By Topic

Load-balancing in a self-organizing server cluster using local leaders

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

1 Author(s)
Samah A. Senbel ; College of Computing and Information Technology, Arab Academy for Science, Technology, and Maritime Transport, Cairo, Egypt

This paper presents a dynamic self-organizing design for a server cluster that supports multiple services with different priorities. Our system dynamically allocates jobs to servers in a server cluster, while maintaining certain QoS objectives for each service. A "Local leader" is dynamically set up for each service, and it controls the membership of its service, and thus the cluster size. It communicates with other local leaders to maintain the system balance. We introduce a new technique for load balancing within a single service to be implemented by nodes with low utilization that drastically reduces the job rejection ratio as well as the average response time. The basic idea is for low-utilization nodes to aggressively advertise themselves to high-utilization nodes as candidate neighbors, and to attempt to "grab" jobs from the job queues of more busy neighbors. Load balancing between services is performed between local leaders to optimize the overall performance. The proposed technique performs well in all performance metrics: rejection ratio, average job response time, average node utilization, Server assignment, and reaction to an overload situation.

Published in:

Informatics and Systems (INFOS), 2010 The 7th International Conference on

Date of Conference:

28-30 March 2010