Telegram trigger not working

Questions
1

1.82.3
Docker

My workflow runs without errors, but at some point, it just stops working. In the executions tab, there are no trigger attempts. If I turn the workflow off and then back on, everything works fine again.

I tried reinstalling n8n on another server, but the issue remains.

As a result, I have a fully functional system, but something breaks outside the workflow. Visually, I can’t see or understand what’s causing the problem.

This issue occurs in any new workflow I create. Everything works perfectly, but at some point, it suddenly stops. On average, it runs for half a day or a full day before I have to restart the workflow to make it work again.

Has anyone experienced a similar issue? Please share any solutions or insights on how to fix this problem.

2

Step-by-Step Explanation and Solution:

Understanding the Issue

Your Telegram-triggered workflows in n8n are silently failing after ~12-24 hours, requiring a restart to function again. This is likely due to webhook deregistration (n8n stops listening for Telegram updates) or resource leaks in long-running workflows. Let’s fix it!

Step 1: Enable Debug Logs

Add these environment variables to your Docker setup to catch silent errors:

environment:  
  - N8N_LOG_LEVEL=debug  # Get detailed logs  
  - N8N_DIAGNOSTICS_ENABLED=true  # Track memory/cpu

Restart n8n and monitor logs for warnings like Webhook deregistered or Telegram API timeout.

Step 2: Switch to Polling (Temporary Workaround)

If webhooks are unstable, force Telegram to use polling instead:

  1. In your Telegram Trigger Node, set:
    "options": { "useWebhook": false }
  2. Add a Cron Node to poll every 5 minutes:
    */5 * * * *  # Cron schedule

Note: Polling is less efficient but avoids webhook issues.

Step 3: Update n8n

Upgrade to n8n v1.84+ (or latest stable). Critical fixes for webhook stability were added post-v1.82.

docker pull n8nio/n8n:latest  
# Recreate your container

Step 4: Check Webhook Heartbeats

Telegram requires webhooks to respond within 30 seconds. If your server has intermittent downtime:

  1. Use a health check service (e.g., UptimeRobot) to ping your n8n URL.
  2. Add a HTTP Request Node to /rest/health (n8n’s health endpoint) every 10 minutes.

Step 5: Database Configuration

If using SQLite (default), switch to PostgreSQL to avoid locks/timeouts:

# docker-compose.yml snippet  
environment:  
  - DB_TYPE=postgresdb  
  - DB_POSTGRESDB_DATABASE=n8n  
  - DB_POSTGRESDB_HOST=postgres  
  - DB_POSTGRESDB_USER=postgres  
  - DB_POSTGRESDB_PASSWORD=your_password

Step 6: Monitor Resource Usage

Use docker stats or tools like cAdvisor to check for:

  • Memory leaks: If n8n’s memory grows steadily, add a restart policy:
    deploy:  
  resources:  
    limits:  
      memory: 1024M  # Limit to 1GB  
  restart_policy:  
    condition: on-failure

Step 7: Re-register Webhooks Periodically

Add a Cron Node to re-register Telegram webhooks daily:

  1. Use an HTTP Request Node to call:
    POST https://api.telegram.org/bot<YOUR_TOKEN>/setWebhook?url=<YOUR_WEBHOOK_URL>
  2. Schedule it to run once a day.

Step 8: Verify Network Stability

If you’re using ngrok or Cloudflare Tunnels:

  • Upgrade to a paid plan (free tiers drop connections).
  • Switch to a static IP service like bore.pub or localhost.run.

Final Workflow Adjustments

  • Add a Catch Node to log silent failures.
  • Disable “Save Manual Runs” in workflow settings to reduce DB load.

Still stuck? Share:

  1. A docker logs n8n snippet from when the trigger died.
  2. Output of curl -X POST http://localhost:5678/rest/webhook/test/<WEBHOOK_ID>.

Hang in there!

Best,
Dandy