David Chappell


Get the Feed! Subscribe

Application Platforms  
# Monday, April 09, 2012
The idea of an application platform can be slippery. In the broadest sense, it includes anything that we use to create and run an application. And slippery or not, it's worth spending some time thinking about application platforms--they're fundamental to software development.

Toward this end, I've written a couple of short papers on this topic, both sponsored by Microsoft. The first, What is an Application Platform?, offers a concrete way to look at the idea. Starting with a general model, it then uses this model to take a high-level look at the Microsoft application platform today.

The second paper, Application Platforms and Business Strategy: Making the Connection, is a little more general. Starting with a simple definition of what business strategy really means, the paper ties this idea to application platforms, using the example of a fictitious rental car company that wants to reinvent itself. Working in the bowels of technology, as many of us do, can blind us to the connections between what we do and the larger world. But I don't care how technically focused you are: this kind of blindness is never a good thing. A primary goal of this paper is to help people from both the business world and the IT world see the tight link between what the business is trying to do and the application platform they've chosen to build on.

Application platforms are important, and so is thinking in broader ways about this topic. Both of these papers are meant to help you do that.

2 comments :: Post a Comment



...but Microsoft Private Cloud is bordering to deception .

The ultimate objective of virtualizatin is to form elastic cloud with on-demand computing power scale-up.

How could Microsoft confuse us with private cloud as it is not elasticity-aware . The application must tap to the elasticity-aware API before the application becomes elasctic

This isn't really correct. I think you're referring to what's commonly called autoscaling, the ability of a cloud platform to detect when an application needs more computing resources, then automatically provide those resources. Some public cloud platforms provide this and some don't, but private clouds rarely provide it. And while this might be useful in some cases, it's not the ultimate objective--it's just another potentially useful feature of a cloud.

Also, the Microsoft private cloud actually does do this. The application must be written to allow this kind of elasticity and it must have a service template defined for it, but if you've done this, System Center 2012 can add or remove VM instances for the app automatically as load changes. To see how this works, you might read the paper I wrote on the Microsoft private cloud (http://www.davidchappell.com/writing/white_papers/The_Microsoft_Private_Cloud_v1.0--Chappell.pdf).

Post a Comment

<< Home