Argo Workflows Patterns Skill

Argo Workflows Patterns

When to Use This Skill

Production patterns for Argo Workflows: reusable templates, error handling, concurrency control, workflow composition, and scheduled automation.

Implementation

Define WorkflowTemplates - Create reusable, tested building blocks
Add Error Handling - Configure retry strategies for transient failures
Control Concurrency - Use mutexes or semaphores for shared resources
Compose Workflows - Chain templates into complex pipelines
Schedule Automation - Run workflows on cron schedules

Techniques

Why Argo Workflows?

Kubernetes provides primitives (Pods, Jobs, CronJobs), but building complex automation from primitives is painful. You end up with shell scripts that check Pod status in loops, cleanup logic scattered across multiple places, and debugging that requires correlating logs from dozens of sources.

Argo Workflows provides higher-level abstractions designed for automation. Define workflows declaratively. Let the controller handle scheduling, retries, and cleanup. Visualize execution in a purpose-built UI. Focus on what the automation does, not how to orchestrate it.

Pattern Categories

| Category | Description | | ---------- | ------------- | | WorkflowTemplate Patterns | Reusable workflow definitions with error handling, volumes, and RBAC | | Concurrency Control | Mutex synchronization, semaphores, and TTL strategies | | Workflow Composition | Parent/child workflows, orchestration, and cross-workflow communication | | Scheduled Workflows | CronWorkflow patterns and GitHub integration |

Quick Start

Define WorkflowTemplates - Create reusable, tested building blocks
Add Error Handling - Configure retry strategies for transient failures
Control Concurrency - Use mutexes or semaphores for shared resources
Compose Workflows - Chain templates into complex pipelines
Schedule Automation - Run workflows on cron schedules

Troubleshooting

Workflow Stuck in Pending

Check service account permissions: kubectl describe rolebinding -n argo-workflows
Verify resource quotas: kubectl describe quota -n argo-workflows
Check node resources: kubectl top nodes
Look for mutex waits: kubectl get workflows -l workflows.argoproj.io/sync-id

Workflow Failed with RBAC Error

Verify ServiceAccount exists in workflow namespace
Check ClusterRoleBinding subjects match namespace
Use kubectl auth can-i to test permissions:

kubectl auth can-i patch deployments \
  --as=system:serviceaccount:argo-workflows:my-sa \
  -n target-namespace

Mutex Deadlock

Find workflows waiting on mutex: kubectl get workflows -l workflows.argoproj.io/sync-id
Identify the workflow holding the lock
Check if the holding workflow is stuck or failed
Terminate stuck workflows to release mutex

Prerequisites

Argo Workflows must be installed in your cluster. See the official installation guide for setup instructions.

Argo Events Setup - EventSource, EventBus, and Sensor configuration
ConfigMap as Cache - Volume mounts for zero-API reads
Event-Driven Deployments - The journey to zero-latency automation

Troubleshooting

See troubleshooting.md for common issues and solutions.

Related Patterns

Argo Events Setup
ConfigMap as Cache
Event-Driven Deployments

Agent Skills: Argo Workflows Patterns

Install this agent skill to your local

Skill Files