inference.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: inferencemodule
  namespace: {{gitops.namespace}}
  labels:
    app.kubernetes.io/name: inferencemodule
    app.kubernetes.io/managed-by: kubectl
spec:
  replicas: 1
  selector:
    matchLabels:
      app: inferencemodule
  minReadySeconds: 10
  strategy:
    type: RollingUpdate
    rollingUpdate:
      maxSurge: 1
      maxUnavailable: 1
  template:
    metadata:
      labels:
        app: inferencemodule
      annotations:
        gitops.version: v-{{gitops.version}}
    spec:
      nodeSelector:
        "beta.kubernetes.io/os": linux
      initContainers:
        - name: wait-for-rtsp
          image: busybox
          command: ["sh", "-c", "until nslookup rtspsim.voe.svc.cluster.local > /dev/null; do echo Waiting for rtspsim module; sleep 2; done;"]
        - name: wait-for-cvcapture
          image: curlimages/curl
          command: ["/bin/sh", "-c", 'until [ $(curl --max-time 5 -kw  "%{http_code}" "http://cvcapturemodule:9000/healthz" -o /dev/null) -ne 200 ]; do sleep 2; echo "Health check failed . Waiting for the service..."; done']
        - name: wait-for-upload
          image: curlimages/curl
          command: ["/bin/sh", "-c", 'until [ $(curl --max-time 5 -kw "%{http_code}" "http://uploadmodule:7000/healthz" -o /dev/null) -ne 200 ]; do sleep 2; echo "Health check failed . Waiting for the service..."; done']
        - name: wait-for-predict
          image: curlimages/curl
          command: ["/bin/sh", "-c", 'until [ $(curl --max-time 5 -kw "%{http_code}" "http://predictmodule:7777/healthz" -o /dev/null) -ne 200 ]; do sleep 2; echo "Health check failed . Waiting for the service..."; done']
      containers:
      - name: inferencemodule
        image: ghcr.io/cse-labs/pib-voe-inference:latest
        imagePullPolicy: Always
        ports:
        - containerPort: 5000
        livenessProbe:
          httpGet:
            path: /healthz
            port: 5000
          initialDelaySeconds: 30
          periodSeconds: 60
        readinessProbe:
          httpGet:
            path: /readyz
            port: 5000
          initialDelaySeconds: 30
          periodSeconds: 60
          failureThreshold: 3
        resources:
          requests:
            cpu: 2000m
            memory: 400Mi
          limits:
            cpu: 3000m
            memory: 1500Mi
        env:
          - name: IOTHUB_CONNECTION_STRING
            valueFrom:
              secretKeyRef:
                name: azure-env
                key: IOTHUB_CONNECTION_STRING
          - name: IOTEDGE_DEVICE_CONNECTION_STRING
            valueFrom:
              secretKeyRef:
                name: azure-env
                key: IOTEDGE_DEVICE_CONNECTION_STRING
          - name: IS_OPENCV
            value: "true"
          - name: IS_K8S
            value: "true"

---
apiVersion: v1
kind: Service
metadata:
  name: inferencemodule
  namespace: {{gitops.namespace}}
  labels:
    app.kubernetes.io/name: inferencemodule
    app.kubernetes.io/managed-by: kubectl
spec:
  ports:
  - port: 5000
  selector:
    app: inferencemodule