On July 31, Microsoft, along with 9 other vendors, made a significant announcement…the standardization of a modeling language to be used in describing computing resources.
The new draft specification, called SML (Service Modeling Language), is a key milestone in the realization of the DSI (Dynamic Systems Initiative) vision that Microsoft has been talking about for the past 3 years.
For three years now, Microsoft has been telling a story about how systems can become more intelligent, self-aware, automated and self-managing. In some sense the vision was largely science fiction because many of the elements required to implement this vision hadn’t been invented yet. However, Microsoft has stayed true to the vision and has started releasing products that implement some of the key elements of the vision. One of the foundational pieces has been the ability to create models that describe your environment. The idea was that using models you could create a blueprint of the system and then use that blueprint to do pre-deployment planning, capacity analysis, and eventually an actual deployment…using the models as a template for how the system should be deployed and configured. The vision went on to describe how these models could then be used to define an ideal state so that if / when systems started to drift from their original configuration, the system could recognize it and self-correct.
This concept was even more broadly expressed in the idea of systems being able to use models to identify which applications should be hosted, and where. The system could also understand requirements and dependencies so that the necessary resources could be automatically provisioned to meet service requirements…also defined in the models.
How does all this relate to infrastructure optimization? Well, the realization of these self-managing, dynamic systems is the same destination that the infrastructure optimization model expresses in the most advanced stages of the infrastructure optimization model. DSI is a strategy for building infrastructure that is intelligent, self-aware and dynamic. Infrastructure optimization describes a methodology for implementing the key technology elements that help a system achieve this dynamic state.
The announcement to standardize the language used to describe the models used in a dynamic environment is a watershed event for the industry. Now vendors and customers can come together to build on a common foundation for describing computing elements from any vendor regardless of their orientation. SML will be implemented by multiple vendors in their development tools, system services, and applications. All this means that the vision to create more dynamic computing systems can now include technologies and solutions from multple vendors making the infrastructure optimization journey inclusive of everything.
Over time, we’ll see more vendors, customers and industry groups adopt this standard. Computer Associates (CA) is the latest vendor to endorse the standard and join the SML working group. Others will follow. In the end, everyone benefits, and the IO model is strengthened by its ability to embrace a much broader collection of service offerings.
Pratul Dublish provides some excellent insights into SML which I recommend for full treatment of SML.
Dynamic Systems Initiative