Render Service Management

Render Service Management - Shared Skill

Overview

This skill provides standardized patterns for managing Render platform services, including service provisioning, configuration management, and lifecycle operations.

When to Use

• When creating new Render services
• When configuring render.yaml files
• When provisioning Render databases
• When managing service scaling
• When setting up auto-deployments
• When handling Render-specific deployments

Core Procedures

Service Creation

# render.yaml - Web Service
services:
  - type: web
    name: my-service
    env: node  # or python, go, ruby, etc.
    region: oregon  # or frankfurt, singapore, etc.
    branch: main
    repo: https://github.com/org/repo
    plan: starter  # starter, pro, performance
    
    # Build & Run
    buildCommand: npm ci && npm run build
    startCommand: npm start
    healthCheckPath: /health
    
    # Auto-deploy
    autoDeploy: true
    
    # Scaling
    numInstances: 1
    
    # Environment
    envVars:
      - key: NODE_ENV
        value: production
      - key: API_KEY
        sync: false  # don't sync between environments

Database Provisioning

# render.yaml - PostgreSQL
databases:
  - name: my-database
    databaseName: myapp_db
    plan: starter  # starter, pro
    color: blue
    ipAllowList: []  # restrict access

# render.yaml - Redis
backgroundWorkers:
  - name: my-worker
    branch: main
    repo: https://github.com/org/repo
    buildCommand: pip install -r requirements.txt
    startCommand: python worker.py
    numInstances: 1

Environment Variable Sync

envVars:
  # Sync'd across all environments
  - key: COMMON_SETTING
    value: "value"
  
  # Environment-specific (not synced)
  - key: DATABASE_URL
    fromDatabase:
      name: my-database
      property: connectionString
  
  - key: API_KEY
    sync: false

Custom Domain Configuration

domains:
  - alias: www.example.com
    type: redirect
  - alias: app.example.com
    type: primary

Service Scaling

Auto-Scaling Configuration

services:
  - type: web
    name: my-service
    env: node
    branch: main
    
    # Manual scaling
    numInstances: 2
    
    # Auto-scaling (Pro plan)
    autoScaling:
      minInstances: 1
      maxInstances: 5
      targetCPUPercent: 70
      targetMemoryPercent: 80

Health Check Configuration

healthCheckPath: /health

# Custom health check in code:
app.get('/health', (req, res) => {
  const healthy = checkDependencies();
  res.status(healthy ? 200 : 503).json({
    status: healthy ? 'healthy' : 'unhealthy',
    timestamp: new Date().toISOString()
  });
});

Deployment Patterns

Branch-Based Deployment

services:
  - type: web
    name: production
    branch: main
    autoDeploy: true
  
  - type: web
    name: staging
    branch: develop
    autoDeploy: true

Environment Separation

Production: main branch → production service
Development: dev branch → dev service (or staging)
Preview: feature/* branches → preview services

Monitoring & Logs

Log Access

# Via CLI
render logs my-service

# Via Dashboard
# Render Dashboard → Service → Logs

Metrics

metrics_available:
  - CPU usage
  - Memory usage
  - Request count
  - Latency
  - Error rate

monitoring:
  - Render Dashboard
  - External: Datadog, New Relic, etc.

Common Issues & Solutions

Build Failures

troubleshooting:
  - Check build logs in Render Dashboard
  - Verify Node version matches local
  - Ensure environment variables are set
  - Check for missing dependencies

Slow Response

solutions:
  - Upgrade plan for more CPU/memory
  - Increase instance count
  - Enable auto-scaling
  - Check database query performance
  - Add caching layer (Redis)

Deployment Not Triggering

checklist:
  - [ ] Auto-deploy enabled
  - [ ] Correct branch name
  - [ ] Valid GitHub repo connection
  - [ ] Push event occurred

Pricing Reference

Best Practices

1. Use render.yaml for reproducible deployments
2. Enable health checks for auto-restart on failure
3. Set up auto-scaling for production services
4. Use environment variables for secrets, not in code
5. Monitor logs and set up alerting
6. Test locally before deploying to Render
7. Use staging for testing before production