N8N - Kubernetes complete configuration

n8n-deployment.yml

---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: n8n-deployment
  namespace: standard
  labels: &labels
    app: n8n
    component: deployment
spec:
  replicas: 1
  selector:
    matchLabels: *labels
  template:
    metadata:
      labels: *labels
      annotations:
        prometheus.io/scrape: "true"
        prometheus.io/port: "5678"
    spec:
      containers:
      - name: n8n
        image: n8nio/n8n
        imagePullPolicy: IfNotPresent
        ports:
        - name: http-metrics
          containerPort: 5678
        envFrom:
        - secretRef:
            name: n8n-secrets
        livenessProbe:
          httpGet:
            path: /healthz
            port: 5678
        readinessProbe:
          httpGet:
            path: /healthz
            port: 5678
        resources:
          limits:
            cpu: "1.0"
            memory: "1024Mi"
          requests:
            cpu: "0.5"
            memory: 512Mi

n8n-ingress.yml

---
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
  name: n8n-ingress
  namespace: standard
  labels: &labels
    app: n8n
    component: ingress
  annotations:
    kubernetes.io/ingress.class: "nginx"
spec:
  tls:
  - hosts:
    - your.domain.com
    secretName: {{ YOUR_SECRET_TLS_CONFIG }}
  rules:
  - host: your.domain.com
    http:
      paths:
      - path: /
        backend:
          serviceName: n8n-service
          servicePort: 80

n8n-secrets.yml

apiVersion: v1
kind: Secret
type: Opaque
metadata:
  name: n8n-secrets
  namespace: standard
  labels:
    app: n8n
    component: secrets
stringData:
  DB_TYPE: "postgresdb"
  DB_POSTGRESDB_USER: "n8n"
  DB_POSTGRESDB_DATABASE: "n8n"
  DB_POSTGRESDB_PASSWORD: "{{ PASSWORD }}"
  DB_POSTGRESDB_HOST: "{{ HOST }}"
  DB_POSTGRESDB_PORT: "5432"
  # Basic auth credentials
  N8N_BASIC_AUTH_ACTIVE: "true"
  N8N_BASIC_AUTH_USER: "n8n"
  N8N_BASIC_AUTH_PASSWORD: "{{ PASSWORD }}"

  N8N_HOST: "your.domain.com"
  N8N_ENCRYPTION_KEY: "{{ PASSWORD }}"
  GENERIC_TIMEZONE: "Europe/Lisbon"
  WEBHOOK_TUNNEL_URL: "https://your.domain.com/"
  NODE_ENV: "production"
  N8N_METRICS: "true"
  # Increase node max memory
  NODE_OPTIONS: "--max_old_space_size=1024"
  # Set n8n to work as single thread instead of forking to worker threads
  EXECUTIONS_PROCESS: "main"

n8n-service.yml

---
apiVersion: v1
kind: Service
metadata:
  name: n8n-service
  namespace: standard
  annotations:
    prometheus.io/probe: "true"
    prometheus.io/probe-path: "/healthz"
  labels:
    app: n8n
    component: service
spec:
  type: ClusterIP
  selector:
    app: n8n
    component: deployment
  ports:
  - protocol: TCP
    name: http
    port: 80
    targetPort: 5678

postgres.yml

---
apiVersion: v1
kind: Service
metadata:
  name: n8n-postgres
  namespace: standard
  labels: &labels
    app: n8n
    component: database
spec:
  ports:
    - name: postgres
      port: 5432
      targetPort: 5432
  clusterIP: None
  selector:
    app: n8n
    component: database
---
apiVersion: apps/v1beta2
kind: StatefulSet
metadata:
  name: n8n-postgres
  namespace: standard
  labels: &labels
    app: n8n
    component: database
spec:
  serviceName: "n8n-postgres"
  replicas: 1
  selector:
    matchLabels: *labels
  template:
    metadata:
      labels: *labels
    spec:
      containers:
      - name: postgresql
        image: postgres:10
        ports:
        - name: postgres
          containerPort: 5432
        env:
        - name: PGDATA
          value: /var/lib/postgresql/data/pgdata
        - name: POSTGRES_USER
          value: n8n
        - name: POSTGRES_DB
          value: n8n
        - name: POSTGRES_PASSWORD
          value: {{ PASSWORD }}

        volumeMounts:
        - name: data
          mountPath: /var/lib/postgresql/data

  volumeClaimTemplates:
  - metadata:
      name: data
    spec:
      accessModes: [ "ReadWriteOnce" ]
      storageClassName: "gp2"
      resources:
        requests:
          storage: 1Gi

Do you know why a path like "/n8n" won't work?
Thank you for this btw, helped guide my configuration.

Hi @Kokolokoli
Could you elaborate more on the problem? Have you tried to run docker run -it --rm n8nio/n8n /bin/bash?

Simply put, I have set up n8n in a k8s pod and it runs with no issue. I have made it accessible through the kubernetes maintained Nginx Ingress Controller and the following ingress ressource:

# Ingress config file 
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: nginx-ingress
spec:
  ingressClassName: nginx
  tls:
  - hosts: 
      - subdomain.myhost.com
  rules:
  - host: subdomain.myhost.com
    http:
      paths:
        - path: /
          pathType: Prefix
          backend:
            service:
              name: n8n-service
              port: 
                number: 8080

and i can access it via my subdomain.myhost.com (although I notice it adds a "/workflow" after the url as if n8n redirected me).
I have ran it with docker as well in the past and worked out nicely. I also made it work with a NodePort service instead of the current clusterIP service + Ingress.

However, if I change the path in my ingress ressource file from "/" to "/n8n":

• accessing the host returns 404 (as it should)
• accessing subdomain.myhost.com/n8n returns a blank page with no redirection to "subdomain.myhost.com/n8n/workflow" and no 404 message (understandable since nginx finds the path to send the request to).

What might be causing this blank page?

hmm got it
have you tried changing ENV vars?
have a look at the configuration doc https://docs.n8n.io/reference/environment-variables.html#deployment

Sorry for the late reply, haven't had much time to try it as the focus of my tasks somewhat switched. That is a very interesting idea! I think this might work, I will let you know how it goes.
Thank you!

Well, I guess it didn't work.
Tried changing the N8N_PATH to subdomain.myhost.com/n8n and it didn't work; now both the subdomain.myhost.com and subdomain.myhost.com/n8n return blank pages...

I also tried N8N_HOST=subdomain.myhost.com/ && N8N_PATH=/n8n, same result.