Archival Best Practices
This article presents the best practices for Archival Activity.
Insurance policy administration systems necessarily generate large volumes of data. As such, archiving old data that is not needed for daily operations can be a critical method for maintaining the performance and usability of the system.
Data archiving means moving data out of the operational database, the data present is is never left in an inconsistent state, either during a successful archival run or during a run where there is an error scenario. To reduce the possibility of inconsistent data due to any unforeseen error scenarios, the following recommendations are made:
- It is recommended to familiarize with the data changes that happen during the archival process and have a proper understanding of the changes happening in the OIPA database when an entity record is archived and should inform the operations team what to expect both in the database and in the OIPA user interface.
- It is recommended to test a small dataset with the archival request parameters, to select a smaller number of entities to archive so the process completes within an acceptable time window.
- Performance testing is critical, especially in the case of an initial archival run where there may be many years worth of data to archive.
- With the configurable nature of OIPA, users might have to adjust indexes to maximize the performance in their instance. After the initial run, assuming the archival process is run on a periodic basis, this should be less of an issue due to the lesser volume of records to be archived during any one archival run.
- It is recommended to initiate a full backup, (especially prior to the initial run) for the archival process on a production database.
- An archival run should not be performed while the system is otherwise available to eliminate the possibility of resource contention during the archival process.