This week, the Open Data Center Alliance announced itself. Yet another industry association? I hear you yawning already. Don't nod off, though. Within two years, the ODCA is likely to be the most important, powerful, and useful industry association in information technology. How is that possible?
The association is creating a culture of sharing between global IT leaders, allowing them to learn from each other and cooperatively influence tens to hundreds of billions of dollars worth of annual IT procurements. That is a wow-worthy new world order.
It's true that IT is already chock-a-block with industry associations. We have groups defining a broad range of standards (DMTF, IETF, OASIS, and W3C) and groups focused on developing specific technologies (PCI-SIG or USB Device Working Group). We have ecosystem development groups (Blade.org or the Itanium Solutions Alliance). We have groups aligned with specific vendors (most user groups), industries (ITU-T for telecommunications), nations national standards bodies such as ANSI and DIN), and even all humanity (the International Standards Organization).
What don't we have? We ain't got dames. Uh, sorry--wrong song.
What we don't have are organizations aligned with, populated with, and driven by IT users. Virtually all of IT industry organizations have been thoroughly vendor-driven. Vendors have been the ones with the obvious business case, will, and resources to make up-front investments. They pay the bills, and they provide the visions, directions, staffs, and expertise that make industry organizations tick. Therefore, they get to frame the issues and set the standards that drive IT forward.
This has left user organizations with essentially no independent, broad-based, coordinated voice. To be sure, users share experiences and techniques in user groups surrounding particular products and vendors. There are plenty of after-work inter-firm discussions over drinks, and the mobility of IT personnel from job to job helps cross-pollinate lessons learned. But when it comes to complex macro issues--for example, how data centers should be organized, workloads coordinated, or software licensed--haphazard informal sharing is insufficient, and every enterprise essentially has to chart its own course. Until now.
The Open Data Center Alliance is an independent user group that, while kicked off by a seed grant from Intel, will be funded and operated by large IT users. They will define common use cases and deployment models for operating modern data centers and virtualized/cloud infrastructures. The initial areas of focus include scale-out storage, unified networking, policy-based power management, trusted computing pools, provisioning, dynamic workload placement, licensing models, and infrastructure compliance. That list of topics is sure to grow.
All told, the alliance represents more than $50 billion in annual IT spending. I have no doubt that there will be well more than 100 members and $100 billion in IT spending before long.
The list of members is even more impressive. The Steering Committee includes BMW, China Life, Deutsche Bank, JPMorgan Chase, Lockheed Martin, Marriott International, National Australia Bank, Shell, Terremark, and UBS. There are currently just more than 60 other members, including AT&T, the Bombay Stock Exchange, CERN, ING, and Savvis. All told, the alliance represents more than $50 billion in annual IT spending. I have no doubt that there will be well more than 100 members and $100 billion in IT spending before long.
And the real kicker? The members have agreed to follow the resulting Usage Model Roadmap to guide their planning and purchasing decisions. Common IT procurement definitions and requirements across a large swath of global IT customers? Sacre bleu!; RFPs will never be the same again.
Why now? Because business requirements and IT capabilities have evolved. In the old days, apps weren't integrated, systems weren't linked, resources weren't shared--within individual enterprises, much less between them. But everything evolves.
On the business side, trends like globalization, integrated supply chains, and straight-through processing have become the norm. On the technology side, networks, virtualization, and cloud computing have risen to enable and promote integration, consolidation, and sharing. Organizational cultures take a few years to catch up, but the silos and fiefdoms of Old IT are rapidly giving way to a shared-resources, service-led model of IT. This shift isn't really optional; the scale points, quality-of-service requirements, and budgetary pressures at or under which IT now operates gives little choice.
There's also a more a touchy-feely enabler: Humanity's new culture of sharing. SMS/texting, blogs, Twitter, Facebook, YouTube, and other participatory, social media have rapidly expanded the boundaries of what we feel is appropriate and desirable to share. Those broad social attitudes combine with the realization that the "let's hide what we're doing, lest our competitors steal our competitive advantage!" that IT has operated under hasn't really led to additional competitive advantage. It actually hurts, as it deprives users of the ability to share knowledge and create leverage in dealing with vendors.
Users gain power and leverage through this banding together; vendors, on balance, lose power and influence.
Many many look to the ODCA's documents, specifications, and other formal outputs as its key deliverables. In fact, while formal documents will be more often cited in procurements, the increasingly open discussions, interactions, cooperation, and learning that will occur among ODCA members will be far more powerful and disruptive. Users gain power and leverage through this banding together; vendors, on balance, lose power and influence.
To be sure, the ODCA is just getting off the ground. The basic organization is set, initial members are onboard, and five Technical Workgroups (Infrastructure, Management, Security, Services, and Government & Ecosystem) have been formed. The group is working to have an initial set of Roadmap documents available in early 2011. But it remains the early days. Even with the best of intentions and resources, it will take time to create solid, deep specifications, to have the members agree on them, and to polish and bullet-proof them. It may be a year or more before the DCMA's written documents begin to flesh out and start to be routinely seen as procurement requirements.
And there are some risk factors. The large number of members, and the desire to remain process-, industry-, technology-, and vendor-neutral could make the road map documents too bland and inclusive, diluting their precision, prescriptive power, and value. The many hosting/cloud service provider members could sway the road maps too far from the enterprise-y toward the hyper-scale.
Enterprises, long accustomed to having vendors do the heavy lifting of defining new approaches, could be reticent to taking up the task themselves. And, of course, there's scope; the ODCA relates to data center and infrastructure, the proverbial "ops" side of IT, leaving "apps" largely alone. But such risks and constraints are par for the course and quite manageable. Given the resources and players already assembled, I'm optimistic.
The Open Data Center Alliance is an exciting, important change to IT's status quo. I cannot name another genuinely powerful, broad-based group of IT users. A group of IT leaders getting together to learn from each other and chart a collective course about where the new face of IT is heading? And they'll each base their IT plans and purchases on this collective understanding? Wow. Watch out vendors! There's a new sheriff in town.