Argo Workflows Module

What is This?

This module provides the Helm values configuration for deploying Argo Workflows to a Kubernetes cluster. Argo Workflows is a container-native workflow engine for orchestrating parallel jobs on Kubernetes.

When to Use This Module

• Need to run multi-step workflows (data pipelines, ML training, batch jobs)
• Want to orchestrate parallel container tasks
• Building CI/CD pipelines
• Running scheduled or event-driven jobs

What is Argo Workflows?

Argo Workflows lets you:

• Define workflows as Kubernetes YAML
• Run tasks in parallel or sequence
• Handle retries and error handling
• Pass data between workflow steps
• Integrate with Kubernetes resources

Example Use Cases:

• Data processing pipelines
• Machine learning model training
• Batch job processing
• Multi-step deployments
• Simulation workloads

What This Module Provides

• Helm Values File: Pre-configured values for Argo Workflows
• Production Settings: Resource limits, persistence, TTL
• Security: Non-root user, proper security contexts
• Monitoring: Metrics enabled

How It Works

Workflow YAML
   │
   ▼
┌─────────────────────────────┐
│  kubectl apply              │
└─────────────────────────────┘
   │
   ▼
┌─────────────────────────────┐
│  Argo Workflow Controller  │
│  - Validates workflow       │
│  - Creates pods for steps   │
│  - Manages state            │
└─────────────────────────────┘
   │
   ▼
┌─────────────────────────────┐
│  Kubernetes Pods            │
│  - Execute workflow steps   │
│  - Pass artifacts          │
└─────────────────────────────┘

Usage

Using Helm Directly

# Add Helm repository
helm repo add argo https://argoproj.github.io/argo-helm
helm repo update

# Install Argo Workflows
helm install argo-workflows argo/argo-workflows \
  -f modules/kubernetes/argo-workflows/helm-values.yaml \
  -n argo --create-namespace

Verify Installation

kubectl get pods -n argo
kubectl get svc -n argo

Access the UI

# Port forward to access UI
kubectl port-forward svc/argo-workflows-server 2746:2746 -n argo
# Open http://localhost:2746

Configuration Explained

Resource Limits

resources:
  limits:
    cpu: 500m      # Maximum CPU
    memory: 512Mi  # Maximum memory
  requests:
    cpu: 100m      # Guaranteed CPU
    memory: 128Mi  # Guaranteed memory

Workflow TTL

ttlStrategy:
  secondsAfterCompletion: 86400  # Delete after 24 hours
  secondsAfterSuccess: 86400
  secondsAfterFailure: 604800    # Keep failures for 7 days

This automatically cleans up completed workflows.

Persistence

persistence:
  archive: true           # Store workflow history
  accessMode: ReadWriteOnce
  size: 10Gi             # Storage size

Submitting Workflows

Simple Workflow

apiVersion: argoproj.io/v1alpha1
kind: Workflow
metadata:
  generateName: hello-world-
  namespace: argo
spec:
  entrypoint: whalesay
  templates:
    - name: whalesay
      container:
        image: docker/whalesay:latest
        command: [cowsay]
        args: ["Hello from Argo!"]

Apply it:

kubectl apply -f workflow.yaml
kubectl get workflows -n argo

Files

• helm-values.yaml - Helm chart configuration values
• README.md - This documentation

Requirements

• Kubernetes Cluster: Version 1.20 or higher
• Helm: Version 3.x
• kubectl: Configured to access your cluster
• Storage: Persistent volume for workflow history (optional)

Components Deployed

1. Workflow Controller: Manages workflow execution
2. Workflow Server: Provides API and UI
3. Service Account: For workflow execution
4. ConfigMaps: Configuration
5. Services: For UI and API access

Security Features

• Non-root User: Workflows run as non-root (UID 8737)
• RBAC: Proper role-based access control
• Service Accounts: Isolated service accounts per namespace
• Network Policies: Compatible with network policies

Monitoring

Metrics are enabled by default. Access them:

kubectl port-forward svc/argo-workflows-server-metrics 9090:9090 -n argo

Troubleshooting

Workflows Stuck in Pending

# Check controller logs
kubectl logs -n argo -l app=workflow-controller

# Check workflow events
kubectl describe workflow <workflow-name> -n argo

Cannot Access UI

# Check service
kubectl get svc argo-workflows-server -n argo

# Port forward as workaround
kubectl port-forward svc/argo-workflows-server 2746:2746 -n argo

Workflow Fails

# View workflow details
kubectl describe workflow <workflow-name> -n argo

# Check pod logs
kubectl logs <pod-name> -n argo

Learn More

• Argo Workflows Documentation
• Workflow Examples
• Workflow Concepts

Related Modules

• argo-workflows-airgap - Offline deployment version
• ingress-nginx - Expose UI via ingress