Purge guidelines

These purge guideline topics include design considerations and an approach for managing the archiving and purge of expired transaction data from the FTM database. A purge function is not included as a core part of FTM because each FTM customer application is likely to have differing requirements. Most customers want to purge the expired data from the live database.

Some customers want to move the expired data to an archive or data warehouse. The data to be transferred to a warehouse might be only a synopsis or subset of the live data. In other cases, the important history data might be archived by the systems that the FTM application interacts with, so the FTM data might not need to be archived.

Typically, a customer wants to maintain the live data in the database for a few days. This time period can vary from a shorter range of 3 - 7 days to a longer period in the range of 30 - 90 days. The expiration criteria is usually based on a combination of the business objectives and the volume of data.

When data is purged, the process must be reliable and must be done in a manner that maintains the referential integrity of the data. In addition, the purge of large numbers of records might have performance implications. Because the purge and archive processes normally involve intense database activity, consider the timing of when the process is run to minimize any disruption to normal FTM processing.

The following topics describe only the considerations and methods to account for when a purge process is designed. The same considerations and methods can be applied to designing an archive process.