Is ILM simple or complex?

Click here to enlarge image

When you-know-who started banging the information life-cycle management (ILM) drum, I paid careful attention to the definition…and then the many definitions that followed. I started getting déjà vu all over again: ILM sure looked and smelled like hierarchical storage management (HSM).

(For an interesting take on the ILM-HSM issue, see “Is ILM HSM in disguise?”, InfoStor, January 2005, p. 22.)

HSM always made a lot of sense, but for whatever reasons it never took off as expected and as time went by, it gathered a negative connotation. So, in an effort to dress up ILM in something other than the emperor’s clothes (HSM), vendors started making it complex by weaving all kinds of other technologies and disciplines into the ILM motley coat.

Click here to enlarge image

That brings me to last month’s ILM Summit conference, a two-day event dedicated solely to ILM (www.ilmsummit.com). The presentations and sessions were great, but by the end of the event I had the impression that the vendor community was trying to scare users away from the ILM concept by making it so complex (which translates into “expensive”).

Of course, the conference was full of the main mantra: “ILM is not a product, it’s a process” (then what were the exhibitors trying to sell? Psssst, wanna buy a process?). Also, many experts were telling us that ILM is about more than storage. No kidding; it’s apparently about waaaay more than storage.

Click here to enlarge image

For example, at the conference I learned that for a full ILM implementation users will need to sew together asset management, security, workflow management, compliance, risk management, data movement/migration, and predictive analysis software. But, wait: There’s more! ILM solutions will have to interface with document management, e-mail management and archiving, database archiving, enterprise resource planning (ERP), enterprise content management (ECM), records management, storage resource management (SRM), data life-cycle management (DLM), content-addressable storage (CAS), data classification and, of course, HSM.

Whoa! I tried to put myself in an IT storage manager’s seat and started thinking about the enormous multi-vendor integration and interoperability hassles, the costs, and the (argh) need to work with dozens of other IT, and non-IT, departments.

I think the vendor community should keep ILM simple for the time being; otherwise, users are going to run, not walk, away from the concept.

Similarly, my advice to users is to keep it simple: I’d say start with something that makes sense regardless of ILM, and is doable, such as a tiered storage architecture.

Click here to enlarge image

For some insight into users’ views on, and plans for, ILM, see the accompanying pie charts, which show the results of a conference survey taken by the Gartner IT consulting firm late last year.

Dave Simpson

This article was originally published on March 01, 2005