An important challenge for SAP administrators is to find out the appropriate data for archiving. It is important to have knowledge of the business processes involved in the enterprise, prior to data archiving, so that appropriate archive objects are created. This helps in identifying database table growth and simplifies the process of identifying objects for archiving.
Three Phase Archiving Process
Data archiving process comprises three major phases. They are:
Creating an archive file
Removing the archived data from the database
Transferring the archived files to a location outside the SAP database
Phase I - Creating Archive Files
The archived files of data are created in the SAP database by the Archiving Management system. The Management system reads the data from the database and writes it to the archive files in the background. In instances of archived files exceeding the maximum specified limit, or if the number of data objects exceed the stipulated limit in the system, then the system automatically creates new archive files.
At the end of the process of saving data into archive files, ADK triggers the system event SAP_ARCHIVING_WRITE_FINISHED, which is an indicator to the system to start next phase of archiving process.
Phase II - Deleting files from the Database
While archiving management system writes data on the archive files, another program deletes it from the database permanently. The program checks whether the data has been transferred to the archive. It is quite important too as it is the last check performed by the system before deleting data permanently from the database. Several deletion programs run simultaneously, because the archiving program is much more faster than the deletion programs. This is important as it increases the efficiency of archiving process.
Phase III - Transferring Archive Files
Once the Archive management system has finished archiving the data, the next step is to save the archived files at a different location other than the SAP database. This can be accomplished by an automated process in the system or by a manual process. This step is optional since many enterprises may wish to keep the archived files within the current database. However, large enterprises transfer their data periodically as a part of their data archiving processes.