At the beginning of an Information Lifecycle Management (ILM) project for my client’s Application and Data warehouse databases, my dialog begins with Records Management and the executive team to assess their ILM and Data Governance maturity.  These questions were briefly mentioned in my previous blog.  Here is some background on why the answers can dictate an ILM project’s success.

Are data retention schedules defined and are they assigned to a business owner?

Data targeted for ILM needs a business owner who is accountable and responsible for the data lifecycle – including defining when the data can be archived or deleted.  If data retention schedules do not exist or aren’t enforced, data volumes grow uncontrollably causing problems in the data center.  IT then owns the problem but isn’t able to address the solution unless business tells them what data can go where.  If data needs to be retained for longer periods of time, the business needs to provide IT with access requirements so they can properly design a database archive.

Do data destruction procedures assure proper sign off with an audit log?

If data retention schedules exist, what are the procedures for destroying data? If database data is considered a legal record, it needs to follow data destruction procedures defined by Records Management that incorporates all data.  Destruction of records in a database is much more involved than files or emails. The business needs to assure that when records are removed from a database only targeted rows and tables are removed without impacting the entire application. This requires intimate knowledge of the application data model.  Also, all copies of the data need to be removed including data in non-production and test databases, backups, and archives. IT needs to be able to identify where all copies of the data is in order to properly execute the purge process.

Is IT involved in defining the data destruction procedures?

IT and Records Management need to have direct communications on how data is to be destroyed or preserved such that IT can make sure the technology is configured properly.  Features such as digital shredding, evidence of authenticity, and the need to present data in the business context can significantly impact IT’s technology selection criteria.  If data needs to be retained with the ability to present it leveraging the original application, storing database data in an online archive may be the preferred architecture.

Are Data Objects defined in the application such that when data is tiered or purged from the database, referential integrity is maintained?

A Data Object is a table or set of tables and criteria that completely defines a business entity, transaction, or legal record.  It is the most critical component when it comes to implementing ILM in databases.  Vendors provide these definitions for packaged applications.  For custom applications, the business needs to work with developers and DBAs to document the Data Object definition.  During an archive or purge cycle, data are selected based on Data Object definitions ensuring data referential integrity is maintained, the application continues to function properly, and if archived, the archive contains business context required for future eDiscovery needs.

Julie Lockner, President & Founder, CentricInfoTM
www.CentricInfo.com