Merge branch 'master' into codegen_v2

Author: adrianmoisey

.github/dependabot.yml

@@ -64,3 +64,74 @@ updates:
     - "area/dependency"
     - "release-note-none"
     - "ok-to-test"
+# Go - Cluster Autoscaler
+- directory: "/cluster-autoscaler"
+  package-ecosystem: "gomod"
+  open-pull-requests-limit: 5
+  schedule:
+    interval: "daily"
+    time: "15:00"
+    # Use America/New_York Standard Time (UTC -05:00)
+    timezone: "America/New_York"
+  commit-message:
+    prefix: "dependabot"
+    include: scope
+  ignore:
+    # Update providers manually.
+    - dependency-name: "k8s.io/cloud-provider-aws/*"
+    - dependency-name: "github.com/aws/*"
+    - dependency-name: "k8s.io/cloud-provider-gcp/*"
+    - dependency-name: "cloud.google.com/go/compute/*"
+    - dependency-name: "sigs.k8s.io/cloud-provider-azure/*"
+    - dependency-name: "github.com/Azure/*"
+    - dependency-name: "github.com/Azure/go-autorest/autorest/*"
+    - dependency-name: "github.com/digitalocean/*"
+    # Update dependencies exclusively used by providers manually
+    - dependency-name: "github.com/gofrs/uuid"
+    - dependency-name: "github.com/google/go-querystring"
+    - dependency-name: "github.com/jmattheis/goverter"
+    - dependency-name: "github.com/jmespath/go-jmespath"
+    - dependency-name: "github.com/vburenin/ifacemaker"
+    - dependency-name: "golang.org/x/oauth2"
+    - dependency-name: "golang.org/x/sys"
+    - dependency-name: "google.golang.org/api"
+    - dependency-name: "gopkg.in/gcfg.v1"
+    - dependency-name: "sigs.k8s.io/yaml"
+    # Maintain k8s version skew compatibility manually.
+    - dependency-name: "k8s.io/kubernetes"
+    # Maintain k8s.io staging dependencies manually.
+    - dependency-name: "k8s.io/api"
+    - dependency-name: "k8s.io/apiextensions-apiserver"
+    - dependency-name: "k8s.io/apimachinery"
+    - dependency-name: "k8s.io/apiserver"
+    - dependency-name: "k8s.io/cli-runtime"
+    - dependency-name: "k8s.io/client-go"
+    - dependency-name: "k8s.io/cloud-provider"
+    - dependency-name: "k8s.io/cluster-bootstrap"
+    - dependency-name: "k8s.io/code-generator"
+    - dependency-name: "k8s.io/component-base"
+    - dependency-name: "k8s.io/component-helpers"
+    - dependency-name: "k8s.io/controller-manager"
+    - dependency-name: "k8s.io/cri-api"
+    - dependency-name: "k8s.io/cri-client"
+    - dependency-name: "k8s.io/csi-translation-lib"
+    - dependency-name: "k8s.io/dynamic-resource-allocation"
+    - dependency-name: "k8s.io/endpointslice"
+    - dependency-name: "k8s.io/externaljwt"
+    - dependency-name: "k8s.io/kms"
+    - dependency-name: "k8s.io/kube-aggregator"
+    - dependency-name: "k8s.io/kube-controller-manager"
+    - dependency-name: "k8s.io/kube-proxy"
+    - dependency-name: "k8s.io/kube-scheduler"
+    - dependency-name: "k8s.io/kubectl"
+    - dependency-name: "k8s.io/kubelet"
+    - dependency-name: "k8s.io/metrics"
+    - dependency-name: "k8s.io/mount-utils"
+    - dependency-name: "k8s.io/pod-security-admission"
+    - dependency-name: "k8s.io/sample-apiserver"
+    - dependency-name: "k8s.io/sample-cli-plugin"
+    - dependency-name: "k8s.io/sample-controller"
+  labels:
+    - "ok-to-test"
+    - "area/cluster-autoscaler"
+    - "area/dependency"

.github/workflows/ci.yaml

@@ -43,12 +43,6 @@ jobs:
         env:
           GO111MODULE: auto
 
-      - name: golangci-lint - vertical-pod-autoscaler
-        uses: golangci/golangci-lint-action@v8
-        with:
-          args: --timeout=30m
-          working-directory: ${{ env.GOPATH }}/src/k8s.io/autoscaler/vertical-pod-autoscaler
-
       - name: Test
         working-directory: ${{ env.GOPATH }}/src/k8s.io/autoscaler
         run: hack/for-go-proj.sh test

.github/workflows/vpa-golangci-lint.yaml

@@ -0,0 +1,36 @@
+name: Lint
+
+on:
+  push:
+    paths:
+      - 'vertical-pod-autoscaler/**'
+  pull_request:
+    paths:
+      - 'vertical-pod-autoscaler/**'
+
+env:
+  GOPATH: ${{ github.workspace }}/go
+
+permissions:
+  contents: read
+
+jobs:
+  golangci:
+    name: golangci-lint - VPA
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v5
+        with:
+          path: ${{ env.GOPATH }}/src/k8s.io/autoscaler
+      - uses: actions/setup-go@v6
+        with:
+          go-version: '1.25.0'
+          cache-dependency-path: |
+             ${{ env.GOPATH}}/src/k8s.io/autoscaler/vertical-pod-autoscaler/go.sum
+             ${{ env.GOPATH}}/src/k8s.io/autoscaler/vertical-pod-autoscaler/e2e/go.sum
+
+      - name: golangci-lint - vertical-pod-autoscaler
+        uses: golangci/golangci-lint-action@v8
+        with:
+          args: --timeout=30m
+          working-directory: ${{ env.GOPATH }}/src/k8s.io/autoscaler/vertical-pod-autoscaler

