production/k8s/deployment-api.yml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: ml-intern-api
  namespace: ml-intern
  labels:
    app: ml-intern-api
    version: v1
spec:
  replicas: 2
  strategy:
    type: RollingUpdate
    rollingUpdate:
      maxSurge: 1
      maxUnavailable: 0
  selector:
    matchLabels:
      app: ml-intern-api
  template:
    metadata:
      labels:
        app: ml-intern-api
        version: v1
      annotations:
        prometheus.io/scrape: "true"
        prometheus.io/port: "8000"
        prometheus.io/path: "/metrics"
    spec:
      affinity:
        podAntiAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
            - weight: 100
              podAffinityTerm:
                labelSelector:
                  matchExpressions:
                    - key: app
                      operator: In
                      values:
                        - ml-intern-api
                topologyKey: kubernetes.io/hostname
      containers:
        - name: api
          image: ml-intern:latest
          imagePullPolicy: Always
          ports:
            - containerPort: 8000
              name: http
          envFrom:
            - configMapRef:
                name: ml-intern-config
          env:
            - name: DATABASE_URL
              valueFrom:
                secretKeyRef:
                  name: ml-intern-secrets
                  key: DATABASE_URL
            - name: REDIS_URL
              valueFrom:
                secretKeyRef:
                  name: ml-intern-secrets
                  key: REDIS_URL
            - name: HF_TOKEN
              valueFrom:
                secretKeyRef:
                  name: ml-intern-secrets
                  key: HF_TOKEN
            - name: ANTHROPIC_API_KEY
              valueFrom:
                secretKeyRef:
                  name: ml-intern-secrets
                  key: ANTHROPIC_API_KEY
            - name: OPENAI_API_KEY
              valueFrom:
                secretKeyRef:
                  name: ml-intern-secrets
                  key: OPENAI_API_KEY
            - name: GROQ_API_KEY
              valueFrom:
                secretKeyRef:
                  name: ml-intern-secrets
                  key: GROQ_API_KEY
            - name: NVIDIA_API_KEY
              valueFrom:
                secretKeyRef:
                  name: ml-intern-secrets
                  key: NVIDIA_API_KEY
          resources:
            requests:
              memory: "512Mi"
              cpu: "250m"
            limits:
              memory: "2Gi"
              cpu: "2000m"
          livenessProbe:
            httpGet:
              path: /health
              port: 8000
            initialDelaySeconds: 30
            periodSeconds: 10
            timeoutSeconds: 5
            failureThreshold: 3
          readinessProbe:
            httpGet:
              path: /health
              port: 8000
            initialDelaySeconds: 5
            periodSeconds: 5
            timeoutSeconds: 3
            failureThreshold: 3
          lifecycle:
            preStop:
              exec:
                command: ["/bin/sh", "-c", "sleep 15"]
      terminationGracePeriodSeconds: 60
---
apiVersion: v1
kind: Service
metadata:
  name: ml-intern-api
  namespace: ml-intern
  labels:
    app: ml-intern-api
spec:
  type: ClusterIP
  ports:
    - port: 8000
      targetPort: 8000
      protocol: TCP
      name: http
  selector:
    app: ml-intern-api
---
apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: ml-intern-api-hpa
  namespace: ml-intern
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: ml-intern-api
  minReplicas: 2
  maxReplicas: 20
  metrics:
    - type: Resource
      resource:
        name: cpu
        target:
          type: Utilization
          averageUtilization: 70
    - type: Resource
      resource:
        name: memory
        target:
          type: Utilization
          averageUtilization: 80
  behavior:
    scaleUp:
      stabilizationWindowSeconds: 60
      policies:
        - type: Percent
          value: 100
          periodSeconds: 60
    scaleDown:
      stabilizationWindowSeconds: 300
      policies:
        - type: Percent
          value: 50
          periodSeconds: 120