Maximizing Profit in an Autonomic Computing System: A Fuzzy Control approach

eCommerce is an area where an Autonomic Computing system could be very effectively deployed. The growth of eCommerce has created demand for services with financial incentives for service providers. Revenues accrue if the admitted requests are processed within the specified deadline and costs are incurred if the admitted requests are not processed within the deadline. The number of requests to be admitted depends on system capacity and workload. This paper describes an approach to automate the admission control decisions in an Apache web server for maximizing profit. A fuzzy controller that implements hill climbing logic is used. This is an illustration of the self-optimizing characteristic of an autonomic computing system.