Home // CLOUD COMPUTING 2013, The Fourth International Conference on Cloud Computing, GRIDs, and Virtualization // View article

Scalable Store and Forward Messaging

Authors:
Ahmed El Rheddane
Noel De Palma
Alain Tchana

Keywords: JMS; message queues; scalability; load balancing; flow control

Abstract:
Since the emergence of the Internet, and particularly with the outburst of cloud computing, the production of reliable and scalable distributed applications is an important area of research. Various middleware technologies were designed for that purpose, among which we find Message-Oriented Middleware (MOM), which provides reliable asynchronous communication through message queueing techniques. MOMs have been standardized using the AMQP protocol, and in the Java world, with the JMS API. In this paper, we extend a store and forward mechanism to improve the scalability of an end-to-end reliable asynchronous messaging infrastructure while remaining compliant to the standard JMS API. We design a flow control based load balancing policy that, on the one hand, reduces the risk of consumer queues' failures while maintaining a near optimal throughput; and on the other hand, insures the scalability of our load balancing mechanism on the producer's side. We report the evaluation of our solution deployed on a cloud computing infrastructure and implemented within Joram, an open source implementation of the JMS API and the AMQP queuing protocol. This work is now part of the Joram distribution available on the OW2 consortium.

Pages: 156 to 161

Copyright: Copyright (c) IARIA, 2013

Publication date: May 27, 2013

Published in: conference

ISSN: 2308-4294

ISBN: 978-1-61208-271-4

Location: Valencia, Spain

Dates: from May 27, 2013 to June 1, 2013