Siebel Analytics Installation and Configuration Guide > Configuring Authentication in Analytics and Operational Applications >

Position and Organization Initialization Variables in Analytics Applications

The Siebel Analytics repository contains two preconfigured initialization blocks for Position-based and Organization-based security. These initialization blocks populate the variables PRIMARY_POSTN_ID and ORGANIZATION. The variable ORGANIZATION is a row-wise initialized variable that contains a list of the organizations that the user belongs to. (For more information on row-wise initialized variables, see Siebel Analytics Server Administration Guide.) You can set up a similar variable that is a union of all the positions and organizations that the user has access to. To set up such a variable, do one of the following:

• Use the W_PARTY_LOGIN table in the data warehouse.
• Get this variable directly from the Siebel transactional database for a real-time list.

Getting the variable directly from the Siebel transactional database requires an understanding of the Siebel data model with reference to security. For each user login, the W_PARTY_LOGIN table has a recursive list of all positions and organizations. The logical equivalent of this setup is:

W_PARTY_LOGIN.PARTY_ID = valueof(NQ_SESSION.POSTN_ORG_UNION)

W_OPTY_PARTY.PARTY_ID = W_PARTY_LOGIN.PARTY_ID

W_OPTY_PARTY.OPTY_ID = W_OPTY_ID.ROW_ID

In an alternate implementation, use an additional join between W_PARTY_LOGIN and W_[DimensionName]_PARTY instead of using the row-wise initialization. In the case of Opportunities, for example, this can be set up as shown:

W_PARTY_LOGIN.LOGIN = valueof(NQ_SESSION.USER)

W_OPTY_PARTY.PARTY_ID = W_PARTY_LOGIN.PARTY_ID

W_OPTY_PARTY.OPTY_ID = W_OPTY_ID.ROW_ID

These two joins have the same effect as the earlier row-wise initialization variable solution, but contain an additional table join. For more information about the Siebel data model and security, see Siebel Data Warehouse Data Model Reference and Siebel Data Warehouse Data Model Reference for Industry Applications.