IKM When Staging Area is on the Target Server

1. Execute Set-Oriented SELECT Statement:
• The IKM begins by executing a single set-oriented SELECT statement.
• This statement applies the staging area and target declarative rules to all C$ tables and local tables (e.g., D in the figure).
• This step generates a result set containing the transformed data.
2. Write Data to Target Table (Simple IKMs):
• For simple "append" IKMs, the result set generated in the previous step is directly written into the target table.
3. Create Temporary "I$" Table (Complex IKMs):
• For more complex IKMs, an "I$" temporary table is created to store the result set instead of directly writing it to the target table.
4. Check Data Against Target Constraints (Optional):
• If necessary, the IKM will call a CKM (Check Knowledge Module) to check the data flow against target constraints.
• The CKM isolates any erroneous records and cleanses the "I$" table by removing or correcting invalid data.
5. Write Data to Target Table:
• The IKM writes the final records from either the "I$" table (if used) or the result set directly into the target table, following the defined strategy (e.g., incremental update, slowly changing dimension, etc.).
6. Drop Temporary "I$" Table:
• After writing the data, the IKM drops the "I$" temporary table to free up resources.
7. Optional Final Consistency Check:
• Optionally, the IKM can call the CKM again to ensure the consistency of the target datastore after the data load.

Key Characteristics:

• These types of IKMs do not manipulate data outside the target server.
• The data processing is set-oriented, which maximizes efficiency when handling large volumes of data.