As automatic processes match demand to resources, reconfigure the environment as needed, and reroute work around failing components, the human focus will shift to ensuring that services and business processes are functional. The availability of any particular server, disk or network component will be nearly irrelevant. These technical changes will result in changes to IT processes. Today, applications are often characterized through a bottom-up process, where the stack of an application is developed then tested to determine the configuration of the server that will provide the optimal performance. In an environment where applications are assembled using a number of shared services, and each service might exist in multiple instances and be used by many applications, this approach will give way to a top-down process that relies on capacity and performance measurements of services and on aggregations of resources as well as platforms.

Impact on monitoring and measurement

When measuring performance of an IT solution, much of the focus has been on the health of individual components, such as application servers and the applications themselves. This is often supplemented by end-to-end user experience metrics, and in many cases, by directly measuring the performance of business processes.

Measurement of business processes is already important: even if all servers are up, all services are operating, and all communication methods are available, there are other factors that can cause problems with performance or interfere with the completion of a business process. Measuring billing backlog or units in inventory may detect anomalies in a way that checking response times or server status will not.

In a self-managed environment where resources are assigned and reassigned dynamically, there may be no direct connection between the performance of a particular component and the health of the business process. On the one hand, sharing of a service means that its performance might affect many business processes. On the other hand, since services may exist in multiple instances, failure of a particular instance (or of an element of a pool of resources) may not have as much impact as the failure of a unique instance.

This means that measurement of business processes moves from important to critical. As a self-managed infrastructure assumes the responsibility for monitoring its components and working around problems, IT will increasingly turn its focus to the business processes that it supports. More attention can be given to monitoring business processes and end-to-end experience. This will help identify patterns of service demands that can be anticipated based on the business process cycle, and will allow IT to bring support of business processes to the forefront.

IT Roles

As AC becomes more entrenched in the IT infrastructure, the duties and responsibilities of IT professionals will also change. For example, in today's environments, there may be proportionately few people engaged in business process design and implementation, and many more involved in application implementation and server configuration. This will change over time due to the presence of the following:

• Facilities that assist business process implementers to map business requirements to reusable services that will accomplish their objectives.
• Tools that map the required services and other components to a logical representation of the infrastructure.
• A service-oriented AC infrastructure that can dynamically map the logical representation onto physical components, and keep services running through self-managing capabilities such as automatic allocation of additional resources as needed and migration of workload from failing components.

As a result, infrastructure design and implementation should be less resource-intensive, and generally should be accomplished outside of the context of a given project or program. The benefits of features such as self-configuration and self-correction will ease the routine operation and maintenance of the infrastructure. The result will be that IT personnel can focus on solving business problems rather than technical ones.