At its core, data archiving involves the systematic removal of completed business process data from the live database to compressed archive files. This process is governed by , which logically group related data tables—such as header and line items—to ensure business integrity is maintained during extraction.
Check the manifest file:
The goal of pre-processing is to identify and mark data that is ready for archiving. This is often done by setting a deletion indicator. bit660 data archiving pdf 23
The central element (defined in T-code AOBJ) that determines which data is archived (tables) and which programs are used.
Aligning archiving schedules with business legal retention policies. At its core, data archiving involves the systematic
Run archiving jobs periodically (e.g., monthly or quarterly) rather than a massive, risky cleanup once every five years.
Rules that tell the system exactly which data to pick (e.g., "all orders from 2018"). 4. The Two-Step Dance (The Archiving Process) This is often done by setting a deletion indicator
For an archiving project to be successful and sustainable, following established best practices is essential. This includes:
A crucial final check, as highlighted in the PDF, is to verify the Archive Information System. After an archiving run, you should see a green status light indicating the info structures have been updated successfully. This step is not optional—if you skip this check, users might not be able to read archived documents later, leading to support tickets and frustrated business teams.
For example, one version of this PDF uses a real-world example with a Materials Management (MM) object—specifically, a purchase order (MM_EKKO)—to walk a user through the entire archiving process. It is designed to be "applicable equally to any PP object (PR_ORDER-Process order)", meaning the steps are general enough to be adapted for various SAP modules like Finance (FI), Sales and Distribution (SD), and Controlling (CO).
This is the cleanup and final verification stage. Some archiving objects have secondary data like indexes or extra logs that need to be cleaned up after the main data is gone. The post-processing job handles this automatically.