Abstract. Traditional workflow systems don't handle dynamic scenarios well, as they are centralized and pre-defined at the start of the project. In this paper, we present a P2P framework to support dynamic workflows, which uses contracts to deal with unexpected changes in the flow of activities. DynaFlow is an agent based framework, where agents take action when exceptions occur. DynaFlow could provide adequate support for E-Science experiments mapped into workflows instances, with tasks operating in distributed environments and diverse types of computational resources and data.
Abstract. Many scientific experiments deal with data-intensive applications and the orchestration of computational workflow activities. These can benefit from data parallelism exploited in parallel systems to minimize execution time. Due to its complexity, robustness and efficiency to exploit data parallelism, grid infrastructures are widely used in some e-Science areas like bioinformatics. Workflow techniques are very important to in-silico bioinformatics experiments, allowing the e-scientist to describe and enact experimental process in a structured, repeatable and verifiable way. The main purpose of this paper is to describe our experience with Tavena Workbench and PeDRo, which are part of my Grid project. Taverna is provided with a workflow toolset and enactor, allowing the specification of processing units, data transfer and execution constraints. As a data entry tool, PeDRo provides a model, a controlled vocabulary and field validations for Web Services descriptions, leveraging the knowledge associated to the workflows. The main contribution of this work is a summary of some considerations drawn by our experience with the use of these tools, emphasizing its advantages and negative aspects, together with proposals for some future improvements.
Many projects are characterized by their flexibility and high number of changes before a definitive solution is implemented. In these scenarios, the people involved may change, as may deadlines, assignments and roles. Traditional workflow systems don't handle dynamic scenarios well, as they are centralized and predefined at the start of the project. To address these problems, we propose an agent-based approach to dynamic workflow management, where participants may join or leave and roles may change depending on the situation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.