Special Issue: First International Workshop on Workflow Systems in Grid Environments (WSGE2006)

Abstract

This special issue of Concurrency and Computation: Practice and Experience contains selected high-quality papers from the First International Workshop on Workflow Systems in Grid Environments (WSGE2006) which was sponsored by Hunan University of Science and Technology in China and held in the university on 23 October 2006 1. The WSGE workshop series aims to provide an international forum for the presentation and discussion of research and development trends to support workflow in Grid environments. WSGE2006 attracted a number of international attendants. A keynote speech entitled ‘Time Management in Workflow Systems’ was delivered by Professor Yunshun Fan from Tsinghua University in China. A panel discussion was also part of this event, comparing ‘Business Workflow’ vs ‘Grid Workflow’. One of the conclusions reached by the panel members was that a Grid workflow is not significant in any way compared to other such approaches, with the key additional aspects being the computation and data resources used (over a Grid) to enact the workflow. Grid workflow has been under investigation for several years 2-6. In particular, GGF (Global Grid Forum) 10 was focused on it in 2004. GGF 10 was then concluded as a special issue in Concurrency and Computation: Practice and Experience. The title of the special issue is ‘Workflow in Grid Systems’ 7, 8 and it was edited by Professor Geoffrey C. Fox and Professor Dennis Gannon from Indiana University in U.S.A. This WSGE2006 special issue is a follow-up of the GGF 10 special issue. Many existing research and development efforts have focused on the functional issues of Grid workflow, such as support for modelling languages or Grid workflow enactment engines 2, 4-6. Gradually, more and more attention is being paid to non-functional issues such as QoS (Quality of Service), time management, or verification and validation 9-12. Accordingly, this special issue has a focus on the non-functional issues. This special issue contains six papers based on those that were presented at WSGE2006. They are listed as 13-18. Particular attention was paid to the non-functional issues. Besides, research problems must be analysed systematically. For specific approaches or models, evaluation must be presented. Based on these, the six papers were selected and also reviewed thoroughly. They are briefed as follows. Zheng et al. 13 propose an enhanced workflow simulation model supporting the simulation of event-based interaction between composed (Grid) services. Using the proposed approach, the behaviour between composed services can be simulated and in particular evaluation of QoS for service composition can be carried out. An example is presented on the implemented prototype to demonstrate the feasibility of the proposed method. Brandic et al. 14 argue that there is a lack of a holistic environment for Grid workflows that supports QoS in all phases of the workflow lifecycle, i.e. from specification to execution. Existing Grid workflow systems either support the whole workflow lifecycle but lack QoS support, or provide only partial QoS support for certain stages of the workflow lifecycle. Hence, Brandic et al. 14 present ‘Amadeus’—a novel service-oriented environment for QoS-aware Grid workflows. A distinguishing feature of Amadeus is the holistic approach to QoS support during all stages of the workflow lifecycle from specification to execution. The architecture of Amadeus is described in detail. For each stage of the workflow lifecycle, experiments are conducted to evaluate the corresponding Amadeus components. Chen and Yang 15 propose a taxonomy of Grid workflow verification and validation. It classifies different issues in Grid workflow verification and validation into a particular set of categories. The relationship of some existing projects to the proposed taxonomy is also presented. Lee and Min 16 propose a PLC (Peak Load Control)-based orchestration system that can stably execute hybrid Grid services. The PLC mechanism uses the delay time algorithm for controlling a workflow engine's heavy peak load caused by request congestion for a short period of time. With PLC, the performance of workflow engines can be improved. Comparison and simulation analysis further demonstrate this point. Liu et al. 17 propose a time model for large-scale workflow execution spanning different working time zones. The model is called WdCM (Workday Calendar Model). WdCM can allow for the difference in world-wide working time zones. A case study is presented to demonstrate the feasibility of WdCM. Rygg et al. 18 present an intuitive workflow environment called GPFlow to support scientists with their research. The workflow in GPFlow wraps legacy tools and presents a high-level interactive Web-based graphical interface to scientists, so that they can access and execute their scientific workflows. Two bioinformatics experiments are performed to demonstrate its flexibility and simplicity.