Buliding N1 Grid Solutions Preparing, Architecting, and Implementing Service-Centric Data Centers

The problems associated with managing today's data center infrastructure are a consequence of typical architectures and a set of technology trends. Server-centric applications have evolved into network-centric services, and the data center network has become a fabric of resources, rather than a network of discrete systems. It is this latter observation that is key to providing a solution.

The previously mentioned trends indicate that services are deployed and managed within the context of the data center network and that the data center network is the context for the fabric of resources onto which these services are mapped. Thus, the problems with managing infrastructures today are perhaps less those of managing a collection of traditional systems than one of managing a new type of system specifically, a system that is built out of the network, rather than one simply attached to it. In this new system, the servers, network switches, and storage arrays become components of the system that uses the network as its virtual backplane. Such a system needs an operating environment that hides the underlying complexity and enables the infrastructure and services to be managed at a higher level of abstraction. This results in services being delivered with greater efficiency and agility, as well as more cost effectively.

Taking a systemic approach to solving these problems means that the root causes of the problems might be addressed directly. More traditional approaches, such as a typical system management solution, simply deal with the symptoms and fail to take into account the transformation in the nature of the data center.