Earth System Modelling with Windows Workflow Foundation

Fairman, Matthew J., Price, Andrew R., Xue, Gang, Molinari, Marc, Nicole, Denis A., Marsh, Robert, Takeda, Kenji and Cox, Simon J. (2008) Earth System Modelling with Windows Workflow Foundation Future Generation Computer Systems

Download

Full text not available from this repository.

Description/Abstract

The GENIE project has built a Grid-enabled Earth system modelling framework that facilitates the integration, execution and management of component models for the study of the Earth system over millennial timescales. The existing framework supports collaborative study of GENIE models across heterogeneous compute grids through scripted workflows in the Matlab environment. While the scripting approach achieves simplicity and flexibility, it suffers from an essentially passive approach to work unit management and from a heavy reliance on a central database to provide fault tolerance. The Windows Workflow Foundation (WF) technology provides a rich set of features to support the authoring and execution of workflows, tracking services that enable the monitoring of a running workflow, and state persistence services that allow workflows to be recovered and resumed upon failure. We demonstrate how the Windows Workflow Foundation has been applied to build a complementary simulation management system which provides rapid composition, event driven logic and reliable hosting of the scientific workflows while interfacing to existing infrastructure. We also describe how the adoption of WF enables the application of a number of associated technologies to provide better interoperability and accessibility for the simulation system. These improvements are demonstrated through a parametric study of the bi-stability of the oceanic thermohaline circulation in a GENIE model where the effects of a new carbon cycle are studied.

Item Type:

Article

Keywords:

Workflow management, Earth and atmospheric sciences, Windows Workflow Foundation

Organisations:

Electronics & Computer Science

ePrint ID:

266022

Date :