Workflow not getting triggered even after being in active state, Docker setup

Questions
1

Hi community,

I’m running n8n 1.111.0 self-hosted on an AWS EC2 instance with Postgres as the database (via Docker).

The issue

When I create a workflow (either via UI or API) and mark it as active:

It appears active in the UI.

But when I hit the production webhook URL, I always get:

{
“code”: 404,
“message”: “The requested webhook "POST " is not registered.”
}

What I tried

On my local Docker setup, everything works fine (workflows register, webhooks trigger).

On EC2, no workflow (UI-created or API-created) ever registers.

Tried with both API key and session cookie for API requests.

Env vars include:

N8N_USER_MANAGEMENT_DISABLED=false

N8N_SECURE_COOKIE=false

NODE_ENV=production

Question

Why would workflows show as active but never actually register webhooks in runtime?

Could this be related to environment variables, queue mode, or something else?

2

Hey, can you post your full list of env variables?

3

Hey, sure @krisn0x
Sharing the entire docker file with you.
version: “3.8”

services:

postgres:

image: postgres:14

restart: unless-stopped

environment:

  - POSTGRES_USER=n8n

  - POSTGRES_PASSWORD=n8n

volumes:

  - postgres_data:/var/lib/postgresql/data

  - ./init.sql:/docker-entrypoint-initdb.d/init.sql

ports:

  - "5432:5432"

networks:

  - internal

n8n-dev:

image: n8nio/n8n:1.111.0

restart: unless-stopped

environment:

  - DB_TYPE=postgresdb

  - DB_POSTGRESDB_HOST=postgres

  - DB_POSTGRESDB_PORT=5432

  - DB_POSTGRESDB_DATABASE=n8n_dev

  - DB_POSTGRESDB_USER=n8n

  - DB_POSTGRESDB_PASSWORD=n8n

  - N8N_BASIC_AUTH_ACTIVE=true

  - N8N_BASIC_AUTH_USER=dev

  - N8N_BASIC_AUTH_PASSWORD=dev123

  - N8N_HOST=0.0.0.0

  - N8N_PORT=5678

  - N8N_PROTOCOL=http

  - WEBHOOK_URL=http://dev.n8n.local:5678/

  - N8N_SECURE_COOKIE=false

  - N8N_USER_MANAGEMENT_DISABLED=true   # 👈 disables teams/users, makes API mode work like local

  - N8N_API_KEY=supersecretapikey       # 👈 backend uses this to create/activate workflows

ports:

  - "5678:5678"

depends_on:

  - postgres

networks:

  - internal

n8n-prod:

image: n8nio/n8n:1.111.0

restart: unless-stopped

environment:

  - DB_TYPE=postgresdb

  - DB_POSTGRESDB_HOST=postgres

  - DB_POSTGRESDB_PORT=5432

  - DB_POSTGRESDB_DATABASE=n8n_prod

  - DB_POSTGRESDB_USER=n8n

  - DB_POSTGRESDB_PASSWORD=n8n

  - N8N_BASIC_AUTH_ACTIVE=true

  - N8N_BASIC_AUTH_USER=admin

  - N8N_BASIC_AUTH_PASSWORD=admin123

  - N8N_HOST=0.0.0.0

  - N8N_PORT=5679

  - N8N_PROTOCOL=http

  - WEBHOOK_URL=http://prod.n8n.local:5679/

  - N8N_SECURE_COOKIE=false

  - N8N_USER_MANAGEMENT_DISABLED=true   # 👈 same here

  - N8N_API_KEY=anothersecretapikey     # 👈 separate prod key

ports:

  - "5679:5679"

depends_on:

  - postgres

networks:

  - internal

volumes:

postgres_data:

networks:

internal:

driver: bridge
4

Also @krisn0x sharing one more important thing.
If I go to the UI of my n8n docker instance, do some really small changes in the workflow which I have created via API, after that the production url starts working.
It feels like the active value is being updated in db thus I am seeing that in UI, but at some place it is being missed when I trigger the API from postman/backend service.

5

Thank you, that was key. There is an open issue for this being tracked:

Please drop a comment there.

There’s also a PR in the works it seems:

Not sure if this affects everyone or not, will have to test myself (tomorrow) and report back.

For now, I will appreciate if you mark this as Solution, it will not close the thread.

6

Hey, just checking with one more idea: if you haven’t checked, can you please ensure all workers are on the same version as your main instance and redeploy them if not?

7

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.