carlos-asg/tu-voz-en-ruta

Overview

This skill configures django-health-check for Django apps to expose monitoring endpoints, Docker HEALTHCHECKs, and Kubernetes liveness/readiness probes. It provides patterns for public-schema URL placement, built-in checks (database, migrations, cache), and how to add custom checks. The goal is reliable container orchestration, load-balancer routing, and external uptime monitoring.

How this skill works

It installs and registers django-health-check and its plugins so the app exposes /health/ (HTML or JSON) and component URLs like /health/db/ and /health/migrations/. Health checks run fast, non-authenticated validations (DB connectivity, pending migrations, cache/storage) and can register custom checks via BaseHealthCheckBackend. Docker and Kubernetes probe snippets call /health/ with short timeouts so orchestrators act on failures.

When to use it

• Deploying Django in Docker or Kubernetes and needing liveness/readiness probes
• Adding simple, machine-readable health endpoints for load balancers or uptime monitors
• Verifying DB connectivity and pending migrations in automated deployments
• Integrating monitoring systems (UptimeRobot, Prometheus via metrics)
• Multi-tenant apps that require public-schema health endpoints

Best practices

• Expose health endpoints in the public schema or routing layer so orchestration can reach them
• Keep checks lightweight and fast; avoid expensive queries or tenant-specific work
• Use database and migrations checks to detect schema drift early
• Add Docker HEALTHCHECK in the Dockerfile and match timeouts to app startup behavior
• Configure Kubernetes liveness vs readiness separately (restart vs remove from service) with conservative initialDelaySeconds
• Return JSON for programmatic checks and ensure endpoints require no authentication

Example use cases

• Add /health/ to the public URL conf so Kubernetes can probe without tenant middleware
• Use docker HEALTHCHECK CMD curl -f http://localhost:8000/health/ to mark unhealthy containers
• Register a custom TenantCountCheck to surface non-critical tenant warnings in CI or dashboards
• Monitor /health/?format=json with an external uptime service and alert on 500 or timeouts
• Combine /metrics/ (django-prometheus) and /health/ for both metrics and simple availability checks