Skip to main content

Kubernetes Deployment

MosBot OS can be deployed to Kubernetes using the manifests included in the mosbot-api repository.

Repository structure​

The mosbot-api repository includes Kubernetes manifests under k8s/ (Kustomize layout):

k8s/
├── base/
│   ├── kustomization.yaml
│   ├── namespace.yaml
│   ├── deployment.yaml
│   ├── service.yaml
│   ├── ingress.yaml
│   ├── secret.template.yaml  ← copy to secret.yaml, fill in values
│   └── configmap.yaml
└── overlays/
    └── production/

Deploying​

1. Configure secrets​

cp k8s/base/secret.template.yaml k8s/base/secret.yaml

Edit secret.yaml and fill in your base64-encoded values:

# Encode a value
echo -n "your-value" | base64

Required secrets:

stringData:
  DB_PASSWORD: 'your-db-password'
  JWT_SECRET: 'your-jwt-secret'
  CORS_ORIGIN: 'https://your-dashboard.example.com'

2. Apply the manifests​

kubectl apply -k k8s/base

Or for a specific overlay:

kubectl apply -k k8s/overlays/production

3. Verify​

kubectl get pods -n mosbot
kubectl get svc -n mosbot

# Check API health
kubectl port-forward -n mosbot svc/mosbot-api 3000:3000
curl http://localhost:3000/health

Ingress​

Configure an ingress to expose the API and dashboard publicly. Example with nginx ingress:

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: mosbot
  namespace: mosbot
  annotations:
    nginx.ingress.kubernetes.io/ssl-redirect: 'true'
spec:
  rules:
    - host: api-mosbot.example.com
      http:
        paths:
          - path: /
            pathType: Prefix
            backend:
              service:
                name: mosbot-api
                port:
                  number: 3000
    - host: mosbot.example.com
      http:
        paths:
          - path: /
            pathType: Prefix
            backend:
              service:
                name: mosbot-dashboard
                port:
                  number: 80
  tls:
    - hosts:
        - api-mosbot.example.com
        - mosbot.example.com
      secretName: mosbot-tls

Resource recommendations​

Use the burstable resource strategy — low requests to unblock the scheduler, high limits to allow bursting:

resources:
  requests:
    cpu: '25m'
    memory: '64Mi'
  limits:
    cpu: '500m'
    memory: '512Mi'

Database​

For production Kubernetes deployments, consider using a managed PostgreSQL service (e.g. Cloud SQL, RDS, Supabase) rather than running PostgreSQL in the cluster. This simplifies backups, scaling, and maintenance.

If running PostgreSQL in the cluster, use a StatefulSet with a persistent volume:

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: postgres
spec:
  serviceName: postgres
  replicas: 1
  template:
    spec:
      containers:
        - name: postgres
          image: postgres:15
          env:
            - name: POSTGRES_DB
              value: mosbot
            - name: POSTGRES_USER
              value: mosbot
            - name: POSTGRES_PASSWORD
              valueFrom:
                secretKeyRef:
                  name: mosbot-secrets
                  key: DB_PASSWORD
          volumeMounts:
            - name: data
              mountPath: /var/lib/postgresql/data
  volumeClaimTemplates:
    - metadata:
        name: data
      spec:
        accessModes: ['ReadWriteOnce']
        resources:
          requests:
            storage: 10Gi

OpenClaw integration in Kubernetes​

If OpenClaw also runs in Kubernetes, use in-cluster service DNS names for the integration:

# In mosbot-api secret/configmap
OPENCLAW_WORKSPACE_URL: 'http://openclaw-workspace.openclaw-personal.svc.cluster.local:8080'
OPENCLAW_GATEWAY_URL: 'http://openclaw.openclaw-personal.svc.cluster.local:18789'

See OpenClaw on Kubernetes for the full guide.

Updating​

# Update the image tag in your deployment
kubectl set image deployment/mosbot-api api=ghcr.io/bymosbot/mosbot-api:v1.2.3 -n mosbot

# Or apply updated manifests
kubectl apply -k k8s/base

Migrations run automatically on pod startup.