Home // ICIW 2016, The Eleventh International Conference on Internet and Web Applications and Services // View article
Authors:
Sathyamangalam R Venkatramanan
Rema Hariharan
Ashok S Murthy
Keywords: Software Load Balancer; feedback; SLA; load distribution; minimum nodes; energy optimization
Abstract:
We describe a load distribution algorithm in this paper that uses the current transaction response time as feedback for dynamically routing traffic to a minimal number of machines that run a business function (pool), with the constraint to consistently meet the response time requirements. This enables us to dynamically vary the number of nodes as per traffic levels, traffic mixes, and varying node capacities - a typical scenario in cloud environments. First, we present details of the basic algorithm followed by an extended version. Both have been implemented and tested in the eBay private cloud. We include graphs that show how the number of active nodes vary with incoming traffic volume while preserving the response time requirements. Results of using the extended version illustrate how the performance of the mirror environment closely matches that of the real environment while running production traffic.
Pages: 39 to 45
Copyright: Copyright (c) IARIA, 2016
Publication date: May 22, 2016
Published in: conference
ISSN: 2308-3972
ISBN: 978-1-61208-474-9
Location: Valencia, Spain
Dates: from May 22, 2016 to May 26, 2016