There's a saying in IT that no good deed goes unpunished.
I guess that applies to IT vendors as well. As I cruise various industry sites, there seems to be an occasional theme of negativism around the whole idea of ILM -- information lifecycle management.
For those who make their living writing about controversy, you'd think this is a Big Issue. Examples can be seen here and a particularly vitriolic attack here.
"It's not well defined". "There's no agreed standard". "It's way too complicated". And, my personal favorite, "it's just a conspiracy for vendors to sell more stuff".
Well (other than the conspiracy theory), the first few observations are probably justified. The industry hasn't done a good job of defining what it is, and what it isn't. Relevant standards are perhaps years away. And, like any big IT challenge, it can look pretty complicated at the outset.
But that's the fault of the people who are communicating it, and not indicative of the idea itself.
Let's take two steps back
I think its useful to step back a few years, and remember that ILM was borne of the realization by users and vendors alike that not all information is created equally. Different information has different value at different times in its lifecycle.
Understand that value, and you might have the potential to lower costs and avoid certain kinds of risks. Lower costs by storing, protecting it and managing it the right way, commesurate with its value. And avoid risks by making sure it's protected and secured appropriately. And if you get real good at it, you might even find a way to create new value from the information you already own.
Not a panacea, and isn't trying to be
Is ILM for everyone? No, that's for sure. I've encountered businesses large and small where ILM shouldn't be on their list of top initiatives. Or, in some cases, they've got some *very* serious problems to attend to first.
But, in more cases than not, we see customers who are literally getting buried in information. Simply throwing more hardware at storing it is a nice luxury, but doesn't address the root cause, and won't relieve IT of the responsibility of understanding what the information is, and how it should be managed.
When we laid out our views on ILM several years ago, we broke the process into three phases that seemed to be a natural evolution.
- Create a tiered services catalog for information.
Go through a process of creating "service level buckets" for different kinds of information. Don't create too many buckets, you can always add more later. Specify performance, availability, recoverability and retention. That's your catalog.
Work with vendors to come up with a price tag for managing each class of information. Use this priced catalog to discuss with the business exactly how they'd like their information managed.
Along the way, make sure you've got the right technology to deliver the service levels, and are using ITIL processes and tools to manage your storage environment.
OK, I can sketch it out in a few short sentences, but in large environments, we're talking about a pretty big project. But it delivers value, prepares you for future steps, and really ought to be done anyway.
When we started doing this with customers, some of the cost savings were eye-popping. Not only did they reduce storage costs, they cut storage growth dramatically, and in some cases got order-of-magnitude improvements in operational efficiency. But it was a journey, not an event.
- Tackle a few big applications next
Email is a logical starting point for most people, with large file systems and databases not far behind. Define your policy for how service levels will change over time for information within the application (e.g. when does email get archived, how long retained, etc.). Select application technology that does what you need for application -- but -- remember that you'll probably want a unified environment in the next step.
The hard part here isn't the technology -- it's the policy. A few years back, we'd be talking about email archiving and get hip deep into a technology discussion, and I'd ask "well, what's your email retention policy?" and the conversation would kind of stall. The heavy lifting here is getting consensus for how service levels should change over time for email, files, database information, etc. etc.
One of the things we did was create a portfolio of policies that other businesses were using, and that seemed to speed up the conversation a bit. At least we had a starting point for the discussion.
As far as IT projects go, these go in relatively easily. The payback is pretty visible to multiple constituencies within the business. Now you've created a "win", the value of the approach is better understood, and people are willing to go a bit further ...
- Move to unification and leverage.
It's a classic story. Once the legal department discovers that IT can retain and search old email, they figure out that it's possible to do this for ALL kinds of information. So after doing one or two application ILM projects, the realization dawns that maybe there should be an approach to doing this for ALL kinds of information, not just email.
At the same time, there's usually a parallel realization that most organizations have a treasure trove of information that can be repurposed for a variety of new uses: creating 360 degree views of customers or business processes, collaborating across time and space around projects and ideas, helping with compliance and discovery issues, and so on.
The requirement changes -- one set of tools that finds and classifies information, a virtual repository to store metadata, a common set of policy engines that determine what needs to be done and when, as well as a rich application environment to create new uses for information.
In this phase, ILM isn't about saving money or avoiding risk, it's about bringing more value back to the organization with the information it already owns. And that's the fun part.
Now, I won't bore you with all the EMC products and services that actually do this (it'd be a very long post indeed), but we set out to provide customers with a reasonably comprehensive set of offerings to do all three, and I think we've achieved that to a relatively high degree.
And, oh yes, there's still much more to do ... and, like all large undertakings, it's a journey.
Whether you decide to call it ILM or IDM or an IT conspiracy, that's your choice. The buzzwords don't do much for me, but the ideas do.
A retrospective
Despite all the criticism, there's a pretty compelling body of evidence out there -- thousands of customers in the first two phases, and we're starting to see several hundred customers purposefuly edging towards phase III ILM, usually with a strong vertical focus that lends itself to this approach. There's a few showcase successes in this advanced category, and by this time next year, there'll be dozens and dozens.
And nobody would be doing it unless it solved a real problem, now, would they?
But let's step back a bit again, and wax philosphical for just a moment. For many customers, the root problem isn't going to go away. And ILM (or whatever you want to call it) seems to be an inescapable trend in the industry.
Sooner or later, I think it's inescapable that IT will have to understand the value of their information, and act appropriately.
And I don't think anyone is going to argue with that premise.
Comments