Mr Major	Accessories
01.06.1999	31.08.1999	Mr Calf	Light bulbs
01.09.1999	10.09.1999	Mrs Smith	Light bulbs
11.09.1999	30.09.1999	Mrs Smith	Pumps

Deleting BI Master Data at Single Record Level

Use

Besides creating and changing master data you can also use a deletion mode at single record level.

Procedure

...

1. You get to the deletion mode from the master data maintenance by choosing

InfoObject Tree ® InfoObject ® Context Menu (right mouse button) ® Maintain Master Data.

or

InfoSource Tree ® Application Component ® Master Data InfoSource ® Context Menu (right mouse button) ® Maintain Master Data.

or from the InfoObject maintenance by choosing Maintain Master Data.

A selection screen that you might want to edit appears for restricting the master data.

2. Taking into account the options from input help, select the relevant data.

3. You get to the list overview for the selection and have two options:

¡ In the list select the master data records to be deleted, choose Selete and Save your entries.

¡ First select additional master data using Data Selection, highlight the master data records that are to be deleted, and choose Delete. Repeat the selection as necessary and choose Save to finish.

The records marked for deletion are first written into the deletion buffer. If you choose Save, the system generates a where-used list for the records marked for deletion. Master data that is no longer being used in other objects is deleted.

Deleting BI Master Data and Texts for a Characteristic

Use

You can delete master data and texts directly from the master data table in BW. In contrast to deleting at single record level, you can use this function to delete all the existing master data and texts for a characteristic in one action.

Prerequisites

In order to delete master data there must be no transaction data in BW for the master data in question, it must not be used as an attribute for InfoObjects and there must not be any hierarchies for this master data.

Functions

You reach the Delete Master Data function from the context menu of your InfoObject in the InfoObject tree and also the InfoSource tree.

If you choose the Delete Master Data function, the program checks the entries in the master data table affected to see if they are used in other objects.

When you delete you are able to choose whether entries in the SID table of a characteristic are to be retained or whether they are to be deleted:

If you delete the SID table entry for a particular characteristic value, the SID value assigned to the characteristic value is lost. If you load new attributes for this characteristic value later, a new SID value has to be created for the characteristic value. In general this has a negative effect on the runtime required for loading. In some cases deleting entries from the SID table can also lead to serious data inconsistencies. This occurs if the list of SID values generated from the where-used list is not comprehensive, however, this is rare.

Delete, retaining SIDs

For the reasons given above, you should choose this option as standard. Even if, for example, you want to make sure that individual attributes of the characteristic that are no longer needed are deleted before you load master data attributes or texts, the option of deleting master data but retaining the entries from the SID table is also absolutely adequate.

Delete with SIDs

Note that deleting entries from the SID table is only necessary, or useful, in exceptional cases. Deleting entries from the SID table does make sense if, for example, the composition of the characteristic key is fundamentally changed and you want to swap a large record of characteristic values with a new record with new key values.

Versioning BI Master Data

Attributes and hierarchies are available in two versions, an active (A) version and a modified (M) version. Texts are active immediately after they have been loaded. Existing texts are overwritten when new texts are loaded.

Attribute versions are managed in the P table and in the Q table. Time-independent attributes are stored in the P table and time-dependent attributes are stored in the Q table. From left to right, the P table contains the key fields of the characteristic (for example, 0COSTCENTER: CO_AREA and COSTCENTER), the technical key field OBJVERS (versioning), the indicator field CHANGED ID (versioning), and 0 or more attribute fields that can be display attributes or navigation attributes. The structure of the Q table is identical to the structure of the P table, with the addition of the 0DATEFROM and 0DATETO fields to map the time-dependency.

The OBJVERS and CHANGED fields must always be taken into account in versioning:

If you load master data that does not yet exist, an active version of this data is added to the table. If, when you reload the data, the value of the attribute changes, the active entry is flagged for deletion (CHANGED = D) and the M/I (modified(insert)) version of the new record is added.

You are loading master data for the 0COSTCENTER characteristic. After activation, the P table looks like this:

Later, you load new records. These new records are given the OBJVERS entry M and the CHANGED entry I. The available records, for which new data has been loaded, are given the OBJVERS entry D for “to be deleted”:

Before the new records can be displayed in reporting, you have to start the change run (see System Response Upon Changes to Data: Aggregate). During the change run, the old record is deleted and the new record is set to active.

BI reporting always reads the active version. InfoSets are an exception to this rule, as the most recent reporting function can be switched on in the InfoSet Builder. In such an InfoSet, the most recent records are displayed in reporting, even if they are not yet active.

For more information, see Most Recent Reporting for InfoObjects.

Activating BI Master Data and Texts

Prerequisites

Master data and texts have already been loaded into the BI system using the scheduler.

Procedure

Activating Master Data

When you update master data from an SAP system the master data is imported in an inactive state. You must activate the new master data so that it can be accessed and used for reporting purposes.

For more information, see Versioning Master Data.

Choose the path InfoObject Tree ® Context Menu of Corresponding Characteristic ® Activate Master Data.

Upon Activation there are two scenarios to choose from:

...

