Collaborative and Distributed e-Research Environment for Supporting Scientific Research and the Education Process

Collaborative and Distributed e-Research Environment for Supporting Scientific Research and the Education Process

Dukyun Nam (Korea Institute of Science and Technology Information, Republic of Korea), Junehawk Lee (Korea Institute of Science and Technology Information, Republic of Korea) and Kum Won Cho (Korea Institute of Science and Technology Information, Republic of Korea)
DOI: 10.4018/978-1-4666-0125-3.ch004
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

The efficient use of a scientific application service built on a computing environment requires technology that integrates each application service into a workflow so that the workflow is executed in a cooperative environment. There have been a number of attempts to automate research activities as a scientific workflow. However, there are practical problems in the full automation of research activities for a number of simulation programs and researchers. In the cyber environment for Collaborative and Distributed E-Research (CDER), the types of workflows need to be studied and supported separately and with different methodologies. In this chapter, the authors analyze the scientific research and education processes and categorize them into four types: simulation, experiment, collaborative work, and educational activity. They then describe the applications needed for each category. To justify their categorization of the CDER workflow, they examine the workflow of e-AIRS (e-Science Aerospace Integrated Research System), a problem-solving environment for aerospace research.
Chapter Preview
Top

Directed Acyclic Graph Manager (DAGMan) (Frey, 2002) allows users to submit a number of jobs with workflows using Condor, and provide the interface for various types of workflows. The goal of DAGMan is the automation of managing complex workflows, including the job submission process. The weakness of DAGMan is the lack of supporting control flow, such as conditional branch and iteration, because Directed Acyclic Graph (DAG) is limited in how it represents the dependency of each step. Pegasus (Deelman, et al., 2003), using the DAGMan execution engine, constructs executable workflows based on the information of workflow instances and usable resources. It makes it possible for users to design a workflow at the application level, regardless of the status of computing resources and the execution environment.

Triana (Taylor, et al., 2003) is a workflow environment that provides an intuitive graphical user interface, allows users to build their own services, and also provides several services for control or logic, such as loops and if clauses. In order to access grid infrastructures, Triana uses the Grid Application Toolkit (GAT) (Seidel, et al., 2002). GAT defines generic APIs for accessing Grid services. Because GAT focuses on executing collective services, it is not adequate to design the entire experimental process, which includes experiment planning, offline experiments, and collaborative works with GAT.

Complete Chapter List

Search this Book:
Reset