Kubernetes Deployment Guide

This document describes how to deploy ArkFlow in a Kubernetes cluster.

Prerequisites

• Kubernetes cluster (version >= 1.16)
• kubectl command-line tool
• Built ArkFlow Docker image

Deployment Configuration

ConfigMap

First, create a ConfigMap to store the ArkFlow configuration file:

apiVersion: v1
kind: ConfigMap
metadata:
  name: arkflow-config
data:
  config.yaml: |
    # Place your ArkFlow configuration here

Deployment

apiVersion: apps/v1
kind: Deployment
metadata:
  name: arkflow
  labels:
    app: arkflow
spec:
  replicas: 1
  selector:
    matchLabels:
      app: arkflow
  template:
    metadata:
      labels:
        app: arkflow
    spec:
      containers:
      - name: arkflow
        image: arkflow:latest  # Replace with your image address
        ports:
        - containerPort: 8000
          name: http
        env:
        - name: RUST_LOG
          value: "info"
        resources:
          requests:
            cpu: "100m"
            memory: "128Mi"
          limits:
            cpu: "500m"
            memory: "512Mi"
        livenessProbe:
          httpGet:
            path: /health
            port: http
          initialDelaySeconds: 30
          periodSeconds: 10
        readinessProbe:
          httpGet:
            path: /readiness
            port: http
          initialDelaySeconds: 5
          periodSeconds: 5
        volumeMounts:
        - name: config
          mountPath: /app/etc
          readOnly: true
      volumes:
      - name: config
        configMap:
          name: arkflow-config

Service

apiVersion: v1
kind: Service
metadata:
  name: arkflow
spec:
  selector:
    app: arkflow
  ports:
  - port: 8000
    targetPort: 8000
  type: ClusterIP  # Can be changed to NodePort or LoadBalancer as needed

Deployment Steps

1. Create Configuration Files

# Create namespace (optional)
kubectl create namespace arkflow

# Apply ConfigMap
kubectl apply -f configmap.yaml

2. Deploy Application

# Deploy Deployment
kubectl apply -f deployment.yaml

# Deploy Service
kubectl apply -f service.yaml

3. Verify Deployment

# Check Pod status
kubectl get pods -l app=arkflow

# Check service status
kubectl get svc arkflow

Configuration Details

• Image Configuration: In the Deployment configuration, replace image: arkflow:latest with your actual image address
• Environment Variables: Environment variables can be configured through the env field, currently configured with RUST_LOG=info
• Port Configuration: Service exposes port 8000 by default
• Configuration File: Mounted to the container's /app/etc directory via ConfigMap
• Resource Limits: Default resource requests and limits are set to prevent resource contention
• Health Checks: Liveness and readiness probes are configured to ensure proper application health monitoring

Important Notes

1. Ensure the configuration file format in ConfigMap is correct
2. Adjust the number of replicas according to actual needs
3. Choose appropriate Service type based on your environment
4. Adjust resource limits according to your application's actual resource consumption
5. Modify health check endpoints to match your application's actual health check endpoints

Persistent Storage (Optional)

If your ArkFlow deployment requires persistent storage, you can add a PersistentVolumeClaim:

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: arkflow-data
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 1Gi

Then update your Deployment to use this PVC:

# Add to the volumes section
volumes:
- name: data
  persistentVolumeClaim:
    claimName: arkflow-data

# Add to the volumeMounts section of your container
volumeMount:
- name: data
  mountPath: /app/data

Troubleshooting

If the service fails to run properly after deployment, you can check the issues using these commands:

# View Pod logs
kubectl logs -l app=arkflow

# View detailed Pod information
kubectl describe pod -l app=arkflow