I've known Werner Vogels for several decades, ever since we were both doing our PhDs. Like all good friends and scientists, we don't always agree on everything. Case in point is that Werner doesn't believe Private Clouds are clouds and I think his arguments against are artificial and short sighted. Now of course you could say that he and I take our perspectives on this debate precisely because of our employers. However, that's wrong, at least where I'm concerned.
As I said earlier this year, I think that today's definition of Cloud is limiting and emphasizes Public Cloud precisely because that's what most people have access to. But I also believe that Public Clouds are not going to be as important in the future. Cloud is a natural evolution of hardware and software (middleware) but if you liken the roadmap for Cloud to that of cars, today's Cloud's are like the Model T: showing everyone the potential, but not available to the masses. We should be looking at the equivalent of the next hundred years of evolution in automotive technologies as far as Cloud is concerned, bringing their benefits to the masses (of people and workloads).
This development has to include Private Clouds (which, contrary to what Werner states, don't necessitate corporations having to buy more hardware), but so much more. The true cloud is the collection of processors that exist virtually everywhere you turn, including mobile devices and sensors. That's where the definition of Cloud must go. In many ways it's returning Cloud to one of its progenitors, ubiquitous computing. By that point there won't be a Public, Private or Personal Cloud, there'll be "just" Cloud (or maybe some other term). Where your application is hosted will still remain important, but not because of any artificial reasons due to words such as 'private' or 'public'.
Subscribe to:
Post Comments (Atom)
4 comments:
I agree with the "Automobile Evolution" analogy - and just as lowered car costs drove demand for access, they also drove major changes in roads and support services - similar to the need for major network upgrades and new security services and business models.
Redefining cloud computing might be an option, but the way it is defined today very few companies do have a private cloud, see Private clouds - a great misconception?
Hi Herbjorn. Nice posting, but I'm not sure I agree with the fact that attributes such as elasticity and pay-per-use cannot be available behind the corporate firewall. They may be limited by virtue of a given organisations capacity to pull more machines into a Cloud "pool", for instance, but it's definitely not black-and-white. Plus, many companies have cross-departmental charging policies in place today for a range of things, including IT, so a pay-per-use approach might make a lot of sense and be possible.
As a person who works for Dell I think your article on private cloud is quite impressive. I think Private cloud is a marketing term for a proprietary computing architecture that provides hosted services to a limited number of people behind a firewall.
Post a Comment