Uncertainty on how to support component services has grown and created a greater chasm between application development and operations.  As organizations allow for partners, customers and even other business units to use component services, there can be great success. But when services are reused, and applications are loaded at levels they were not originally built for, what happens to your service level agreements?  You may find yourself with conflicting service levels and loads that the original architects of the systems never accounted for. What do you do with the unpredictable loads on the original applications and how do you manage the service level agreements holistically?

This talk will confront the often forgotten issue of designing the data center to adapt to the changing infrastructure needs of component services in an SOA environment. Primarily through the use of service-level automation, the analysis will cover: The difficulty of predicting component service loads, how to automate component service levels, how to optimize compute power “overhead”, what type of optimization engines are available, where virtualization plays a role, and how to ultimately create a computing “fabric” to support your SOA component needs. Through this talk you will gain the understanding on how to support the growth created by SOA for your organization going forward.