Home // International Journal On Advances in Intelligent Systems, volume 5, numbers 1 and 2, 2012 // View article

Utility Functions in Autonomic Workload Management for DBMSs

Authors:
Mingyi Zhang
Baoning Niu
Patrick Martin
Wendy Powley
Paul Bird

Keywords: Self-Optimization, Utility Function, Autonomic Computing, Workload Management, Database Management Systems

Abstract:
Utility functions are a popular tool for achieving self-optimization in autonomic computing systems. Utility functions are used to guide a system in optimizing its own behavior in accordance with high-level objectives specified by the system administrators. It is, however, difficult to define a new utility function or evaluate whether an existing utility function is appropriate for a specific system management scenario. In this paper, we discuss the fundamental properties of an effective utility function for autonomic workload management in database management systems (DBMSs). We present two concrete examples of utility functions to illustrate the properties. The utility functions are used for dynamic resource allocation and for query scheduling in DBMSs. The utility functions help the systems translate high-level workload management policies into low-level tuning actions, and therefore ensure the workloads achieve their required performance objectives. A set of experiments are presented to illustrate the effectiveness of the two example utility functions.

Pages: 66 to 75

Copyright: Copyright (c) to authors, 2012. Used with permission.

Publication date: June 30, 2012

Published in: journal

ISSN: 1942-2679