In-Depth

When It Comes to ILM, Storage Vendors Want to Lock You In

When it comes to information lifecycle management, storage hardware vendors want to lock you in to their platform.

• By Jon William Toigo
• 07/12/2005

A few weeks ago, I heard a presentation by Nexsan Technologies CEO (and SNIA Data Management Forum co-chair) Philip Black on the topic of content addressable storage systems (CASS) and their place in the storage food chain. Phil is an agreeable person and his company was among the pioneers of the low-cost, high-capacity storage array—offering first ATA-based storage frames, then products based on Serial ATA (SATA).

Even if I didn’t always agree with them (in particular, their product positioning vis-à-vis tape technology), I'd always liked Nexsan,. This was mainly because of Phil, but also because of senior VP Diamond Lauffin, who is probably the best salesperson I ever met in the storage industry, or in any industry for that matter.

A couple of years ago, Black and Lauffin took singular delight in debunking the fear, uncertainty, and doubt (FUD) campaigns waged against them by brand-name vendors of big iron storage (“If you put a bunch of desktop-class disk drives in an array, their vibrations will shake out all of the mounting rail screws within a couple of days!”). Their technique was laudable: when your opponent plays the FUD card, trot out a bunch of satisfied customers, one after another, who have never experienced such disastrous outcomes. Nexsan customers obliged in spades, all stating how delighted they were with the performance of the Nexsan products, despite their bizarre-sounding product names (ATABoy and ATABeast).

In my humble opinion, SATA's huge success last year can be attributed in no small part to Nexsan. They paved the way for the technology to go mainstream whether the big boys liked it or not.

As I was listening to Phil, someone I genuinely like, I became a bit perturbed. It started when he walked through a couple of SNIA slides that had found their way into his presentation deck, I assumed, as a tribute to his role in the Data Management Forum. I wondered if the industry association, in addition to hefty annual membership dues, required that its forum chairpersons fly the flag at all public gatherings.

There was something slightly sinister about the update. The message was that the industry was going to fix all of the problems associated with data management and information lifecycle management (ILM) so that we poor little consumers wouldn’t have to bother ourselves with such things as data naming schema or platform characterization.

The fundamental problem with this idea, as I see it, is that the industry is schizophrenic when it comes to ILM.

Half the industry wants to attack the challenges of data management with software. Nothing wrong with that. A boatload of software will be required to move ILM into the realm of the possible. The problem is that most of the storage software guys are looking at data management from the bottom up—that is, from the perspective of storage. Such an approach is inherently flawed.

Data inherits its characteristics, like so much DNA, from the business processes and applications that generate it. From a storage perspective, data is just a bunch of ones and zeros. In the final analysis, storage management software has little or nothing to do with data management.

Here’s the schizophrenic part: the other half of the storage industry is determined to deal with problems of data management the old fashioned way, by throwing more capacity at them. Many vendors try to side step the issue of granular sorting of data simplistically, by throwing another ten terabytes of capacity into the SAN.

The analogy I like to use is the kitchen junk drawer. Everybody has one: an accumulation point for every extra screw, broken part, old battery, tie wrap, box top coupon, and every Crayola Mona Lisa, rice-and-paste sculpture, and Valentine’s Day card your kids ever made for you. That is what storage has become in the absence of data management: a junk drawer.

The only way to fix the situation is to sort through all the junk in the drawer, throw out what you don’t need, and bag-and-tag the stuff you want to—or have to—save. That takes time and effort and, in some cases, a lot of finesse.

The same is true with data—only the hardware guys want you to add more drawers and distribute the unsorted junk among the new drawers in a more “balanced” way.

The problem with a SNIA forum trying to tackle the nuances of data management is that whatever naming schema they develop for data will likely tie itself back to a specific vendor’s hardware platforms. “This kind of data goes to our big iron array. This kind of data goes to our SATA array. This kind of data goes to our tape automation.”

That was the first thing that irritated me about Phil’s presentation. The other was the lengthy homage he paid to EMC. He singled out Centerra, Hopkinton’s Japanese Mortgage solution to long-term data retention that obligates consumers to buy only Centerra controllers for as long as they need to retain their data (80+ years for some health care companies), as the definitive implementation of CASS. I can think of others off the top of my head that are not hardware-centric, including Avamar Technologies’ AXION product, but there was no mention of them in Nexsan’s speech.

I wondered why Phil would give so much air time to a competitor’s product until his slides alluded to a similar product that would shortly become available from his company. What he was saying was that EMC had opened the door and established a market for proprietary CASS platforms that Nexsan would shortly step through with its own product.

What ultimately undid me was his casual reference to “stickiness.” That is another term for proprietary lock-in: sell a customer “sticky” technology and they are your customer for life. In spite of Nexsan’s long-time championship of consumerism and storage options, Black did not assail the “stickiness” of the EMC solution—far from it, in fact. He stated, almost with a shrug, that most data management solutions would probably be sticky for the foreseeable future, since it is in the interest of vendors who offer them to join the customer at the hip to their wares.

Black showed no concern about the deeper meaning of his words, confirming my worst fears about SNIA’s interest in all of this data management stuff. In the final analysis, regardless of the slogans and platitudes about integrated solutions and customer value, the objective of the cottage industry growing up around ILM (from storage hardware vendors, at least) is to lock you in to their platform.

Given the centrality of data management to the mission of IT, that kind of stickiness is something we’d better pay close attention to. Your comments are welcomed. jtoigo@toigopartners.com