Home // ICAS 2014, The Tenth International Conference on Autonomic and Autonomous Systems // View article

Resource Aware Workload Management for Autonomic Database Management Systems

Authors:
Wendy Powley
Patrick Martin
Natalie Gruska
Paul Bird
David Kalmuk

Keywords: workload management; database management systems; autonomic computing; scheduling

Abstract:
Workloads running in a multi-purpose database environment often compete for system resources causing resource contention, which leads to poor performance. Autonomic database systems will be required to recognize that the system resources are not being utilized optimally and take action to correct the situation. Workload management techniques can be used to choose an appropriate mix of concurrent work to reduce resource contention. We describe a resource aware scheduling approach that predicts the amount of CPU, I/O and sort heap memory that will be required by a query and schedules each query to run only when doing so is unlikely to overwhelm the resources. We present experimental evidence that indicates that overall system performance can be improved using this technique.

Pages: 31 to 36

Copyright: Copyright (c) IARIA, 2014

Publication date: April 20, 2014

Published in: conference

ISSN: 2308-3913

ISBN: 978-1-61208-331-5

Location: Chamonix, France

Dates: from April 20, 2014 to April 24, 2014