Introduction to Control Theory and Its Application to Computing Systems

Our DDPE colleague Joe Hellerstein, along with Tarek Abdelzaher (University of Illinois), Yixin Diao (IBM), Chenyang Lu (Washington University), and Xiaoyun Zhu (HP), is presenting a seminar this week at ACM Sigmetrics.

The session Abstract follows:

Abstract.

Feedback control is central to managing computing systems and networks. For example, feedback is employed to achieve response time objectives by adjusting scheduling priorities and bandwidth allocations. Control theory provides a way to determine if feedback loops are stable (e.g., avoid wild oscillations), accurate in their control (e.g., achieve the desired response time objectives), and settle quickly to their steady state values (e.g., to adjust to workload dynamics). Recently, control theory has been used in the design of many aspects of computing, with a few examples of commercial products designed using control theory. Examples of where control theory has been used include: networking protocols (e.g., new versions of TCP/IP), real time systems, web servers, database servers, multi-tier computing systems, and workload managers.

This tutorial provides an introduction to control theory for computer scientists with an emphasis on applications. The topics covered include: (1) Introduction to control theory concepts and discrete-time linear systems; (2) Application to self-tuning memory management in IBM’s DB2 Universal Database Management System; (3) Application to model-predictive control in distributed real time systems; (4) Application to automated workload management in virtualized data centers; (5) Application to managing power and performance in data centers; and (6) Research challenges.

The Powerpoint slides that accompany the presentation are attached.

Sigmetrics Tutorial on Control Engineering -- Part 3 - Control in Distributed Real Time Systems.ppt