Workflow stuck on execution, observed error "BadRequestError: request aborted"

[Eskizo]

Describe the problem/error/question

Click on Manual trigger to execute the workflow.

The workflow has been tested, and it is working fine (when it works)

The spinning wheel appears with the “executing workflow” but not progress on any node is observed. It seems stuck for minutes.

What is the error message (if any)?

BadRequestError: request aborted
at IncomingMessage.onAborted (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/[email protected]/node_modules/raw-body/index.js:245:10)
at IncomingMessage.emit (node:events:518:28)
at IncomingMessage._destroy (node:_http_incoming:221:10)
at _destroy (node:internal/streams/destroy:122:10)
at IncomingMessage.destroy (node:internal/streams/destroy:84:5)
at abortIncoming (node:_http_server:811:9)
at socketOnClose (node:_http_server:805:3)
at Socket.emit (node:events:530:35)
at TCP. (node:net:346:12)

Please share your workflow

I am not sure if that adds any value, no node is executing when this condition appears.

Share the output returned by the last node

It didnt even executed the first node

Information on your n8n setup

• n8n version: 1.106.3
• Database (default: SQLite): PGSQL
• n8n EXECUTIONS_PROCESS setting (default: own, main): default
• Running n8n via (Docker, npm, n8n cloud, desktop app): Docker
• Operating system: ubuntu 24.04

[barn4k]

hello @Eskizo

What is your n8n configuration?

[Eskizo]

n8n:
image: n8nio/n8n
restart: unless-stopped
container_name: n8n
environment:
DB_TYPE: postgresdb
DB_POSTGRESDB_HOST: postgres
DB_POSTGRESDB_PORT: 5432
DB_POSTGRESDB_DATABASE: xxxx
DB_POSTGRESDB_USER: xxxxx
DB_POSTGRESDB_PASSWORD: xxxx
N8N_HOST: xxxx
N8N_PORT: 5678
N8N_PROTOCOL: http
N8N_SSL_KEY: “”
N8N_SSL_CERT: “”
N8N_SECURE_COOKIE: “false”
N8N_API_ALLOW_PUBLIC: true
N8N_RUNNERS_ENABLED: true
TZ: UTC

I believe runners might not be needed

[barn4k]

Do you have a reverse proxy/ALB in front of the n8n? Like nginx, traefik, and so on.

So far, I can see that you have SSL stuff, which you are not using (N8N_SSL_..)

[Eskizo]

Yes I do, I have caddy. It handles SSL among other things.

Behing Caddy everything is http

[Eskizo]

@barn4k the proxy has something to do on this issue?

[barn4k]

Yes, seems something wasn’t configured. Usually, the main issue is that the reverse proxy is not configured to properly handle wss connections

[Eskizo]

mmm that is interesting, caddy is not doing anything special, just offloading the ssl workload and adding a header to filter the access, which I dont expect to add any issue as the header is included on the browser in any petition to that host.

If anything, the port exposed to the world is not the n8n default port, the default port only works inside the local network, but I can’t see how that would be an issue.

Is there any log or debug option on n8n to research on this further?

Thanks in advance.

[barn4k]

Check the dev console (f12) when you are opening the n8n for any errors

[Eskizo]

It seems that the proxy upfront on another port is also interfering into the telemetry and diagnostics sent to n8n which are using the default port (which is correct in the internal lan), therefore the browser and the whole app is like somehow lagging or stuck. I disabled the telemetry/diagnostics and it seems to be better now.