It’s official: ILM (Information Lifecycle Management), the three year old storage industry marketing thrust, is in critical condition. The good people over at the StorageNetworking.org have politely raised the red flag on ILM in three reports (see Information Lifecycle Management in Perspective: Initial Findings From Surveys of Top Management , Information Lifecycle Management: An Analysis of End User Perspectives and ILM Survey: What Storage, IT and Records Managers Say) on their independent survey research.
The surveys and analysis show that ILM is suffering from several major ailments:
- Definition: prospects don’t have a clear definition of ILM – policy people see it as policy and procedure, while techies see it as a technical and systems management issue
- Cost: prospects believe ILM “. . . has the potential to introduce a greater number of problems than it can solve.”
- Efficacy: a general concern that technology is being thrown at a soft problem
To recap: people don’t know what it is; believe it can create more problems than it solves; and aren’t sure it is a technology problem. For a 3+ year old major industry initiative those are stunningly poor results.
Marketectures: The Good, The Bad, And The Ugly
Technology marketers commonly offer a vision intended to inform customer planning and research. When such visions are thought-through and offer real benefits, they are a powerful catalyst for action. Good “marketectures” can help customers, vendors and analysts focus on a set of important problems and, through dialogue and investment in products and implementation, move the industry forward. They also provide air cover for the folks on the ground who are working, often against enormous organizational inertia and resource constraints, to help justify the investments that move the organization forward.
Poor marketectures, such as ILM, are immensely damaging. First, the industry has blown hundreds of millions of dollars in marketing, sales and customer time. Tens of thousands of customer supporters, or “inside sales”, get bogged down, and sometimes fired, for promoting unworkable schemes. C-level execs lose respect for the vendors as they see a cynical sales program instead of the hoped-for partnership. Finally, the opportunity cost is huge: years that could have been spent solving real problems are lost forever.
Voodoo IT
I’ve deconstructed ILM horsepucky (see ILM is Bunk and Gee, “Users Cite ILM Shortfalls†– Maybe ILM IS Bunk) a couple of times over the years. If I could see it two years ago from a distance, folks inside EMC and other ILM cheerleaders must have seen it much sooner. It points to an incredible lapse of leadership that vendors would knowingly promote something they knew, or should have known, they had no chance of delivering.
Just to be clear: the objection to ILM is that it isn’t implementable. If you could do it it would be a Good Thing. As the SNUG ILM Survey Analysis noted:
Paradoxically, in many industry discussions of the preparations needed for ILM, analysts, business reporters and others [ed. note: egregious example here] typically direct attention to assessing the technical infrastructure, and the fit (alignment) with business demands for improved information. . . . These assessments come perilously close to missing the lessons of past practice. Indeed, the problems brought out in data management and data quality initiatives of the past were largely organizational, not technical. Achieving performance was more a matter of effectively managing people and process, not just the alignment of IS with business requirements.
[bolding added for emphasis.]
Here Are Some Real Problems To Focus On
The ISIC reports list some real problems that customers would love help on. These include:
- Email, Application Archiving and Repositories
- Data Center Consolidation, Storage Consolidation, Backup and Restore
- Data Classification and Migration
- Records Compliance and Management
- Customer Records Management, Data Warehousing
I’d love to see vendors drop ILM and start focusing on how they can help customers prioritize and manage these more mundane problems. I’ll bet the salespeople would love it too.
Update
In the initial post I’d gotten some of the links and titles wrong. They’ve been corrected.
ILM has some value.
Mostly it is overhyped by vendors running out of hardware to sell.
The real answer is CAS (Content Addressed Storage).
Delivered the way Njini does it.
http://www.njini.com/
I developed, Implemented and used a Storage concept called the “Storage Hill”. Later I found out it is called ILM (Information Lifecycle Management).
I spent years laboring in obscurity developing the E2EIoD (End-to-End Information on Demand) concept, of which the “Storage Hill” is a component. Then I find out IBM had done a more complete job and called it SOA.
The “Storage Hill” and commercially available ILM do not have the same objective. The “Storage Hill” solved a common IT shop problem associated with the classic standard IT processes of:
1) Configuration – new equipment or de-commissioned/re-commissioned
2) Replication – covers archiving, migrating, copying
3) Re-configuration – de-commissioned/re-commissioned in-place
The “Storage Hill” is a concept to automate meeting the SLA/SLM requirements for Managed Units of Information. By definition, a Managed Unit of Information must reside on a Managed Unit of Technology.
Managed Units of Information are all ROI.
Managed Units of Technology are all TCO.
Down the “Storage Hill” means lower TCO which is synonymous with slower response and less bandwidth. Up means faster response and flex bandwidth.
As the ROI/TCO ratio drops below the SLA/SLM threshold the Managed Unit of Information slides down the “Storage Hill” to a Managed Unit of Technology with a TCO that generates an acceptable ROI/TCO ratio.
If no such TCO exists in the infrastructure, a human is notified for the SLA/SLM Online “Expiry/Persistence” action authorization.
Typical actions are Nearline and Offline. Both have many options. RAIDVD is one.
If the ROI/TCO ratio of the Managed Unit of Information increases due to
increased ROI the Managed Unit of Information slides up the “Storage Hill”
to an SLA/SLM acceptable TCO.
The “Storage Hill” is normally only applied to those Managed Units of Information that:
1) Generate 80% of the company revenue (ROI)
2) That Information, which if lost, will put the company out of business
If Datacore had done their software right the “Storage Hill” would have been very easy to Implement. If Datacore fails to move to the “right” objective, they will be obsolete.
Why do I say that? Because with the introduction of Virtual Machines and Servers the game has changed again, for the better. I don’t need Datacore to do what needs to be done. I could use their software if the price is right.
Take a look at:
“Virtual machines help with disaster recovery”
How virtual machines can play in a disaster recovery scenario
New Data Center Strategies Newsletter By Andreas M. Antonopoulos, Network World, 07/18/06
[Begin long URL]
http://www.networkworld.com/newsletters/datacenter/2006/0717datacenter1.html
[End long URL]
The focus of this article is Disaster Recovery. Just think about how the same process could be applied to standard classic IT processes?
Nortel… humor
“New Bill Gates charity project: rescue Nortel”
[excerpt]
“If you dump two buckets of dead fish into one large bucket, and swirl them around, what are the chances of producing a living fish?”
Full post at:
[Begin long URL]
http://dialzero.blogspot.com/2006/07/new-bill-gates-charity-project-rescue.html
[End long URL]
don’t complain with your lights on and your mouth full- those who can’t sell…. apparently now they write instead of teach
Dear Not A Fan,
It’s EMC’ers like you that give the company a bad a name. You can’t respond to the substantive comments, or the independent research, so you just have a little hissy fit.
EMC got big and successful by providing a decent product that offered real customer value and aggressively selling and marketing it. But that doesn’t mean that everything the company does meets that criteria today. ILM certainly doesn’t. If you disagree, tell me why. I’d love to have that conversation.