The centralized ecosystem for discovering, learning, and deploying high-throughput computing workflows using Pegasus-WMS.
The Pegasus Workflow Management System is an open-source framework that bridges the scientific domain and the execution environment. It automatically maps high-level abstract workflow descriptions onto distributed execution infrastructures like HTCondor pools, cloud providers, and supercomputers.
Define your workflow logically as a DAG. Pegasus automatically maps it to the available physical hardware resources.
Seamlessly automates data movement, staging, directory creation, and clean-up across distributed nodes.
Built-in job retries and "Rescue DAGs" ensure your workflows survive infrastructure and node failures.
Run everything from a 10-job pipeline on your laptop to a 1,000,000-job ensemble on the Open Science Grid.
Building scalable scientific workflows is historically a daunting task. Researchers often face a steep learning curve when trying to bridge raw computing power with reliable automation. We built this platform to provide a structured, accessible path from beginner to expert.
Browse our curated collection of community-contributed workflows. Find ready-made blueprints for genomics, astrophysics, machine learning, and more.
Scaffold new projects instantly using Cookiecutter or Copier, and confidently deploy them across local containers, HPC clusters, or the Open Science Grid.
Give back to the scientific community. Did you build something amazing? Follow our easy template to open a Pull Request and publish your workflow on PegasusHub.