Poster 89: BeeCWL: A CWL Compliant Workflow Management System

Authors: Betis Baheri (Kent State University), Steven Anaya (New Mexico Institute of Mining and Technology), Patricia Grubel (Los Alamos National Laboratory), Qiang Guan (Kent State University), Timothy Randles (Los Alamos National Laboratory)

Abstract: Scientific workflows are used widely to carry out complex and hierarchical experiments. Although there are many trends to extend the functionality of workflow management systems to cover all possible requirements that may arise from a user community, one unified standard over cloud and HPC systems is still missing. In this paper, we propose a Common Workflow Language (CWL) compliant workflow management system. BeeCWL is a parser to derive meaningful information such as requirements, steps, relationships, etc. from CWL files and to create a graph database from those components. Generated graphs can be passed to an arbitrary scheduler and management system to decide whether there are enough resources to optimize and execute the workflow. Lastly, the user can have control over workflow execution, collecting logs, and restart or rerun some part of a complex workflow.

Best Poster Finalist (BP): no

Poster: PDF
Poster summary: PDF


Back to Poster Archive Listing