Integration tables

In general, the tables are categorized by the prefixes and suffixes described in this table:

Table	Description
ATT_	A prefix indicating a table that contains cube fact data or dynamic attribute data in the format that can be imported directly into a cube. The third letter of the name typically indicates C for cube or D for dimension. For example, the BPCMAIN cube or BPDACCOUNT dimension.
DAT_	A prefix indicating a table containing staged or transformed data. These are not read by Designer when importing dimensions or cube data.
DIM_	A prefix indicating one of the sets of four dimension tables that is required to create a dimension.
INV_	A prefix indicating one of the sets of three inventory tables used by the data integration for Base Data Store to define and manage tables. These tables do not contain source system data.
_ATTRIBUTES	Suffix of one of the four dimension tables that are required to create a dimension. The tables contain attributes for dimension elements.
_ELEMENTS	One of the four dimension tables required to create a dimension. The tables contain the list of dimension elements.
_HIERARCHIES	One of the four dimension tables required to create a dimension. The tables contain the list of hierarchies in the dimension.
_PARENTS	One of the four dimension tables required to create a dimension. The tables contain parent/child relationships between elements.
_STAGED	A table (prefixed with DAT_) to contain the data from the source system that is to be imported. The data in these tables is read by stored procedures ending in _PREPARE_STAGING. These stored procedures validate the data and write it out to a separate copy of the table with a suffix of _TRANSFORMED.
_TRANSFORMED	A table (prefixed with DAT_) that contains validated staged data processed by stored procedures ending in _PREPARE_STAGING. This validated data is in turn processed by a second stored procedure to write it to the required dimension and cube tables. The _TRANSFORMED table may contain values marked by the _PREPARE_STAGING stored procedures as errors. These values are prefixed with _ERROR_ so they can be found with queries. Processing cannot continue until the _ERROR_ values are corrected.