feat(apps/main): add device plugin for intel gpu

qlonik · Feb 17, 2024 · eecf484 · eecf484
1 parent c611648
commit eecf484
Show file tree

Hide file tree

Showing 8 changed files with 128 additions and 0 deletions.
diff --git a/kubernetes/main/apps/tools/device-plugin-intel/app/helmrelease.yaml b/kubernetes/main/apps/tools/device-plugin-intel/app/helmrelease.yaml
@@ -0,0 +1,30 @@
+---
+# yaml-language-server: $schema=https://kubernetes-schemas.pages.dev/helm.toolkit.fluxcd.io/helmrelease_v2beta2.json
+apiVersion: helm.toolkit.fluxcd.io/v2beta2
+kind: HelmRelease
+metadata:
+  name: device-plugin-intel-operator
+spec:
+  interval: 30m
+  chart:
+    spec:
+      chart: intel-device-plugins-operator
+      version: 0.29.0
+      sourceRef:
+        kind: HelmRepository
+        name: intel
+        namespace: flux-system
+  install:
+    crds: CreateReplace
+    remediation:
+      retries: 3
+  upgrade:
+    cleanupOnFail: true
+    crds: CreateReplace
+    remediation:
+      retries: 3
+  uninstall:
+    keepHistory: false
+  dependsOn:
+    - name: node-feature-discovery
+      namespace: tools
diff --git a/kubernetes/main/apps/tools/device-plugin-intel/app/kustomization.yaml b/kubernetes/main/apps/tools/device-plugin-intel/app/kustomization.yaml
@@ -0,0 +1,6 @@
+---
+# yaml-language-server: $schema=https://json.schemastore.org/kustomization
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+  - ./helmrelease.yaml
diff --git a/kubernetes/main/apps/tools/device-plugin-intel/gpu/helmrelease.yaml b/kubernetes/main/apps/tools/device-plugin-intel/gpu/helmrelease.yaml
@@ -0,0 +1,32 @@
+---
+# yaml-language-server: $schema=https://kubernetes-schemas.pages.dev/helm.toolkit.fluxcd.io/helmrelease_v2beta2.json
+apiVersion: helm.toolkit.fluxcd.io/v2beta2
+kind: HelmRelease
+metadata:
+  name: device-plugin-intel-gpu
+spec:
+  interval: 30m
+  chart:
+    spec:
+      chart: intel-device-plugins-gpu
+      version: 0.29.0
+      sourceRef:
+        kind: HelmRepository
+        name: intel
+        namespace: flux-system
+  install:
+    remediation:
+      retries: 3
+  upgrade:
+    cleanupOnFail: true
+    remediation:
+      retries: 3
+  uninstall:
+    keepHistory: false
+  dependsOn:
+    - name: device-plugin-intel-operator
+      namespace: tools
+  values:
+    name: intel-gpu-plugin
+    sharedDevNum: 3
+    nodeFeatureRule: true
diff --git a/kubernetes/main/apps/tools/device-plugin-intel/gpu/kustomization.yaml b/kubernetes/main/apps/tools/device-plugin-intel/gpu/kustomization.yaml
@@ -0,0 +1,6 @@
+---
+# yaml-language-server: $schema=https://json.schemastore.org/kustomization
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+  - ./helmrelease.yaml
diff --git a/kubernetes/main/apps/tools/device-plugin-intel/ks.yaml b/kubernetes/main/apps/tools/device-plugin-intel/ks.yaml
@@ -0,0 +1,42 @@
+---
+# yaml-language-server: $schema=https://kubernetes-schemas.pages.dev/kustomize.toolkit.fluxcd.io/kustomization_v1.json
+apiVersion: kustomize.toolkit.fluxcd.io/v1
+kind: Kustomization
+metadata:
+  name: &app device-plugin-intel
+  namespace: flux-system
+spec:
+  path: ./kubernetes/main/apps/tools/device-plugin-intel/app
+  targetNamespace: tools
+  commonMetadata:
+    labels:
+      app.kubernetes.io/name: *app
+  prune: true
+  sourceRef:
+    kind: GitRepository
+    name: home-kubernetes
+  wait: false
+  interval: 30m
+  retryInterval: 1m
+  timeout: 5m
+---
+# yaml-language-server: $schema=https://kubernetes-schemas.pages.dev/kustomize.toolkit.fluxcd.io/kustomization_v1.json
+apiVersion: kustomize.toolkit.fluxcd.io/v1
+kind: Kustomization
+metadata:
+  name: &app device-plugin-intel-gpu
+  namespace: flux-system
+spec:
+  path: ./kubernetes/main/apps/tools/device-plugin-intel/gpu
+  targetNamespace: tools
+  commonMetadata:
+    labels:
+      app.kubernetes.io/name: *app
+  prune: true
+  sourceRef:
+    kind: GitRepository
+    name: home-kubernetes
+  wait: false
+  interval: 30m
+  retryInterval: 1m
+  timeout: 5m
diff --git a/kubernetes/main/apps/tools/kustomization.yaml b/kubernetes/main/apps/tools/kustomization.yaml
@@ -5,6 +5,7 @@ resources:
   - ./namespace.yaml
   - ./descheduler/ks.yaml
   - ./device-plugin-amd-gpu/ks.yaml
+  - ./device-plugin-intel/ks.yaml
   - ./node-feature-discovery/ks.yaml
   - ./reloader/ks.yaml
   - ./system-upgrade-controller/ks.yaml
diff --git a/kubernetes/main/flux/repositories/helm/intel.yaml b/kubernetes/main/flux/repositories/helm/intel.yaml
@@ -0,0 +1,10 @@
+---
+# yaml-language-server: $schema=https://kubernetes-schemas.pages.dev/source.toolkit.fluxcd.io/helmrepository_v1beta2.json
+apiVersion: source.toolkit.fluxcd.io/v1beta2
+kind: HelmRepository
+metadata:
+  name: intel
+  namespace: flux-system
+spec:
+  interval: 2h
+  url: https://intel.github.io/helm-charts
diff --git a/kubernetes/main/flux/repositories/helm/kustomization.yaml b/kubernetes/main/flux/repositories/helm/kustomization.yaml
@@ -20,6 +20,7 @@ resources:
   - ./hajimari.yaml
   - ./immich.yaml
   - ./ingress-nginx.yaml
+  - ./intel.yaml
   - ./jetstack.yaml
   - ./k8s-gateway.yaml
   - ./kubernetes-dashboard.yaml