Getting "connection lost" error in a deployed n8n instance

[Shikhar_Ashish]

If your question hasn’t been asked before, please follow the template below. Skip the questions that are not relevant to you. →

Hi everyone,

I deployed an n8n instance using this tutorial on YouTube. (Ubuntu + Vultr + NPM). The instance is deployed but when i see “connection lost” message when i try to create a workflow.

Did anyone else face this issue before? please help me in resolving this.

[johnnyn8n]

Try this in your NGINX/NPM advanced config for the n8n host:

• WebSocket support

  • proxy_http_version 1.1;

  • proxy_set_header Upgrade $http_upgrade;

  • proxy_set_header Connection “upgrade”;

• Forward original client info

  • proxy_set_header Host $host;

  • proxy_set_header X-Real-IP $remote_addr;

  • proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;

  • proxy_set_header X-Forwarded-Proto $scheme;

• Disable caching/buffering and avoid timeouts

  • chunked_transfer_encoding off;

  • proxy_buffering off;

  • proxy_cache off;

  • proxy_read_timeout 3600;

  • proxy_send_timeout 3600;

These settings have resolved the exact “connection lost” symptom for multiple users behind NGINX/NPM and Kubernetes ingress (enable WebSocket support in NPM’s UI if you use it)

You can check here:

• Connection lost using Nginx reverse Proxy, Self hosted with Nginx Proxy Manager
• First Workflow Issue: WebSockets on NGINX, Docker on subpath fix

If you’re on a subpath (e.g., /n8n), also set:

• N8N_PATH=/n8n/

• WEBHOOK_URL=https://yourdomain.com/n8n/

• N8N_HOST=yourdomain.com

• N8N_PROTOCOL=https . And do not run with the tunnel flag, which overrides webhook URL selection

[Shikhar_Ashish]

Thanks a lot!