• Sorted by Date • Classified by Publication Type • Classified by Research Category •
Jonathan Wildstrom, Peter Stone, Emmett Witchel, and Mike Dahlin. Adapting to Workload Changes Through On-The-Fly Reconfiguration. Technical Report UT-AI-TR-06-330, The University of Texas at Austin, Department of Computer Science, AI Laboratory, 2006.
[PDF]132.0kB [postscript]358.9kB
High-end servers that can be partitioned into logical subsystems and repartitioned on the fly are now becoming available. This development raises the possibility of reconfiguring distributed systems online to optimize for dynamically changing workloads. This paper presents one approach to solving this online reconfiguration problem. In particular, we learn to identify, from only low-level system statistics, which of a set of possible configurations will lead to better performance under the current unknown workload. This approach requires no instrumentation of the system's middleware or operating systems. We introduce an agent that is able to learn this model and use it to switch configurations online as the workload varies. Our agent is fully implemented and tested on a publicly available multi-machine, multi-process distributed system (the online transaction processing benchmark TPC-W). We demonstrate that our adaptive configuration is able to outperform any single fixed configuration in the set over a variety of workloads, including gradual changes and abrupt workload spikes.
@TechReport{TR06,
author = "Jonathan Wildstrom and Peter Stone and Emmett
Witchel and Mike Dahlin",
title = "Adapting to Workload Changes Through On-The-Fly
Reconfiguration",
institution = "The University of Texas at Austin, Department of
Computer Science, AI Laboratory",
year = 2006,
number = "UT-AI-TR-06-330",
abstract = { High-end servers that can be partitioned into
logical subsystems and repartitioned on the fly are
now becoming available. This development raises the
possibility of reconfiguring distributed systems
online to optimize for dynamically changing
workloads. This paper presents one approach to
solving this online reconfiguration problem. In
particular, we learn to identify, from only
low-level system statistics, which of a set of
possible configurations will lead to better
performance under the current unknown workload. This
approach requires no instrumentation of the system's
middleware or operating systems. We introduce an
agent that is able to learn this model and use it to
switch configurations online as the workload
varies. Our agent is fully implemented and tested on
a publicly available multi-machine, multi-process
distributed system (the online transaction
processing benchmark TPC-W). We demonstrate that our
adaptive configuration is able to outperform any
single fixed configuration in the set over a variety
of workloads, including gradual changes and abrupt
workload spikes. },
badwwwnote = "{\url{http://www.cs.utexas.edu/ftp/pub/AI-Lab/tech-reports/UT-AI-TR-06-330.pdf}}",
}
Generated by bib2html.pl (written by Patrick Riley ) on Tue May 06, 2008 09:31:03