Show TOC

 Physical Structure of the Data in CO-PA

The following sections describe the detailed physical structure of the data (physical insertion sequence) and the typical accesses of the dataset.

In the section Physical Distribution of the Data in CO-PA , we will propose a way to distribute the data effectively over larger areas (table spaces, hard drives, striping, and so on).

Documents

Actual line items are stored in table CE1xxxx (where xxxx = operating concern). The system inserts into this table on an ongoing basis (sorted according to time stamp of the insertion point).

In drilldown reporting, cost center assessment, etc., the system accesses the data by this time stamp. Thus the physical insertion sequence corresponds to the typical read access method.

Plan line items are stored similarly in table CE2xxxx and are accessed in the same way. Consequently, this table will not be referred to separately in the following.

Segment Level

The segment level -- CE3xxxx -- contains the data in its first summarized form. The profitability segment number and the period are key fields (along with additional technical fields: plan version, record type, plan/actual indicator, etc.). The data part consists of the value fields.

The data in the segment level is typically sorted by period when it is created (INSERT) and updated during the period. That is, the insertion sequence depends on the random order in which the business transactions occurred and is only sorted roughly according to period (or period block PERBL).

The typical queries search in the segment level with a few (approximately 100) fixed segment numbers (WHERE PAOBJNR IN <List> ...) and a condition of time (... AND PERBL IN ...). Thus the physical insertion sequence does not correspond to the typical read access sequence! The segment level is always accessed via the primary index CE3xxxx__0.

Segment Table

The segment table CE4xxxx contains a unique profitability segment number (the primary key) for each combination of characteristic values (data part) to which data has been posted.

The records are inserted in the order of the profitability segment numbers (albeit only approximately, due to buffering of the segment number assignment on the application servers) and are practically never changed.

Selections from the segment table can generally be divided into two groups:

  1. Selections of large data volumes for drilldown reporting, planning or cost center assessment

  2. Here the system finds all the profitability segment numbers whose combination of characteristic values meets a specified condition (e.g. in a drilldown report). If a suitable index exists, the system uses it to access table CE4xxxx. In that case the access does not correspond to the physical insertion sequence. If no suitable index exists, the system selects using the primary index. To read about how you can improve performance for this type of access, see Physical Distribution of the Data in CO-PA .

  3. Determination of a profitability segment number based on the known combination of characteristics

In this case the system finds the profitability segment number when the values of all characteristics are known. To read about how you can improve performance for this type of access, see Index Support for Determination of Segment Numbers .