feat: added chart for AWS Inferenria operator (#133)

* feat: added chart for AWS Inferenria operator

charts/tfy-inferentia-operator/Chart.yaml

@@ -0,0 +1,6 @@
+apiVersion: v2
+name: tfy-inferentia-operator
+version: 0.0.1
+description: "Truefoundry Inferentia Operator"
+maintainers:
+  - name: truefoundry

charts/tfy-inferentia-operator/README.md

@@ -0,0 +1,15 @@
+# Tfy-inferentia-operator helm chart packaged by TrueFoundry
+Tfy-inferentia-operator is a Helm chart that facilitates the deployment and management of AWS Inferentia resources in Kubernetes clusters.
+
+Refer to https://github.com/aws-neuron/aws-neuron-sdk/blob/e0ef8a1a780ee798e7f01fe94f1235d571e211c6/src/k8/k8s-neuron-device-plugin.yml#L1
+
+## Parameters
+
+### Configuration for the device plugin responsible for node feature discovery
+
+| Name                                     | Description                                 | Value                                                  |
+| ---------------------------------------- | ------------------------------------------- | ------------------------------------------------------ |
+| `devicePlugin.enabled`                   | Enable device plugin Daemonset.             | `true`                                                 |
+| `devicePlugin.resources.requests.cpu`    | CPU request for device plugin Daemonset.    | `100m`                                                 |
+| `devicePlugin.resources.requests.memory` | Memory request for device plugin Daemonset. | `128MiB`                                               |
+| `devicePlugin.image`                     | Image to use for device plugin Daemonset.   | `public.ecr.aws/neuron/neuron-device-plugin:2.16.18.0` |

charts/tfy-inferentia-operator/templates/neuron-device-plugin.yaml

@@ -0,0 +1,114 @@
+{{- if .Values.devicePlugin }}
+kind: ClusterRole
+apiVersion: rbac.authorization.k8s.io/v1
+metadata:
+  name: {{ .Release.Name }}-neuron-device-plugin
+rules:
+- apiGroups:
+  - ""
+  resources:
+  - nodes
+  verbs:
+  - get
+  - list
+  - watch
+- apiGroups:
+  - ""
+  resources:
+  - events
+  verbs:
+  - create
+  - patch
+- apiGroups:
+  - ""
+  resources:
+  - pods
+  verbs:
+  - update
+  - patch
+  - get
+  - list
+  - watch
+- apiGroups:
+  - ""
+  resources:
+  - nodes/status
+  verbs:
+  - patch
+  - update
+---
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: {{ .Release.Name }}-neuron-device-plugin
+  namespace: {{ .Release.Namespace }}
+---
+kind: ClusterRoleBinding
+apiVersion: rbac.authorization.k8s.io/v1
+metadata:
+  name: {{ .Release.Name }}-neuron-device-plugin
+  namespace: {{ .Release.Namespace }}
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: ClusterRole
+  name: {{ .Release.Name }}-neuron-device-plugin
+subjects:
+- kind: ServiceAccount
+  name: {{ .Release.Name }}-neuron-device-plugin
+  namespace: {{ .Release.Namespace }}
+---
+apiVersion: apps/v1
+kind: DaemonSet
+metadata:
+  name: {{ .Release.Name }}-neuron-device-plugin
+  namespace: {{ .Release.Namespace }}
+spec:
+  selector:
+    matchLabels:
+      name: {{ .Release.Name }}-neuron-device-plugin-ds
+  updateStrategy:
+    type: RollingUpdate
+  template:
+    metadata:
+      labels:
+        name: {{ .Release.Name }}-neuron-device-plugin-ds
+    spec:
+      serviceAccount: neuron-device-plugin
+      tolerations:
+      {{- range .Values.devicePlugin.tolerations }}
+        - {{ . | toYaml | nindent 10 | trim }}
+      {{- else }}
+        []
+      {{- end }}
+      priorityClassName: "system-node-critical"
+      affinity:
+        {{- toYaml (index .Values "devicePlugin" "affinity") | nindent 8 }}
+      containers:
+      - image: .Values.devicePlugin.image
+        imagePullPolicy: Always
+        name: neuron-device-plugin
+        resources:
+          {{- toYaml (index .Values "devicePlugin" "resources") | nindent 10 }}
+        env:
+        - name: NODE_NAME
+          valueFrom:
+            fieldRef:
+              fieldPath: spec.nodeName
+        securityContext:
+          allowPrivilegeEscalation: false
+          capabilities:
+            drop: ["ALL"]
+        volumeMounts:
+          - name: device-plugin
+            mountPath: /var/lib/kubelet/device-plugins
+          - name: infa-map
+            mountPath: /run
+      volumes:
+        - name: device-plugin
+          hostPath:
+            path: /var/lib/kubelet/device-plugins
+        - name: infa-map
+          hostPath:
+            path: /run
+---
+{{- end }}

charts/tfy-inferentia-operator/values.yaml

@@ -0,0 +1,41 @@
+## @section Configuration for the device plugin responsible for node feature discovery
+## @param devicePlugin.enabled Enable device plugin Daemonset.
+## @param devicePlugin.resources.requests.cpu CPU request for device plugin Daemonset.
+## @param devicePlugin.resources.requests.memory Memory request for device plugin Daemonset.
+## @param devicePlugin.image Image to use for device plugin Daemonset.
+devicePlugin:
+  enabled: true
+  resources:
+    requests:
+      cpu: 100m
+      memory: 128MiB
+  image: public.ecr.aws/neuron/neuron-device-plugin:2.16.18.0
+  tolerations:
+      ## @skip devicePlugin.tolerations[0]
+      ## @skip devicePlugin.tolerations[1]
+    - key: CriticalAddonsOnly
+      operator: Exists
+    - key: aws.amazon.com/neuron
+      operator: Exists
+      effect: NoSchedule
+  affinity:
+    ## @skip devicePlugin.affinity.nodeAffinity
+    nodeAffinity:
+      requiredDuringSchedulingIgnoredDuringExecution:
+        nodeSelectorTerms:
+          - matchExpressions:
+              - key: "node.kubernetes.io/instance-type"
+                operator: In
+                values:
+                  - inf1.xlarge
+                  - inf1.2xlarge
+                  - inf1.6xlarge
+                  - inf1.24xlarge
+                  - inf2.xlarge
+                  - inf2.4xlarge
+                  - inf2.8xlarge
+                  - inf2.24xlarge
+                  - inf2.48xlarge
+                  - trn1.2xlarge
+                  - trn1.32xlarge
+                  - trn1n.32xlarge