Argo Workflows
Argo Workflows is a container-native workflow engine for Kubernetes that lets teams define and orchestrate multi-step, parallel jobs — such as CI pipelines or data and ML pipelines — as Kubernetes custom resources.
Definition
Argo Workflows is a container-native workflow engine for Kubernetes that lets teams define and orchestrate multi-step, parallel jobs — such as CI pipelines or data and ML pipelines — as Kubernetes custom resources.
Overview
Each step in an Argo Workflow runs as a container in its own pod, with workflows defined as directed acyclic graphs (DAGs) or sequential steps in YAML. It is part of the Argo Project alongside Argo CD and Argo Rollouts, sharing the same Kubernetes-native philosophy. Argo Workflows is commonly used for CI pipelines, batch data processing, and machine-learning pipelines, and is often discussed alongside general-purpose schedulers like Apache Airflow, though Argo Workflows runs natively as Kubernetes resources rather than a separate scheduling service. A typical pipeline might build a Docker image in one step and hand off deployment to Argo CD in a later stage.
Key Features
- Kubernetes-native workflow execution using custom resources
- DAG- and step-based workflow definitions for parallel or sequential jobs
- Artifact passing between steps via object storage
- Built-in retry, timeout, and conditional logic for pipeline steps
- UI for visualizing workflow execution and logs