Applying SPC to autonomic computing

Statistical process control (SPC) is proposed as the method to frame autonomic computing system. SPC follows a data-driven approach to characterize, evaluate, predict, and improve the system services. Perspectives that are central to process measurement including central tendency, variation, stability, capability are outlined. The principles of SPC hold that by establishing and sustaining stable levels of variability, processes will yield predictable results. SPC is explored to meet and support individual autonomic computing elements´ requirement. One timetabling example illustrates how SPC discover and incorporate domain-specific knowledge, thus stabilize and optimize the application service quality. The example represents reasonable application of process control that has been demonstrated to be successful in engineering point of view.