The Corresponding Master Data Is Already Being Used in Aggregates in the InfoCube:

If you are already using the available master data in aggregates in InfoCubes, you cannot activate the master data individually. In this case, proceed as follows:

1. In the main menu, choose the path Tools ®Hierarchy/Attribute Change.

2. Proceed as described in System Response upon Changes to Data: Aggregate.

The system now automatically restructures and activates the master data and its aggregates.

Please note that this process can take several hours if the volume of data is relatively high. Therefore, you should simultaneously activate all of the characteristics that are affected by changes to their master data, at regular intervals.

The Corresponding Master Data Is Not Used in Aggregates:

Choose the path InfoObject Tree ® Context Menu of Corresponding Characteristic ® Activate.

The system now automatically activates the master data so that it can be used directly in reporting.

Activating Texts

Texts are active immediately and can be used directly in reporting. You do not need to Activate manually.

Simulate Loading of BI Master Data

Use

This function allows you to simulate the loading of a master data packet before loading the data into BW. This means you can be aware of errors in the data loading early on and remove problems in advance.

Integration

You can get to the function from the Monitor Detailed Screen ® Tabstrip Updated Data or All IDocs.

Features

In the case of data without errors, the loading simulation provides you with a detailed description of the processes that are run during loading. The left-hand frame structures the various master data types that can be loaded in a tree. Either:

• Time-dependent texts and/or time-constant texts, or

• Time-dependent master data attributes and/or time-constant texts

On the level below the master data types, you will see the different database operations that are carried out during loading. (For example, modifying, inserting, deleting)

By clicking on a master data type or on a database operation, or by using drag-and-drop on these objects in the right-hand frame, you can obtain a detailed view of the respective uploaded data.

In the case of incorrect data, only the master data types, and not the database operations, are displayed in the left-hand frame.

The corresponding error log appears in the lower frame.

BI Master Data Lock

Use

During the master data load procedure, the master data tables concerned are locked so that, for example, data cannot be loaded at the same time from different source systems, which would bring about inconsistencies.

In certain cases, for example if a program termination occurs during the load process, then the locks are not automatically removed after the load process.

You then have to manually delete the master data locks.

Activities

You get to the master data lock overview via the padlock symbol. Using the context menu (right mouse button), choose the corresponding master data symbol, and delete the master data lock.

Reorganizing BI Master Data

Use

You can reorganize the dataset for texts and attributes belonging to a basic characteristic. The reorganization process finds and removes redundant data records in the attribute- and text- tables. This reduces the volume of data and improves performance.

Functions

For a given basic characteristic, the system firstly compares data in the active and modified versions of the time-dependent and non-time-dependent attributes with each other. If there are no differences between the active and the modified versions, the redundant data is compressed. In a second step, the system checks time-dependent texts and attributes to see whether time intervals exist with identical attribute values or text entries. If this is the case, the affected time intervals are combined into larger intervals.

Firstly, the attribute Cost Center Manager (0RES_PERSON) is changed, as the only attribute, for a Cost Center and then is reset to its original value using a second load process. Therefore, the name of the Cost Center Manager has not actually changed. In this case, the reorganization means that the data record is deleted for the changed version (M version).

For a Cost Center, the same person is entered as Cost Center Manager for the period 01.06.2001-31.12.2001 as for the period 01.01.2002 – 31.03.2002. The process of reorganization combines these two intervals into one, providing that the other time-dependent attributes for the cost center are consistent across both intervals.

You can carry out the master data reorganization process as a process type in process chain maintenance.

Activities

During master data organization for attributes and texts, the system sets locks preventing access to the basic characteristic currently being processed. These locks correspond to the locks preventing the loading of the master data attributes and texts. This means that it is not possible to load, delete or change master data for this characteristic during the reorganization process.

When assigning locks, the system distinguishes between locks for attributes and locks for texts. This means that you can load texts for this characteristic during a reorganization that only affects attributes, and vice versa.

Load Master Data to InfoProviders Straight from Source Systems

In data transfer process (DTP) maintenance, you can specify that data is not extracted from the PSA of the DataSource but is requested straight from the data source at DTP runtime. The Do not extract from PSA but allow direct access to data source indicator is displayed for the Full extraction mode if the source of the DTP is a DataSource. We recommend that you only use this indicator for small datasets; small sets of master data, in particular.

Extraction is based on synchronous direct access to the DataSource. The data is not displayed in a query, as is usual with direct access, but is updated straight to a data target without being saved in the PSA.

Dependencies

If you set this indicator, you do not require an InfoPackage to extract data from the source.

Note that if you are extracting data from a file source system, the data is available on the application server.

Using the Direct Access mode for extraction has the following implications, especially for SAP source systems (SAPI extraction):

● Data is extracted synchronously. This places a particular demand on the main memory, especially in the source system.

● The SAPI extractors may respond differently than during asynchronous load since they receive information by direct access.

● SAPI customer enhancements are not processed. Fields that have been added using the append technology of the DataSource remain empty. The exits RSAP0001, exit_saplrsap_001, exit_saplrsap_002, exit_saplrsap_004 do not run.

● If errors occur during processing in BI, you have to extract the data again since the PSA is not available as a buffer. This means that deltas are not possible.

● In the DTP, the filter only contains fields that the DataSource allows as selection fields. With an intermediary PSA, you can filter by any field in the DTP.