Merge pull request #19 from oracle-quickstart/rg_plugin_helm

Helm for lens plugin

Author: amgowda-oci

oci-scanner-plugin-amd-helm/Chart.yaml

@@ -0,0 +1,16 @@
+apiVersion: v2
+name: oci-scanner-plugin
+description: OCI Scanner Plugin for AMD GPU monitoring and health checks
+type: application
+version: 0.1.0
+appVersion: "1.0.0"
+keywords:
+  - oci
+  - amd
+  - gpu
+  - monitoring
+  - prometheus
+home: https://github.com/oracle/oci-lens-quickstart
+maintainers:
+  - name: OCI Lens Team
+    email: [email protected]

oci-scanner-plugin-amd-helm/README.md

@@ -0,0 +1,33 @@
+# OCI Scanner Plugin Helm Chart
+
+AMD GPU monitoring and health check solution for OCI compute instances.
+
+## Components
+
+- **Go Plugin**: Main metric collector
+- **Node Exporter**: System metrics
+- **AMD GPU Exporter**: GPU metrics
+- **Metrics Push Job**: Automated metrics forwarding
+- **Health Check**: GPU performance testing (optional)
+- **DRHPC**: DRHPC monitoring (optional)
+
+## Configuration
+
+```bash
+# Custom push gateway
+helm install oci-scanner-plugin ./oci-scanner-plugin-helm \
+  --set global.pushGatewayUrl=http://my-pushgateway:9091/
+
+# Enable health check
+helm install oci-scanner-plugin ./oci-scanner-plugin-helm \
+  --set healthCheck.enabled=true
+
+# Uninstall
+helm uninstall oci-scanner-plugin
+```
+
+## Requirements
+
+- Kubernetes cluster with AMD GPU nodes
+- Prometheus Push Gateway accessible from cluster
+- AMD GPU drivers installed on nodes

oci-scanner-plugin-amd-helm/templates/amd-gpu-exporter.yaml

@@ -0,0 +1,139 @@
+{{- if .Values.amdGpuExporter.enabled }}
+---
+apiVersion: apps/v1
+kind: DaemonSet
+metadata:
+  name: amd-gpu-exporter
+  namespace: {{ .Values.namespace.name }}
+  labels:
+    app: amd-gpu-exporter
+    component: gpu-monitoring
+spec:
+  selector:
+    matchLabels:
+      app: amd-gpu-exporter
+  template:
+    metadata:
+      labels:
+        app: amd-gpu-exporter
+        component: gpu-monitoring
+      annotations:
+        prometheus.io/scrape: "true"
+        prometheus.io/port: "5000"
+        prometheus.io/path: "/metrics"
+    spec:
+      nodeSelector:
+        kubernetes.io/arch: amd64
+      
+      hostNetwork: true
+      
+      volumes:
+      - name: dri
+        hostPath:
+          path: /dev/dri
+          type: Directory
+      - name: kfd
+        hostPath:
+          path: /dev/kfd
+          type: Directory
+      - name: sys
+        hostPath:
+          path: /sys
+          type: Directory
+      - name: proc
+        hostPath:
+          path: /proc
+          type: Directory
+      - name: var-lib-dpkg
+        hostPath:
+          path: /var/lib/dpkg
+          type: Directory
+      
+      containers:
+      - name: amd-gpu-exporter
+        image: "{{ .Values.amdGpuExporter.image.repository }}:{{ .Values.amdGpuExporter.image.tag }}"
+        imagePullPolicy: {{ .Values.amdGpuExporter.image.pullPolicy }}
+        ports:
+        - name: metrics
+          containerPort: 5000
+          protocol: TCP
+        
+        volumeMounts:
+        - name: dri
+          mountPath: /dev/dri
+        - name: kfd
+          mountPath: /dev/kfd
+        - name: sys
+          mountPath: /sys
+          readOnly: true
+        - name: proc
+          mountPath: /proc
+          readOnly: true
+        - name: var-lib-dpkg
+          mountPath: /var/lib/dpkg
+          readOnly: true
+        
+        securityContext:
+          privileged: false
+          readOnlyRootFilesystem: false
+          allowPrivilegeEscalation: false
+          capabilities:
+            add:
+            - SYS_ADMIN
+            - NET_ADMIN
+        
+        resources:
+          {{- toYaml .Values.amdGpuExporter.resources | nindent 10 }}
+        
+        livenessProbe:
+          httpGet:
+            path: /health
+            port: 5000
+          initialDelaySeconds: 60
+          periodSeconds: 30
+          timeoutSeconds: 10
+          failureThreshold: 3
+        
+        readinessProbe:
+          httpGet:
+            path: /health
+            port: 5000
+          initialDelaySeconds: 30
+          periodSeconds: 10
+          timeoutSeconds: 5
+          failureThreshold: 3
+        
+        env:
+        - name: ROCR_VISIBLE_DEVICES
+          value: "all"
+        - name: HSA_OVERRIDE_GFX_VERSION
+          value: "11.0.0"
+      
+      tolerations:
+      - key: amd.com/gpu
+        operator: Equal
+        value: "present"
+        effect: NoSchedule
+      
+      restartPolicy: Always
+      terminationGracePeriodSeconds: 30
+
+---
+apiVersion: v1
+kind: Service
+metadata:
+  name: amd-gpu-exporter
+  namespace: {{ .Values.namespace.name }}
+  labels:
+    app: amd-gpu-exporter
+    component: gpu-monitoring
+spec:
+  selector:
+    app: amd-gpu-exporter
+  ports:
+  - name: metrics
+    port: 5000
+    targetPort: 5000
+    protocol: TCP
+  type: ClusterIP
+{{- end }}

