Applications are retired (sunset or decommissioned) when they become dormant or read-only. This occurs as a result of mergers and acquisitions or through modernization efforts and is a natural part of Application Information Lifecycle Management. While the applications may be no longer needed, the data they contain cannot be discarded. As a result, many organizations have hundreds, even thousands, of defunct applications that are consuming budget dollars, taking up data center space, complicating IT management, and generally just getting in the way. The challenge is getting rid of applications without getting rid of the data which is tightly coupled to them.
The application retirement process involves extracting data from its original source and loading it to a new more cost effective infrastructure “as-is”. To state this more boldly, it is important that every row in every table from the source database be an exact replica of the original. This is essential because data archived for compliance reasons is often retained for a future audit or legal use, not to mention ongoing business use.
The data must 100% accessible and intact. Auditors and legal eDiscovery requests for retired application data will require that data can be quickly retrieved for analysis. In order to use the retrieved data, it is essential that the owners of the retired data can prove that the data is in its original form and that no changes or modifications have occurred either during or after the archive process. With hundreds of applications to retire, each of which can include hundreds and thousands of tables which each can include hundreds of columns and millions of rows, the verification step is a daunting task. In fact, the verification step is more onerous than the archive step itself.
With the volumes that are involved, testing manually by business users is not sufficient nor cost effective. To meet the audit and legal requirement, more comprehensive testing is required. An effective application retirement verification process performs 100% testing of the values in every column and every row between the original source and the new archive destination. This validation insures that the data is intact and is an essential step before beginning the application decommissioning process.
These best practices are imbedded in Informatica’s approach to data archive. Value comparisons are simple and easy to analyze with an option for random sampling of high volume tables. Verification reports are produced which can be used as proof that the data was not transformed during the archive process and does indeed match the original source exactly. When the Informatica ILM solution is combined with Lean Integration, the result in an Application Retirement Factory. Stay tuned for future blog postings about how to effectively implement a factory approach.