An EJB-based very large scale workflow system and its performance measurement

10Citations
Citations of this article
9Readers
Mendeley users who have this article in their library.
Get full text

Abstract

The design and implementation of a workflow management system is typically a large and complex task. Decisions need to be made about the hardware and software platforms, data structures, algorithms, and the network interconnection of various modules utilized by various users and administrators. These decisions are further complicated by requirements such as flexibility, robustness, modifiability, availability, usability, and scalability. The scalability among them is the most impeccable issue, in recent, as the sizes of workflow systems and its applications are becoming much larger. Also, the organizations are recognizing that the de facto standard of workflow architectures, the activity-oriented workflow architecture that is the joint-flow reference model proposed by OMG, and its off-the-shelf and more prevalent in the near future. In this paper1, we propose a solution for the scalability issue - The e-Chautauqua VLSW architecture and system. For the realization of scalable workflow systems, we propose an architecture that is called workcase-oriented workflow enactment architecture, and implement an EJB-based workflow management system preserving the property of the architecture. We analytically and experimentally demonstrate that the architecture gives the best performance for the very large number of workcases (instances of workflow procedures). That is, our analytic results showed that (a) software architecture is as important as hardware architecture for VLSW performance enhancement, and (b) the proposed architecture is much more scalable. Based upon the analytic results, we have conceived the workcase-oriented work-flow architecture, and experimentally demonstrated that our system is much better scalable and gives the reasonable performance for the very large workflow applications. © Springer-Verlag Berlin Heidelberg 2005.

Cite

CITATION STYLE

APA

Kim, K. H., & Ahn, H. J. (2005). An EJB-based very large scale workflow system and its performance measurement. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 3739 LNCS, pp. 526–537). https://doi.org/10.1007/11563952_46

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free