Performance of Service-Discovery Architectures in Response to Node Failures

Published

June 1, 2003

Author(s)

Christopher E. Dabrowski, Kevin L. Mills, Andrew L. Rukhin

Abstract

Current trends suggest future software systems will rely on service-discovery protocols to combine and recombine distributed services dynamically in reaction to changing conditions. We investigate the ability of selected designs for service-discovery protocols to support real-time distributed control applications by detecting and recovering from failure of remote services. We model two architectures (two-party and three-party) underlying most commercial service-discovery systems. We use simulation to quantify functional effectiveness achieved by the two architectures as the rate of failure increases for remote services. We further decompose non-functional periods into failure-detection delay and restoration delay. Our quantitative measurements suggest that a two-party architecture yields better robustness than a three-party architecture. We discuss the underlying causes for this outcome.

Proceedings Title

2003 International Conference on Software Engineering and Practice (SERP'03)

Volume

1

Conference Dates

June 23-26, 2003

Conference Location

June 23-26

Conference Title

(Las Vegas, Nevada)

Pub Type

Conferences

Download Paper

Local Download

Keywords

distributed software systems, failure recovery, service discovery, software architecture

Citation

Dabrowski, C. , Mills, K. and Rukhin, A. (2003), Performance of Service-Discovery Architectures in Response to Node Failures, 2003 International Conference on Software Engineering and Practice (SERP'03), June 23-26, -1, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=50787 (Accessed December 15, 2024)