EMC, Isilon, and the Rebirth of Storage

Yesterday’s storage industry news is certainly exciting. EMC has signed a definitive agreement to acquire Isilon Systems. This comes on the heels of HP’s acquisition of 3Par. With these two acquisitions, three broad themes are clear:image

1) Storage is “sexy” again.

This is not surprising in an environment where unstructured data is growing by 60 percent or more each year and 50 percent of all applications are expected to be virtualized in the next three years.

2) In particular, the scale-out approach to storage is gaining momentum.

3Par, of course, is known for its innovations in thin provisioning and scale-out SAN, and Isilon virtually defined the market for scale-out NAS.

With scale-out architectures like 3Par, Isilon or Gluster, disk, compute and I/O resources can be added independently as required for increasing capacity and performance. Contrast this to the scale-up approach where enterprises must upgrade to bigger and bigger boxes as capacity or performance needs require and/or significantly over-provision for needs that may or may not materialize over the next three years.

3) The forces that have fundamentally transformed the computing world are beginning to transform the storage world, creating a unique opportunity for innovative small firms.

This is perhaps the most important theme. Now is not just an exciting time for mid-sized storage players—it is an extremely exciting time for startups as well.

Over the past 10 years, enterprises have seen enormous gains as they migrated from proprietary, monolithic server architectures to architectures that are virtualized, open source, standardized and commoditized. The move to cloud architectures promises even greater advances in both flexibility and economics. I can think of no better encapsulation of the current computing trends than the following quote about VMware:image

VMware’s philosophy on the future of enterprise IT is this: Corporations must be more platform agnostic. Every application will be fully virtualized, mostly automated and controlled and administered via the Web, able to be hosted practically anywhere, and moved around at the drop of a hat.” –Carl Brooks, SearchStorage.com, August 2010.

Unfortunately, storage has not kept pace with computing. The proprietary, monolithic and scale-up solutions that dominate the storage industry today do not deliver the economics, flexibility and scaling that modern data centers need in a hyper-growth, virtualized and cloud-based world.

The scale-out approach demonstrated by newer firms like 3Par and Isilon certainly addresses part of this shift.

However, the trends discussed above have created a unique opportunity for even more radical innovations, and from smaller firms.

I believe it is inevitable that storage must evolve not only to support the new compute environment but must also begin to “look” more like the computing world. In other words, storage is about to become open sourced, commoditized, standardized and virtualized.

Such periods of creative destruction are much rarer in the enterprise space than they are in the consumer space but are stunning when they occur. Fortunately, for small firms, the recipe for succeeding in a disruptive market is clear. Consider, for example, the path taken by companies like Red Hat in disrupting the server O/S market, Salesforce in disrupting the CRM/ERP market and MySQL in disrupting the RDBMS market. The rules for radical disruption become fairly obvious:

Rule One: Market conditions must be right. It is much easier for small, egg-eating mammals to thrive when the dinosaurs are getting knocked off their feet by asteroids. Similarly, it is much easier for small companies to thrive when market conditions are upending the incumbents, and the incumbents are either unable or unwilling to adapt to fulfill the new niches exploited by the small players. Google was able to disrupt the publishing and advertising worlds, but it took the Internet bringing publishing and distribution costs to near zero before they could do so.

Rule Two: To disrupt a well-established market, innovations should be radically better. Being incrementally cheaper or incrementally faster will not do much good against the reputation, operating history and rich feature sets of large competitors. For example, MySQL was not only radically cheaper than existing RDBMS (i.e. an entry price of $0), it was also significantly better for PHP applications and enabled the entire LAMP stack to be commoditized. Linux was not just a cheaper version of Solaris; it also allowed the entire RISC-based hardware stack to be replaced by commodity X-86 boxes.


    • A corollary is that you won’t win by trying to match an incumbent’s full feature set. Instead, aim for parity on the most important features and be radically better on the few dimensions that really matter.

Rule Three: Innovate not only on price and product but also on development, sales and distribution methodologies.

    • Since incumbents also have larger sales forces, bigger engineering staffs, deeper channel relationships and bigger marketing budgets than star- ups, it is often critical to fight the sales, development and distribution battle on different ground as well.
    • For example, Salesforce’s SaaS model radically innovated by creating a sales and deployment model that could bypass MIS departments, creating a solution that could be deployed in days rather than involving years-long centralized projects.
    • Similarly, open source projects allow the complicated licensing and direct sales models to be bypassed in favor of high volume trial and conversion processes. By being willing to cede a little control and give value back, a startup like Gluster can effectively acquire a global staff of thousands of developers, testers, marketers and salespeople. This radically shortens both development and sales cycles. Over 95 percent of Gluster’s sales now happen over the phone, from prospects who heard about us from other passionate users, who successfully piloted the product for free in their labs and who have come to us ready-to- buy as they go into production.

In the long run, of course, this pattern of disruptive innovation is nothing new. As Clayton Christensen pointed out in his study of disruptive innovation, not one of the independent disk drive companies that existed in 1976 survived to 1995.  In the history of the hard disk drive industry, the leaders stumbled at each point of disruptive technological change: when the diameter of disk drives shrank from the original 14 inches to 8 inches, then to 5.25 inches, and finally to 3.5 inches. Of course, the pattern of disruption continues, with solid state storage, scale-out and cloud storage as a few prominent examples.

So, while this is challenging news for incumbents, it’s great news for startups. More importantly, this period of radical innovation should provide to be fantastic news for the industry as a whole.