Exhibits, Demos & Posters
Open Grid Computing Environment’s Workflow Suite for E-Science Projects
Authors
-
Suresh Marru, School of Informatics, Indiana University
-
Marlon Pierce, Community Grids Lab, Indiana University
-
Chathura Herath, School of Informatics, Indiana University
- Srinath Perera, School of Informatics, Indiana University
Abstract
E-Science users will want to construct, share, execute, and monitor sequence of tasks. These tasks may execute on machines ranging from their local workstations to high-end, grid-enabled compute resources. Often, these tasks are legacy applications written in various programming and scripting languages and are designed to be run in a single user environment rather than as a Web application. These tasks often need to be tied together into composite applications that need to span multiple computing resources. The majority of the scientific experiments in e-Science involve orchestrating multiple tasks in the correct fashion to produce scientific computational experiments.
Scientific workflows have proven to be a coherent and abstract framework that is capable of capturing such scientific experiments and hence have gained popularity among the scientific community. Many of the current e-Science projects have adopted different workflow systems and consequently have adopted custom workflow execution and description standards. We argue that the core of the execution semantics used by most scientific workflows can be expressed as a small common set, and we further investigate the possibility of having a higher level workflow description language that is capable of capturing and possibly bridging these different workflow systems.
In this demonstration we present a set of e-Science development tools for managing these problems: a high-level workflow composition monitoring and enactment tool, a toolkit that is capable of providing Web Service interfaces for command-line scientific applications, and a service registry that is used as a repository for sharing. This domain independent workflow suite will allow users from wide verity of environments to selectively and securely share their applications as web services and construct workflows with these services. The suite also features on-demand service creation, workflow orchestration and monitoring. Furthermore, these components can be used individually or collectively to build a small to large-scale e-Science infrastructures.
