DAGMan
DAGMan is a workflow management tool within HTCondor that coordinates complex job dependencies. It allows users to define tasks as nodes in a directed acyclic graph, where execution order is based on data or process dependencies. DAGMan is used in large-scale research projects that require structured, automated workflows.
About
DAGMan (Directed Acyclic Graph Manager) is an HTCondor utility for defining and executing complex workflows where jobs must run in a specific order, with dependencies between tasks. Users describe workflows as directed acyclic graphs, with each node representing a job submission and each edge representing data or dependency relationships, allowing DAGMan to manage retries, error handling, and sequencing automatically.
DAGMan is an important tool for teaching and running automated, reproducible workflows on campus HTC/HPC resources, enabling researchers to scale from single jobs to robust pipelines without building custom orchestration tools. It is particularly useful for domains that involve multi-step simulations, data processing chains, or ensemble runs.