More on the simple architecture movement (expanding to the simple computing movement)

I have been talking about the simplicity of computing and the concepts around making architectures easier to use and reuse. In fact I have a concept that I wonder in fact if Microsoft and others should embrace.

Simple doesn’t mean small.

My proposed concept for Microsoft is to create a cloud testing area. Windows 10 is a very intriguing upgrade. But as a former consultant in that space I know that getting to the ability to upgrade can be an extremely complex process.

Why not create a cloud based solution that allows an organization to borrow 10 windows 10 systems and over the course of a week load their apps and see what doesn’t work. At the point they are done, release the machines (say one week free) and let the next customer have a shot. Given that the number of systems pre-deployed with windows 10 could be managed to say 1000 or so, you could in giving each customer 5 free systems easily create a test environment. Using the cloud solutions, resetting the environment to untouched is also easy.

The only thing the customer couldn’t test is domain connectivity.

Of course if a cloud broker were to offer this service then the customer could also test automation, and directory connectivity. It is another way that cloud brokers are going to be huge in the future. IT migrations that in the past were tough, will be a lot easier when you can deploy a test environment in a matter of minutes. Create a puppet or chef run book with all your applications, connect to the cloud you are going to use and away you go. The physical/virtual  creation and use of the cloud assets will take longer than IT will have to spend in setting up the test environment.

Simple is more than reduction of complexity. It is also automation of complexity to remove it. In the case of testing scenarios using a Cloud Broker to automate the creation of the assets is a simple answer to what has often been a complex problem.

Ultimately moving towards simple should save an organization money. First off you don’t have to have a reinforced shelf to hold your enterprise architecture. Or a spinning disk that stores something that hasn’t been touched since 1923. The living breathing document that represents your enterprise architecture allows you to plug testing and automation into your systems quickly.

Simple can equal speed.

Of course, cutting waste is another way to save money and the drive to simple helps there as well. Complex problems like testing and automation can be reduced by considering the automation of a cloud broker as a baseline. Documenting only the Delta’s in your architecture from the reference architectures of the solutions you are using reduces the page count of architectures.

I suspect Enterprise Planning solutions will start moving along this path eventually as well. Empowering IT to deliver simple solutions effectively. There are a set of simple changes to enterprise planning software that in the end would make them highly effective.

  • Automated data collection. A tool that will tell you everything deployed in your enterprise. From desktops to servers. What do I really have today.
  • Move that data into a presentation layer that I can zoom in and out of. IE I can evaluate the applications installed in a single location, or across the entire enterprise.
  • I can input the cost of helpdesk and the cost of infrastructure (network, director etc.) 1 time and have it applied to all applications.
  • IT personnel costs are applied once and then based on hours input into the time management system. Sure in the short run a change in business process makes things a little slower, but over time when you see actual IT costs for applications it may change the well future deployments of applications.
  • Enterprise Capabilities Architecture that comes out of the collection of the data.

Sometimes a short term cost like the additional complexity of changing your time management system has an impact (complexity). But knowing not only everything that is deployed but also knowing which applications cost the most to support can be an IT game changer.

The business of IT is the business. Knowing what the business needs but also knowing what that actually costs is a game changer.

.doc

Scott Andersen

IASA Fellow.