Boeing "design" by 8-year-old Harry Winsor
My incredible passion for anything related to aviation has burned an indelible problem into my existence.
I have a recurring dream.
It happens every few months, rarely with much deviation from the plot, yet it is guaranteed to end in the same way – me waking with a smile on my face. In the dream, I am wandering around the Pike Place Market in Seattle when I come across an open briefcase containing a set of detailed plans for how to build a Boeing airliner. The dream isn’t specific enough to give the model a 7×7 identifier, but I am deep enough in the moment to know that this is an important find. I have everything I need to build my very own Boeing – from mechanics to avionics – I only have to find the money, the parts, a place to construct it and someone to help me maintain it when I have built it. Where do I start ?
It’s at the end of that very question that reality jolts me awake. I smile. What use are the plans if I can’t afford the reality and practicality of the construction or the maintenance of the finished product ?
This week’s announcement from facebook around the launch of the OpenCompute project was in some ways a weird intersect of my dream with my daily reality as a technology architect.
In simple terms, the unveiling of the blueprints of facebook’s new Princeville, Oregon facility are akin to my discovery of the Pike Place briefcase. I could, theoretically, now build myself a brand new, state of the art datacenter as an exact replica of the new facebook facility – but – just as I find myself waking up to the reality of having to find the money, the parts, a place to construct and someone to help me maintain my dream airliner, the practicalities are strikingly similar…just as it makes no sense for me to build my own Boeing, it makes no sense for an enterprise to build a facebook data center.
Bottom line ? To today’s large corporate IT shops; those who either have, or will continue to operate on-premise or co-located “private cloud” environments, the excitement levels around the OpenCompute project (if anyone actually hears of it at all) will be all-to-familiarly low as sadly, to wake some of these sleeping giants, it will take more than a poke from the very same company who’s website their IT teams are trying to prevent employees from accessing.
OK, so you could argue that I am completely missing the point of OpenCompute, but I can assure you I am not. I’m simply giving a personal perspective of the gaping void between where we (as large enterprises) are versus where they (being the growing band of web scale success stories) believe we need (and have the resources) to get to.
My observation is that in some ways, the nascent cloud environment and all of its bountiful choice is making enterprise IT even more risk averse as the fear, uncertainty and doubt that can not easily, nor readily, be addressed to many a CxO’s comfort level. Can you imagine a conversation where the idea of building one’s own hardware, free from the benefits of $$$ of R&D / support of the large OEM vendors, to run enterprise LoB applications is floated up to board level ? My experience tells me we are a long way away from that being answered in the affirmative, even if it could be argued and proven that there may be a lower overall cost.
In response to Mark Thiele’s observations in his blog – Facebook’s New Data Center: What Can We Learn From It – I offered this additional, simple summary thought:
In enterprise thinking, there is a big chasm between “we set the standards” and “we are the standards”
The rate of change can quite honestly not be assimilated by many.
That said, I genuinely believe that for traditional service providers who are making investments in new areas and offerings, XaaS providers, OEM hardware vendors and those with plans to become giants in the next generation(s) of Systems Integrators that the OpenCompute project is a huge step forward and will be a fantastic success story over the next few years as the community and its innovations grow and tangible benefits emerge.
To finish up my aviation analogy, consider that today’s public cloud services and co-location providers are today’s equivalent of commercial airlines, providing their own multi-tenant services, price structures and user experiences on top of just a handful of airframe and engine manufacturers. OpenCompute has the potential to influence the efficiency and effectiveness of those manufacturers by helping to contribute towards ideas and potentially standards that can be adopted across the industry.
Do I think OpenCompute will instantly become a competitor to the proverbial IT equivalents of Boeing, Airbus, Bombardier or Embraer ? No. Although it could quite easily be the manufacturer of the new, ubiquitous Jet A1 fuel that helps the manufacturers and ultimately their customers to run a cleaner, greener offering – and that would be good for everyone.
I recently read an interesting article from ZDNet that proclaimed Hardware vendors prepare for the day when no one builds a data center – interesting in title, and substance of the write up, but I can’t help focusing on this, from Savvis senior figure, Bill Fathers:
“Enterprises started to build these large private clouds, but then decided to outsource or go cloud. The question becomes whether cloud can be adopted enterprise-wide. Why spend $50 billion and build your own cloud? A third of our wins over the last three quarters have come from companies that had some flavor of a private cloud.”
I have no idea from which planet Mr Fathers orginates, but it certainly didn’t cost us $50 billion to build our very successful private cloud. It is precisely this kind of misdirected rubbish that adds to my concerns over enterprise risk aversion. Not a smart move. I lost Faith in our Fathers.
As long as there are clouds there are skies. As long as their are skies there will be commercial and private jets. Each serve a purpose. Get used to it.
