Scalable service composition execution by means of an asynchronous paradigm

  • Vandikas K
  • Quinet R
  • Levenshteyn R
 et al. 
  • 3

    Readers

    Mendeley users who have this article in their library.
  • 2

    Citations

    Citations of this article.

Abstract

In this paper we describe our findings and experiences with regards to scalable service composition execution. More specifically we detail the process of redesigning our original core for service composition execution to adhere to a more asynchronous, event-driven, non-blocking paradigm. The outcome of this work is an asynchronous core that is scalable; limited by the amount of available memory and not by the number of available processing threads. Another characteristic of the asynchronous core is robustness as system starvation is not possible. Our findings are evaluated with performance results comparing the new asynchronous core to its predecessor. The evaluation focuses on two aspects; throughput and overhead. Our results show that the overhead introduced by service composition is very small: 4 to 12 msec. In addition, the new asynchronous core provides better throughput ranging from +3% to +49% for moderate load and no upper limit for higher load.

Author-supplied keywords

  • HTTP
  • SIP
  • Service Composition
  • Web 2.0
  • asyncrhony

Get free article suggestions today

Mendeley saves you time finding and organizing research

Sign up here
Already have an account ?Sign in

Find this document

Authors

  • Konstantinos Vandikas

  • Raphaël Quinet

  • Roman Levenshteyn

  • Jörg Niemöller

Cite this document

Choose a citation style from the tabs below

Save time finding and organizing research with Mendeley

Sign up for free