Kubernetes Deployment

yaml-workflow provides a Docker image and Helm chart for running workflows on Kubernetes.

Docker

Run yaml-workflow in a container without installing Python:

# Run a workflow
docker run --rm -v $(pwd)/workflows:/app/workflows ghcr.io/orieg/yaml-workflow run /app/workflows/hello.yaml

# Start the web dashboard
docker run -p 8080:8080 -v $(pwd)/workflows:/app/workflows ghcr.io/orieg/yaml-workflow

# Validate a workflow
docker run --rm -v $(pwd)/workflows:/app/workflows ghcr.io/orieg/yaml-workflow validate /app/workflows/hello.yaml --format json

Helm Chart

Deploy the web dashboard and run workflows on Kubernetes:

# Install from the repo
helm install my-workflows ./helm/yaml-workflow

# With inline workflow definitions
helm install my-workflows ./helm/yaml-workflow \
  --set-file workflows.files.pipeline\\.yaml=workflows/pipeline.yaml

# Access the dashboard
kubectl port-forward svc/my-workflows 8080:8080

Configuration

Key values in values.yaml:

Parameter	Description	Default
image.repository	Container image	ghcr.io/orieg/yaml-workflow
image.tag	Image tag	Chart appVersion
serve.port	Dashboard port	8080
workflows.files	Inline workflow YAML definitions	{}
workflows.existingConfigMap	Use an existing ConfigMap	""
runs.persistence.enabled	Enable PVC for run history	true
runs.persistence.size	Storage size	1Gi
ingress.enabled	Enable Ingress	false
cronjob.enabled	Enable scheduled runs	false
cronjob.schedule	Cron expression	""
cronjob.workflow	Workflow file to run	""

Inline Workflows

Define workflows directly in values.yaml:

workflows:
  files:
    data_pipeline.yaml: |
      name: Data Pipeline
      secrets:
        - API_KEY
      steps:
        - name: fetch
          task: http.request
          inputs:
            url: "https://api.example.com/data"
            auth:
              type: bearer
              token_env: API_KEY
        - name: process
          task: python_code
          depends_on: [fetch]
          inputs:
            code: |
              data = steps["fetch"]["result"]["json"]
              result = {"count": len(data)}

Scheduled Runs

Run workflows on a schedule via Kubernetes CronJob:

cronjob:
  enabled: true
  schedule: "0 9 * * 1-5"  # Weekdays at 9am
  workflow: "data_pipeline.yaml"
  params: "env=production"

ArgoCD Integration

The Helm chart is GitOps-ready. Point ArgoCD at your repo:

apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
  name: yaml-workflow
  namespace: argocd
spec:
  project: default
  source:
    repoURL: https://github.com/your-org/your-repo
    path: helm/yaml-workflow
    targetRevision: main
    helm:
      valueFiles:
        - values-production.yaml
  destination:
    server: https://kubernetes.default.svc
    namespace: workflows
  syncPolicy:
    automated:
      prune: true
      selfHeal: true

Workflow YAML changes committed to Git are automatically synced to the cluster.

Argo Workflows Integration

Use yaml-workflow as a step inside Argo Workflows:

apiVersion: argoproj.io/v1alpha1
kind: WorkflowTemplate
metadata:
  name: yaml-workflow-step
spec:
  templates:
    - name: run-pipeline
      container:
        image: ghcr.io/orieg/yaml-workflow:latest
        command: ["yaml-workflow", "run"]
        args:
          - /app/workflows/pipeline.yaml
          - --format
          - json
          - "env=production"
        volumeMounts:
          - name: workflows
            mountPath: /app/workflows