feat: ListNodeImageVersions + shared image gallery support

Makefile-az.mk

@@ -8,6 +8,7 @@ else
   AZURE_ACR_NAME ?= $(COMMON_NAME)
 endif
 
+AZURE_SIG_SUBSCRIPTION_ID ?= $(AZURE_SUBSCRIPTION_ID)
 AZURE_CLUSTER_NAME ?= $(COMMON_NAME)
 AZURE_RESOURCE_GROUP_MC = MC_$(AZURE_RESOURCE_GROUP)_$(AZURE_CLUSTER_NAME)_$(AZURE_LOCATION)
 
@@ -46,7 +47,8 @@ az-mkacr: az-mkrg ## Create test ACR
 az-acrimport: ## Imports an image to an acr registry
 	az acr import --name $(AZURE_ACR_NAME) --source "mcr.microsoft.com/oss/kubernetes/pause:3.6" --image "pause:3.6"
 
-az-cleanenv: az-rmnodeclaims-fin  ## Deletes a few common karpenter testing resources(pods, nodepools, nodeclaims, aksnodeclasses)
+az-cleanenv: az-rmnodeclaims-fin  ## Deletes a few common karpenter testing resources(pods, nodepools, nodeclaims, aksnodeclasses) 
+	kubectl delete deployments -n default --all
 	kubectl delete pods -n default --all
 	kubectl delete nodeclaims --all
 	kubectl delete nodepools --all
@@ -136,6 +138,11 @@ az-perm: ## Create role assignments to let Karpenter manage VMs and Network
 	az role assignment create --assignee $(KARPENTER_USER_ASSIGNED_CLIENT_ID) --scope /subscriptions/$(AZURE_SUBSCRIPTION_ID)/resourceGroups/$(AZURE_RESOURCE_GROUP)    --role "Network Contributor" # in some case we create vnet here
 	@echo Consider "make az-configure-values"!
 
+az-perm-sig: ## Create role assignments when testing with SIG images 
+	$(eval KARPENTER_USER_ASSIGNED_CLIENT_ID=$(shell az identity show --resource-group "${AZURE_RESOURCE_GROUP}" --name "${AZURE_KARPENTER_USER_ASSIGNED_IDENTITY_NAME}" --query 'principalId' -otsv))
+	az role assignment create --assignee $(KARPENTER_USER_ASSIGNED_CLIENT_ID) --role "Reader" --scope /subscriptions/$(AZURE_SIG_SUBSCRIPTION_ID)/resourceGroups/AKS-Ubuntu/providers/Microsoft.Compute/galleries/AKSUbuntu
+	az role assignment create --assignee $(KARPENTER_USER_ASSIGNED_CLIENT_ID) --role "Reader" --scope /subscriptions/$(AZURE_SIG_SUBSCRIPTION_ID)/resourceGroups/AKS-AzureLinux/providers/Microsoft.Compute/galleries/AKSAzureLinux
+
 az-perm-subnet-custom: az-perm ## Create role assignments to let Karpenter manage VMs and Network (custom VNet)
 	$(eval VNET_SUBNET_ID=$(shell az aks show --name $(AZURE_CLUSTER_NAME) --resource-group $(AZURE_RESOURCE_GROUP) | jq -r ".agentPoolProfiles[0].vnetSubnetId"))
 	$(eval KARPENTER_USER_ASSIGNED_CLIENT_ID=$(shell az identity show --resource-group "${AZURE_RESOURCE_GROUP}" --name "${AZURE_KARPENTER_USER_ASSIGNED_IDENTITY_NAME}" --query 'principalId' -otsv))

karpenter-values-template.yaml

@@ -37,6 +37,12 @@ controller:
       value: ""
     - name: AZURE_NODE_RESOURCE_GROUP
       value: ${AZURE_RESOURCE_GROUP_MC}
+
+    # managed karpenter settings 
+    - name: USE_SIG
+      value: "false"
+    - name: SIG_SUBSCRIPTION_ID
+      value: ${SIG_SUBSCRIPTION_ID}
 serviceAccount:
   name: ${KARPENTER_SERVICE_ACCOUNT_NAME}
   annotations:

pkg/cloudprovider/drift.go

@@ -145,21 +145,17 @@ func (c *CloudProvider) isImageVersionDrifted(
 
 	if vm.Properties == nil ||
 		vm.Properties.StorageProfile == nil ||
-		vm.Properties.StorageProfile.ImageReference == nil ||
-		vm.Properties.StorageProfile.ImageReference.CommunityGalleryImageID == nil ||
-		*vm.Properties.StorageProfile.ImageReference.CommunityGalleryImageID == "" {
-		logger.Debug("not using a CommunityGalleryImageID for nodeClaim %s", nodeClaim.Name)
+		vm.Properties.StorageProfile.ImageReference == nil {
 		return "", nil
 	}
+	CIGID := lo.FromPtr(vm.Properties.StorageProfile.ImageReference.CommunityGalleryImageID)
+	SIGID := lo.FromPtr(vm.Properties.StorageProfile.ImageReference.ID)
+	vmImageID := lo.Ternary(SIGID != "", SIGID, CIGID)
 
-	vmImageID := *vm.Properties.StorageProfile.ImageReference.CommunityGalleryImageID
+	var imageStub imagefamily.DefaultImageOutput
+	imageStub.PopulateImageTraitsFromID(vmImageID)
 
-	publicGalleryURL, communityImageName, _, err := imagefamily.ParseCommunityImageIDInfo(vmImageID)
-	if err != nil {
-		return "", err
-	}
-
-	expectedImageID, err := c.imageProvider.GetImageID(ctx, communityImageName, publicGalleryURL)
+	expectedImageID, err := c.imageProvider.GetLatestImageID(ctx, imageStub)
 	if err != nil {
 		return "", err
 	}

pkg/fake/communityimageversionsapi.go

@@ -18,9 +18,11 @@ package fake
 
 import (
 	"context"
+	"fmt"
 
 	"github.com/Azure/azure-sdk-for-go/sdk/azcore/runtime"
 	"github.com/Azure/azure-sdk-for-go/sdk/resourcemanager/compute/armcompute/v5"
+	"github.com/samber/lo"
 
 	"github.com/Azure/karpenter-provider-azure/pkg/providers/imagefamily"
 )
@@ -51,6 +53,17 @@ func (c *CommunityGalleryImageVersionsAPI) NewListPager(_ string, _ string, _ st
 	return runtime.NewPager(pagingHandler)
 }
 
+func (c *CommunityGalleryImageVersionsAPI) Get(_ context.Context, location string, publicGalleryName string, galleryImageName string, galleryImageVersionName string, options *armcompute.CommunityGalleryImageVersionsClientGetOptions) (armcompute.CommunityGalleryImageVersionsClientGetResponse, error) {
+	// TODO: Add case where this get doesn't work or succeed
+	return armcompute.CommunityGalleryImageVersionsClientGetResponse{
+		CommunityGalleryImageVersion: armcompute.CommunityGalleryImageVersion{
+			Identifier: &armcompute.CommunityGalleryIdentifier{
+				UniqueID: lo.ToPtr(fmt.Sprintf("/CommunityGalleries/%s/images/%s/versions/%s", publicGalleryName, galleryImageName, galleryImageVersionName)),
+			},
+		},
+	}, nil
+}
+
 func (c *CommunityGalleryImageVersionsAPI) Reset() {
 	if c == nil {
 		return