Add demo for Amazon EKS

demo/clusters/eks/create-cluster.sh

@@ -0,0 +1,40 @@
+#!/bin/bash
+
+CURRENT_DIR="$(cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd)"
+PROJECT_DIR="$(cd -- "$( dirname -- "${CURRENT_DIR}/../../../.." )" &> /dev/null && pwd)"
+
+set -ex
+set -o pipefail
+
+# We extract information from versions.mk
+function from_versions_mk() {
+    local makevar=$1
+    local value=$(grep -E "^\s*${makevar}\s+[\?:]= " ${PROJECT_DIR}/versions.mk)
+    echo ${value##*= }
+}
+DRIVER_NAME=$(from_versions_mk "DRIVER_NAME")
+
+: "${AWS_REGION:=us-east-1}"
+: "${CLUSTER_NAME:=${DRIVER_NAME}-cluster}"
+: "${EKS_VERSION:=1.33}"
+: "${INSTANCE_TYPE:=g6e.4xlarge}"
+
+export AWS_REGION
+export CLUSTER_NAME
+export EKS_VERSION
+export INSTANCE_TYPE
+
+export EKS_CP_AZS=$(aws ec2 describe-availability-zones \
+      --region ${AWS_REGION} \
+      --filters "Name=opt-in-status,Values=opt-in-not-required" \
+      --query "AvailabilityZones[?ZoneId!='use1-az3'].[ZoneName][:3]" \
+      --output text | sed 's/ /, /g; s/^/  - /')
+
+## Create eksctl configuration file
+envsubst < eksctl.yaml > ${CLUSTER_NAME}-${AWS_REGION}.yaml
+
+## Create EKS cluster using eksctl
+eksctl create cluster -f ${CLUSTER_NAME}-${AWS_REGION}.yaml --install-nvidia-plugin=false
+
+## Setup EKS cluster credentials
+aws eks update-kubeconfig --name ${CLUSTER_NAME} --region ${AWS_REGION}

demo/clusters/eks/delete-cluster.sh

@@ -0,0 +1,24 @@
+#!/bin/bash
+
+CURRENT_DIR="$(cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd)"
+PROJECT_DIR="$(cd -- "$( dirname -- "${CURRENT_DIR}/../../../.." )" &> /dev/null && pwd)"
+
+set -ex
+set -o pipefail
+
+# We extract information from versions.mk
+function from_versions_mk() {
+    local makevar=$1
+    local value=$(grep -E "^\s*${makevar}\s+[\?:]= " ${PROJECT_DIR}/versions.mk)
+    echo ${value##*= }
+}
+DRIVER_NAME=$(from_versions_mk "DRIVER_NAME")
+
+: "${CLUSTER_NAME:=${DRIVER_NAME}-cluster}"
+: "${AWS_REGION:=us-east-1}"
+
+export CLUSTER_NAME
+export AWS_REGION
+
+## Delete EKS cluster using eksctl
+eksctl delete cluster --name ${CLUSTER_NAME} --region ${AWS_REGION} --wait

demo/clusters/eks/eksctl.yaml

@@ -0,0 +1,63 @@
+apiVersion: eksctl.io/v1alpha5
+kind: ClusterConfig
+metadata:
+  name: ${CLUSTER_NAME}
+  region: ${AWS_REGION}
+  version: "${EKS_VERSION}"
+availabilityZones:
+${EKS_CP_AZS}
+managedNodeGroups:
+  - name: system
+    amiFamily: AmazonLinux2023
+    instanceType: m6g.large
+    privateNetworking: true
+    desiredCapacity: 2
+    minSize: 2
+    maxSize: 4
+    labels:
+      role: system
+    taints:
+      - key: CriticalAddonsOnly
+        effect: NoSchedule
+    updateConfig:
+      maxUnavailable: 1
+  - name: gpu
+    amiFamily: AmazonLinux2023
+    instanceType: ${INSTANCE_TYPE}
+    privateNetworking: true
+    desiredCapacity: 1
+    minSize: 1
+    maxSize: 4
+    labels:
+      role: gpu
+      nvidia.com/gpu.present: "true"
+    taints:
+      - key: nvidia.com/gpu
+        effect: NoSchedule
+    updateConfig:
+      maxUnavailable: 1
+    overrideBootstrapCommand: |
+      apiVersion: node.eks.aws/v1alpha1
+      kind: NodeConfig
+      spec:
+        kubelet:
+          config:
+            featureGates:
+              DynamicResourceAllocation: true
+cloudWatch:
+  clusterLogging:
+    enableTypes: ["*"]
+addons:
+  - name: vpc-cni
+    version: latest
+  - name: coredns
+    version: latest
+    configurationValues: |-
+      nodeSelector:
+        role: system
+  - name: eks-pod-identity-agent
+    version: latest
+  - name: kube-proxy
+    version: latest
+iam:
+  withOIDC: true

demo/clusters/eks/install-dra-driver-gpu.sh

@@ -0,0 +1,16 @@
+#!/bin/bash
+
+CURRENT_DIR="$(cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd)"
+PROJECT_DIR="$(cd -- "$( dirname -- "${CURRENT_DIR}/../../../.." )" &> /dev/null && pwd)"
+
+helm upgrade -i --create-namespace --namespace nvidia-dra-driver-gpu nvidia-dra-driver-gpu ${PROJECT_DIR}/deployments/helm/nvidia-dra-driver-gpu \
+  --set image.pullPolicy=Always \
+  --set gpuResourcesEnabledOverride=true \
+  --set controller.tolerations\[0\].key="CriticalAddonsOnly" \
+  --set controller.tolerations\[0\].operator=Exists \
+  --set controller.tolerations\[0\].effect=NoSchedule \
+  --set controller.nodeSelector.role=system \
+  --set controller.affinity=null \
+  --set kubeletPlugin.tolerations\[0\].key="nvidia.com/gpu" \
+  --set kubeletPlugin.tolerations\[0\].operator=Exists \
+  --set kubeletPlugin.tolerations\[0\].effect=NoSchedule