oci-scanner-plugin-amd-helm/templates/drhpc.yaml

@@ -0,0 +1,70 @@
+{{- if .Values.drhpc.enabled }}
+---
+apiVersion: apps/v1
+kind: DaemonSet
+metadata:
+  name: oci-lens-dr-hpc
+  namespace: {{ .Values.namespace.name }}
+  labels:
+    app: oci-lens-dr-hpc
+    component: gpu-monitoring
+spec:
+  selector:
+    matchLabels:
+      app: oci-lens-dr-hpc
+  template:
+    metadata:
+      labels:
+        app: oci-lens-dr-hpc
+    spec:
+      nodeSelector:
+        kubernetes.io/arch: amd64
+      
+      affinity:
+        nodeAffinity:
+          requiredDuringSchedulingIgnoredDuringExecution:
+            nodeSelectorTerms:
+            - matchExpressions:
+              - key: amd.com/gpu
+                operator: In
+                values: ["true", "present"]
+      
+      tolerations:
+      - key: amd.com/gpu
+        operator: Exists
+        effect: NoSchedule
+
+      priorityClassName: system-node-critical
+      terminationGracePeriodSeconds: 0
+      
+      volumes: 
+      - name: root
+        hostPath:
+          path: "/"
+      
+      hostPID: true
+      hostIPC: true
+      hostNetwork: true
+      
+      containers:
+      - name: oke-dr-hpc-prod
+        image: "{{ .Values.global.imageRegistry }}/{{ .Values.drhpc.image.repository }}:{{ .Values.drhpc.image.tag }}"
+        imagePullPolicy: {{ .Values.drhpc.image.pullPolicy }}
+        
+        securityContext:
+          privileged: true
+          capabilities:
+            add: [SYS_ADMIN]
+        
+        volumeMounts: 
+        - name: root
+          mountPath: /host
+        
+        resources: {}
+        
+        env:
+        - name: PUSH_GATEWAY
+          value: {{ .Values.global.pushGatewayUrl | quote }}
+        - name: JOB_NAME
+          value: "oci_lens_drhpc_metrics"
+{{- end }}