Zombies are taking over your IT department. Not the shuffling flesh-eating kind. Something far worse. Zombie applications. They look like their alive but they are not. They used to be living breathing systems but now they are the undead of your application portfolios.
Every IT department has a list as long as your arm of applications that are no longer used but can’t be deleted because they contain data that has to be retained. They spin there on your disks taking up space and eating the brains of the developers maintaining them. These applications are a drain on time, attention,capacity and other scarce IT resource but there is a way to kill them. Application decommissioning with xDB. (warning – shameless plug for EMC stuff follows)
I spent last week in Pleasanton learning about an approach for application decommissioning from Jeroen van Rotterdam, the new Chief Architect for IIG. The idea is to identify the apps in your portfolio that are no longer creating new data but are still needed for some period of time for compliance reasons. You take the data out of the app using any appropriate extraction method, transform it into XML and store it in xDB.
Using special Documentum Delivery Services templates you can quickly create forms that let you search, apply simple date based retention or legal holds. You store with the data a chain of custody that records the extraction and transformation validation. With these basics in place you can finally put an end to the undead app and free up those resources supporting legacy portfolios.
It is a straight forward value proposition. You can move off these legacy apps quickly rather than creating complex data models and implementing an enterprise wide master data management programs that take years to implement. There is nothing preventing an MDM strategy. In fact this is a great way to get started. The goal however is to turn off the old code as quickly as possible while still maintaining access to the data.
By extracting the data, persisting it in a vendor neutral self describing structure and providing a completely open standards access method (XQuery) you can begin to shorten the list of unused legacies that are bogging your IT organization down.
The approach has two qualities that I think make it especially appealing.
- The Problem is pervasive – every IT shop has it. Mergers and acquisitions, ERP implementations and just the passage of time all generate apps in need of retirement.
- The Solution addresses the simplest form of the problem – xDB Application Decommissioning is not intended as formal records management or even really for unstructured data in file systems. EMC already has apps for that. (Documentum Records Manager & Kazeon). This is aimed at a very specific business problem of structured data that lives in aging systems in need of retirement that needs to be retained.
I found using xDB to manage retired structured data to be a very effective use of the technology. I should note that there a more complex competitive solutions are on the market offering conceptually similar capabilities. Tolerating variable schemas and searching across disparate data sets though can be very difficult in other approaches and focusing on “just the basics” of data retirement allows for a more rapid and cost effective implementation relative to other products.
If you want to see it or chat further ping your friendly neighborhood EMC IIG rep or drop me a line and I’ll connect you to the right people.
Doesn’t EMC already have a sales pitch for Application decommissioning. I believe PS has actually done some work in this area as well, though the idea to use xDB is NICE.
I am beginning to prove to clients the wonders of xDB/xHive. As said before I like the responsiveness of searches in xDB.
I was saddedned last year when one of the EMC folks was complaining that response to xDB in the market was not very good. I feel most customers don’t understand the use cases for using the product.
With cases like you have provided maybe xDB usage will appreciate.