🚀 Render Keep-Alive Worker

A production-ready Cloudflare Worker that automatically pings your Render applications every 10 minutes (24/7) to prevent them from going to sleep on the free tier.

🎯 Problem & Solution

The Challenge: Render's free tier spins down web services after 15 minutes of inactivity, resulting in 30-50 second cold starts and slow response times for users.

The Solution: This intelligent Cloudflare Worker pings your application every 10 minutes around the clock. By running every 10 minutes, it provides a crucial 5-minute safety margin against Cloudflare cron execution jitter and network retries, ensuring your service never sleeps.

✨ Features

• 🆓 Zero Cost - Runs 100% within Cloudflare's generous 100,000 requests/day free tier (only 144 runs/day used)
• ⏰ 24/7 Intelligent Keep-Alive - Executes every 10 minutes (*/10 * * * *) to stay well within Render's 15-minute timeout
• 🩺 Universal Health Parser - Correctly validates any HTTP 200-299 response, recognizing ok: true, status: "ok" | "healthy" | "UP", success: true, and plain text bodies
• 📊 Rich Metrics Extraction - Extracts and visualizes server uptime, active connections, rooms, players, and versions
• 🔄 Automatic Root Fallback - If /healthz fails or returns 404, automatically attempts a fallback ping to / to reset Render's sleep timer
• 🛡️ Production-Grade Reliability - Built-in retries with exponential backoff, request timeouts, and timing diagnostics
• 📋 Live Status Dashboard - Visual web UI with live metrics cards, ping history table, and manual trigger button
• ⚡ Zero External Dependencies - Pure modern ES module Cloudflare Worker with native Node.js test suite

🚀 Quick Start

Prerequisites

• Node.js (v18 or later)
• Git
• Cloudflare account (free)
• Render application URL (e.g. https://songless-unlimited.onrender.com)

1. Clone & Setup

# Clone the repository
git clone https://github.com/kavinthangavel/cloudflare-render-ping.git
cd cloudflare-render-ping

# Install dependencies
npm install

2. Configure Cloudflare

# Login to Cloudflare (opens browser for authentication)
npx wrangler login

# Set your Render application URL as a secret (optional, defaults to songless-unlimited.onrender.com)
npx wrangler secret put RENDER_APP_URL
# When prompted, enter your Render URL: https://your-app.onrender.com

3. Deploy

# Deploy the worker to Cloudflare
npx wrangler deploy

4. Verify Deployment

# Test the worker manually via JSON API
curl https://your-worker.workers.dev/ping

# View the visual status dashboard
# Visit in browser: https://your-worker.workers.dev/

# View real-time logs
npx wrangler tail

# Check diagnostics and configuration
curl https://your-worker.workers.dev/debug

🔧 Configuration

All settings can be configured in wrangler.toml under [vars] or overridden via Cloudflare Dashboard secrets / environment variables.

Variable	Default	Description
RENDER_APP_URL	https://songless-unlimited.onrender.com	Base URL of your Render web service
HEALTH_ENDPOINT	/healthz	Primary health check endpoint
FALLBACK_ENDPOINT	/	Fallback endpoint to ping if /healthz fails
TIMEOUT_MS	30000	HTTP request timeout in milliseconds
RETRY_ATTEMPTS	2	Number of attempts before attempting fallback
ENABLE_FALLBACK	true	Whether to attempt fallback ping to root /

Setting Secrets via Wrangler

# Set your target application URL
npx wrangler secret put RENDER_APP_URL

# Customize endpoints (optional)
npx wrangler secret put HEALTH_ENDPOINT
npx wrangler secret put TIMEOUT_MS

Schedule: Why Every 10 Minutes?

In wrangler.toml:

[triggers]
crons = ["*/10 * * * *"]

• Render puts inactive free services to sleep after 15 minutes.
• A 14-minute cron schedule risks sleep when Cloudflare scheduled execution jitter (30-90s) or slow network retries occur.
• 10 minutes provides a reliable 5-minute safety margin.
• 10 minutes = 6 runs/hour = 144 runs/day, consuming only 0.14% of Cloudflare's 100,000 daily free requests.

📊 Endpoints

Endpoint	Method	Response	Description
/ or /status	GET	HTML	Live status dashboard with metrics cards, ping history, and manual trigger
/ping	GET / POST	JSON	Manually trigger health check and return rich diagnostics & metrics
/debug	GET	JSON	Inspect worker isolate uptime, configuration, last ping, metrics, and history

Example /ping Response

{
  "success": true,
  "message": "Health check completed successfully",
  "timestamp": "2026-11-04T07:22:28.707Z",
  "durationMs": 388,
  "status": "ok",
  "httpStatus": 200,
  "endpoint": "/healthz",
  "isFallback": false,
  "metrics": {
    "uptimeSeconds": 3509,
    "uptimeFormatted": "58m 29s",
    "version": 1,
    "connections": 1,
    "pendingImports": 0,
    "rooms": 2,
    "privateRooms": 0,
    "players": 0
  },
  "result": { ... }
}

🧪 Testing

Run the automated test suite using Node.js native test runner:

npm test

📁 Project Structure

cloudflare-render-ping/
├── src/
│   ├── index.js     # Worker entrypoint (fetch & scheduled cron handlers)
│   ├── pinger.js    # Core ping engine, response parser, metrics extractor & fallback
│   ├── state.js     # In-memory isolate state and ping history management
│   └── ui.js        # HTML status dashboard template
├── test/
│   └── pinger.test.js # Comprehensive unit tests
├── wrangler.toml    # Cloudflare Worker configuration & crons
├── package.json     # Node.js metadata & scripts
└── README.md        # Documentation

📄 License

MIT License.