In Joe Weinman's "10 Laws of Cloudonomics" he does a great job of enumerating the strategic advantage of cloud computing for the public cloud providers. Not surprisingly, each law hinges on operating at a scale that no private cloud or enterprise data center can match.
When evaluating cloud computing providers for your business, consider these three simple facts and how they relate to scale:
First and foremost cloud computing is, according to Tim O'Reilly, "a crappy low-margin business". Jeff Bezos and Amazon understand this better than most. How many really profitable online booksellers can you name?
If you spend any time in the Cloud Computing Google Group you will hear a common argument against the value proposition of public clouds. It goes something like this:
- Cloud providers must always maintain enough excess capacity to meet the greatest aggregate computing demand of their customers.
- This is enormously expensive.
- That expense will be passed on to you, the customer.
- You're better off building/leasing a tailored solution.
Smaller providers who cannot reach this massive scale, and thus compete on lower margins, will need to push the aforementioned group hard on cloud related standards. Otherwise, they will be pushed back towards a hybrid hosting model. The OCCI cloud interoperability group is working hard on standards that should help keep the ecosystem healthy but the biggest providers have not yet indicated a willingness to participate.
Second, vendors will lock you in (on some level).
Which cloud providers are open and transparent enough to earn your trust? This one is tough because cloud providers aren't equal in terms of what they provide, e.g. infrastructure, application, service, platform, etc. Quite simply the higher you move up the stack, from infrastructure to platform, the more you are at risk for lock-in.
And don't forget data lock-in. Once you acquire Terrabytes of data in a cloud it may not be quick, cheap, or easy to move the data to another provider.
Third, if you are going to invest considerable resources into moving applications and data to the cloud, you have to place a bet.
The provider you choose must be capable of withstanding fierce competition from all sides. They have to finish the race and it would be nice if they placed.
The on-demand pay-by-the-drink model is very appealing on the surface, but it doesn't account for the substantial investment in training for your staff nor does it begin to cover required changes to IT processes. We need more data on the less obvious (and sometimes hidden) costs of moving to the cloud.
We're in the early stages of a highly capitalized build out of cloud infrastructure and services. As George Reese says in his post "The First Step Into the Cloud"
If you are a small or medium-sized business (SMB), this discussion is moot—it's foolish for an SMB to ever own any servers for any reason. The risks and costs of managing these servers will always exceed the risks and costs associated with leveraging cloud equivalents.For enterprises it is more complicated. Take baby steps. Start building cloud competency in your organization by deploying non-mission critical applications to the cloud.
