How to enable queue mode in n8n?

The short, direct answer:
To enable queue mode in n8n, you must run n8n in “queue” mode by setting the environment variable N8N_EXECUTIONS_MODE=queue and then run the separate queue worker process with N8N_EXECUTIONS_MODE=queue and N8N_OVERRIDE_EXECUTIONS_PROCESS=main. You also need a Redis instance because queue mode uses Redis as the job broker.

 

What Queue Mode Actually Is

 

Queue mode is n8n’s way of splitting the workload into two roles:

• Main process: receives triggers and schedules jobs
• Workers: pick up jobs from Redis and execute workflows

This lets you run multiple workers, spread load across servers, and keep the main process responsive. It’s used in real production setups where workflows are heavy or need scaling.

 

How To Enable Queue Mode (Step by Step)

 

Below is the real setup you use in production. No shortcuts.

• Step 1: Deploy a Redis instance
  Queue mode requires Redis. It doesn’t have to be part of the same container. Any normal Redis works.
• Step 2: Run the n8n “main” pod/container in queue mode
  You set an environment variable so the main container does not execute workflows itself — it only queues them.

// Main n8n server
export N8N_EXECUTIONS_MODE=queue
export QUEUE_BULL_REDIS_HOST=redis
export QUEUE_BULL_REDIS_PORT=6379

n8n

• Step 3: Run at least one worker
  The worker is a separate container/process that pulls jobs from Redis and actually runs workflows.

// Worker process
export N8N_EXECUTIONS_MODE=queue
export QUEUE_BULL_REDIS_HOST=redis
export QUEUE_BULL_REDIS_PORT=6379

n8n worker

• Step 4: (Optional but recommended) Run the webhook process
  In some deployments you split the “main” server and “webhook” server. For simple setups, the main process handles webhooks already, so this is optional.

// Webhook process (optional)
export N8N_EXECUTIONS_MODE=queue
export QUEUE_BULL_REDIS_HOST=redis
export QUEUE_BULL_REDIS_PORT=6379

n8n webhook

 

What Most People Forget

 

• The worker must run n8n worker — not just “n8n”. This is the most common mistake. The main process runs “n8n”, workers run “n8n worker”.
• All components must point to the same Redis. If workers connect to another instance, nothing will execute.
• Queue mode cannot run without Redis. There is no fallback.
• Your workflows will not execute in the main process anymore. This is intended. Everything gets offloaded to workers.

 

Why Production Teams Use Queue Mode

 

In real deployments, queue mode is used when workflows are long-running, heavy, or triggered frequently. The main server stays responsive, and you can scale by simply adding more workers.

• Avoids blocking triggers or webhooks
• Allows running dozens of workers
• More stable under load

 

Minimal docker-compose Example

 

This is the cleanest working example I use in production:

version: '3.8'

services:
  redis:
    image: redis:7

  n8n-main:
    image: n8nio/n8n
    environment:
      N8N_EXECUTIONS_MODE: queue
      QUEUE_BULL_REDIS_HOST: redis
    ports:
      - "5678:5678"
    depends_on:
      - redis

  n8n-worker:
    image: n8nio/n8n
    command: n8n worker
    environment:
      N8N_EXECUTIONS_MODE: queue
      QUEUE_BULL_REDIS_HOST: redis
    depends_on:
      - redis

 

The moment this deploys, queue mode is active.