gpu-memory-exporter/k8s-deploy.yaml

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: gpu-memory-exporter
  namespace: monitoring
  labels:
    app.kubernetes.io/component: gpu-memory-exporter
    app.kubernetes.io/instance: gpu-memory-exporter
    app.kubernetes.io/name: gpu-memory-exporter
    app: gpu-memory-exporter
spec:
  selector:
    matchLabels:
      app.kubernetes.io/component: gpu-memory-exporter
      app.kubernetes.io/instance: gpu-memory-exporter
      app.kubernetes.io/name: gpu-memory-exporter
      app: gpu-memory-exporter
  template:
    metadata:
      labels:
        app.kubernetes.io/component: gpu-memory-exporter
        app.kubernetes.io/instance: gpu-memory-exporter
        app.kubernetes.io/name: gpu-memory-exporter
        app: gpu-memory-exporter
    spec:
      affinity:
        nodeAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            nodeSelectorTerms:
              - matchExpressions:
                  - key: qgpu-device-enable
                    operator: In
                    values:
                      - enable
      containers:
        - env:
            - name: MY_NODE_NAME
              valueFrom:
                fieldRef:
                  apiVersion: v1
                  fieldPath: spec.nodeName
          image: gpu-memory-exporter:v1
          imagePullPolicy: IfNotPresent
          livenessProbe:
            failureThreshold: 3
            tcpSocket:
              port: 8080
            initialDelaySeconds: 5
            periodSeconds: 5
            successThreshold: 1
            timeoutSeconds: 5
          name: gpu-memory-exporter
          ports:
            - containerPort: 8080
              name: metrics
              protocol: TCP
          readinessProbe:
            failureThreshold: 3
            tcpSocket:
              port: 8080
            initialDelaySeconds: 5
            periodSeconds: 5
            successThreshold: 1
            timeoutSeconds: 5
          resources:
            requests:
              cpu: 100m
              memory: 128Mi
          securityContext:
            capabilities:
              add:
                - SYS_ADMIN
            runAsNonRoot: false
            runAsUser: 0
          terminationMessagePath: /dev/termination-log
          terminationMessagePolicy: File
          volumeMounts:
            - mountPath: /var/lib/kubelet/pod-resources
              name: pod-gpu-resources
              readOnly: true
            - mountPath: /var/run/docker.sock
              name: sock
              readOnly: true
      dnsPolicy: ClusterFirst
      restartPolicy: Always
      schedulerName: default-scheduler
      securityContext: {}
      imagePullSecrets:
        - name: lenz-dockerimg
      terminationGracePeriodSeconds: 30
      volumes:
        - hostPath:
            path: /var/lib/kubelet/pod-resources
            type: ""
          name: pod-gpu-resources
        - hostPath:
            path: /var/run/docker.sock
            type: ""
          name: sock

---
apiVersion: v1
kind: Service
metadata:
  name: gpu-memory-exporter
  namespace: monitoring
  labels:
    app.kubernetes.io/component: gpu-memory-exporter
    app.kubernetes.io/instance: gpu-memory-exporter
    app.kubernetes.io/name: gpu-memory-exporter
spec:
  selector:
    app: gpu-memory-exporter
  ports:
    - name: http
      port: 80
      targetPort: 8080
  type: ClusterIP