In-Depth
Innovation 2008, Part 1: Can Ecosystems Trump Stovepipes?
Why the ISE ecosystem is the new face of storage.
By the time you read this, Xiotech will have announced the latest additions to its partner ecosystem. Here's why the Xiotech ISE ecosystem is on my radar and why I view it as one of the big storage innovations of 2008.
Let's begin by defining a few terms. To many, innovation is used interchangeably with invention or even inspiration. There may be an overlap at some level between these terms, but to innovate, technically, means to convert knowledge into new economic or new social benefit. It doesn't necessarily require the creation of any new product (the essence of invention), but it may involve combining new or existing products in new ways or to better fit a process. Inspiration, by contrast, refers to an epiphany by which an innovative idea or a newly invented product might come about.
This year we’ve seen many inventions. They continue to be announced: new magnetic recording techniques, new file systems, new functions for compressing/deflating data so more of it can fit on the same media, and new techniques for replicating data on hardware to safeguard it and to share it more effectively. The list goes on.
Inspiration is also in great supply. Given the current realities of economic, regulatory, and even environmental trends, something is building fires under a lot of IT people to find ways to do more with less, ensure the preservation and immutability of data, and reduce kilowatt consumption.
Translating inspiration and invention into innovation is the challenge. There are a myriad of new products as well as ideas about what to build with them, but innovation drives us into the realm of what can be accomplished reasonably and with diligence. Ripping out and replacing all storage today, and starting over with the latest product, might seem to some to be the best path, but it is not realistic. Innovators recognize practical constraints and try to work within them to create meaningful and strategic change over time that returns its investment in short term gains.
That's the long way of saying that real innovation unleashes value. Its nemesis, in storage at least, is stovepiping, which constrains value.
Stovepiping is old school and involves the creation of technologies that only work on one platform and in one way. Stovepipes deliver value, but in a very narrow context: to realize the promised gain or benefit, consumers must obtain all components from one vendor (or perhaps its cadre of like-minded vendors). Stovepipes contribute to top-line growth, but first and foremost for the vendors selling the stovepipes. It is zero-sum thinking that locks in consumers and locks out competitors, and it is sweetened by vendor promises that everything will work and there will be no finger pointing when anything goes wrong.
Which brings me back to the subject of this column: Xiotech is announcing a burgeoning ecosystem of vendors who are adding value around the ISE platform, a Lego-like building-block array (discussed in this column several times) that provides a great resiliency story (failing disks are refurbished on board the array) and a great management story (the product talks W3C Web Services standards).
ISE, however, stops short of emulating its cousin vendors: there is no extra "value-add" functionality on the ISE array itself. Instead of stovepiping these functions on the array to increase its cost, Xiotech is endeavoring to get the developers of value-add software functionality simply to enable their products to plug in to the ISE infrastructure. Sometimes plugging in means attaching to convenience ports in the box itself; sometimes functionality is best created off the array, in the pathways that data will take on its way to the ISE disk packs.
This is an innovative use of the term "ecosystem," by the way. Everyone talks ecosystem because it sounds so green and friendly. VMware has issued a cattle call to vendors seeking to join up with the virtual server ecosystem. Such terminology may just be "brochure speak" that is intended to signal that the joining partners' products will work with VMware -- which would be the case even if they didn't join the club.
Cisco Systems also has an ecosystem of vendors willing to bend their product designs to fit function sets that Cisco uses to make its IP switches a bit different than competitors’ IP switches and to support a higher price for the sale of gear that has become commoditized over time. In my mind, the term ecosystem usually brings to mind a bunch of small fries following the big shark looking for scraps of food that slipped through the jaws.
Xiotech can't lay claim to being a big fish -- not yet, at least. It hasn't penetrated the market with a sales blitz costing tens of millions of dollars like VMware has. It doesn't command the market share, for example, of a Cisco Systems within enterprise data centers. What Xiotech's ISE ecosystem is, frankly, can be compared to a farmers' market. The consumer decides what he wants to cook for dinner and goes to a convenient aggregation of vendors to buy just the ingredients needed to make the meal.
Unifying the Xiotech farmers' market are W3C standards, which ultimately enable the discrete ingredients to plug and play and to be manageable given their different pedigrees. The burgeoning ISE ecosystem is doing what the industry could not, despite many well-intentioned (and more than a few half-hearted) efforts within the Storage Networking Industry Association: it is helping consumers see that stovepipes are archaic remnants of another time and that every bit of the functional value proffered by stovepipe vendors can be obtained -- and at a lower cost -- by selecting your own ingredients to design your solution.
EMC didn't want this to happen. History will be written about how EMC tried to kill ISE in its cradle while the technology was under development at Seagate. They tried to pull the plug on the behind-closed-doors meetings that took place year after year at Storage Networking World between analysts, OEMs, and key consumers and the Advanced Technology Group. Simply put, they didn't like where ISE was going because it didn't support the stovepipe zero-sum game model of the big iron providers.
The ISE ecosystem is storage for the future. It is how IT innovators will build infrastructure so that it meets specific performance and functionality requirements without requiring consumers to buy more than they need. This innovation delivers enormous value, especially in the form of cost containment, and from what I've been told, it is resonating with IT innovators. Sales of the ISE platform are outstripping supply and exceeding the wildest expectations of Xiotech management and investors.
It is also, by coincidence, causing an increasing number of vendors of both storage software and hardware -- from the small component people to the finished-frame vendors to the software mavens -- to begin investigating what this W3C Web Services thing is all about and what it can do for their offerings.
Web services have been largely discussed as tools for combining application components on Web pages and cell phone displays. They have another, richer dimension as middleware for allocating services to applications from underlying infrastructure. Prior to his acquisition of ISE, Xiotech CEO Casey Powell was already moving in the Web Services standards-based management of his gear by hiring some of the most talented W3C Standards programmers on the planet to work out a Web services-based storage management facility on his Magnitude product line. ISE advances this innovative vision of the best way to build storage.
DataCore Software and Data Domain are early sign-ons to the ISE ecosystem. The former has long been an inventor and innovator in its own right, providing one of the best storage virtualization solutions around. Though I am not a fan of de-duplication except where it is judiciously applied, Data Domain has for the past year or so been setting the price in the market for de-dupe solutions. ISE gives them a target for a de-duplicating gateway service instead of requiring customers to buy a specialty platform for their shops.
Both of these ecosystem partners have been doing well with their own products, of course, but see an opportunity to attract more business in a new reality described by Web Services-managed data and infrastructure. I would not be surprised if we find more array product vendors adopting this line in the coming year.
Xiotech, with its ISE ecosystem, gets my nod as one of 2008's most important innovators. Your opinion is welcome: jtoigo@toigopartners.com.