cluster-autoscaler/README.md

@@ -51,6 +51,7 @@ Starting from Kubernetes 1.12, versioning scheme was changed to match Kubernetes
 
 | Kubernetes Version | CA Version               | Chart Version |
 |--------------------|--------------------------|---------------|
+| 1.34.x             | 1.34.x                   |9.51.0+|
 | 1.33.x             | 1.33.x                   |9.47.0+|
 | 1.32.x             | 1.32.x                   |9.45.0+|
 | 1.31.x             | 1.31.x                   |9.38.0+|
@@ -86,24 +87,40 @@ Starting from Kubernetes 1.12, versioning scheme was changed to match Kubernetes
 
 ## Schedule
 
-Cluster Autoscaler releases new minor versions shortly after OSS Kubernetes release
-and patches for versions corresponding to currently
-supported [Kubernetes versions](https://kubernetes.io/releases/) on a roughly 2
-month cadence. Currently planned schedule is below. Please note that target
-dates listed below are approximate and we expect up to a week difference between
-target ETA and the actual releases.
-
-| Date       | Maintainer Preparing Release | Backup Maintainer | Type  |
-|------------|------------------------------|-------------------|-------|
-| 2025-06-11 | jackfrancis                  | gjtempleton       | 1.33  |
-| 2025-07-16 | gjtempleton                  | towca             | patch |
-| 2025-08-20 | towca                        | BigDarkClown      | patch |
-| 2025-09-17 | BigDarkClown                 | x13n              | 1.34  |
-| 2025-10-22 | x13n                         | jackfrancis       | patch |
-| 2025-11-19 | jackfrancis                  | gjtempleton       | patch |
-
-Additional patch releases may happen outside of the schedule in case of critical
-bugs or vulnerabilities.
+Cluster Autoscaler synchronizes its releases with the [Kubernetes release schedule](https://kubernetes.io/releases/).
+
+For Cluster Autoscaler releases of new minor versions, expect a release date of up to
+one month after the corresponding Kubernetes release. This is due the fact that upstream
+integrations of Kubernetes into Cluster Autoscaler can't be finalized until the Kubernetes
+release is official, and the time required to test and validate those integrations.
+
+Cluster Autoscaler will also release patch versions in accordance with Kubernetes patch
+releases to ensure rapid integration of upstream Kubernetes fixes. The overhead to integrate
+and validate Kubernetes patch releases is less costly, and thus the Cluster Autoscaler
+release date should follow the corresponding Kubernetes release by no more than 1-2 weeks.
+
+Bug fixes and Cloud Provider features to Cluster Autoscaler itself will be continually
+backported into the supported release branches (n - 3, where n is the latest release).
+Backporting into older release branches can be requested as an exception by filing an issue
+and bringing the request [to the official SIG Autoscaling Community](https://github.com/kubernetes/community/blob/master/sig-autoscaling/README.md).
+
+Finally, additional Cluster Autoscaler patch releases may happen outside of the above schedule
+in case of critical bugs or vulnerabilities.
+
+In summary, users should not be guided by a strict patch version equivalency between Kubernetes
+and Cluster Autoscaler (for example, there is no strict requirement to use Cluster Autoscaler v1.34.1 w/ a Kubernetes v1.34.1 cluster). Rather, we recommend that users always use the _latest_
+Cluster Autoscaler release that corresponds to the minor version of Kubernetes that their cluster
+is running.
+
+For example, if the latest (hypothetical) Cluster Autoscaler releases are
+`v1.100.1`, `v1.99.5`, `v1.98.10`, and `v1.97.16`, any of the below scenarios follows the recommended guidance:
+
+| Kubernetes Version  | CA Version               |
+|---------------------|--------------------------|
+| 1.100.0             | 1.100.1                  |
+| 1.99.4              | 1.99.5                   |
+| 1.98.4              | 1.98.10                  |
+| 1.97.16             | 1.97.16                  |
 
 # Notable changes
 

cluster-autoscaler/cloudprovider/gce/gce_price_info.go

@@ -83,6 +83,7 @@ var (
 		"n2":  0.031611,
 		"n2d": 0.027502,
 		"n4":  0.030821,
+		"n4d": 0.032578,
 		"t2d": 0.027502,
 		"z3":  0.0496531,
 	}
@@ -101,6 +102,7 @@ var (
 		"n2":  0.004237,
 		"n2d": 0.003686,
 		"n4":  0.004131,
+		"n4d": 0.3,
 		"t2d": 0.003686,
 		"z3":  0.0066553,
 	}
@@ -119,6 +121,7 @@ var (
 		"n2":  0.007650 / 0.031611,
 		"n2d": 0.002773 / 0.027502,
 		"n4":  0.007976 / 0.030821,
+		"n4d": 0.0130312 / 0.032578,
 		"t2d": 0.006655 / 0.027502,
 		"z3":  0.0165 / 0.0496531,
 	}
@@ -127,18 +130,21 @@ var (
 		"n1":  0.033174,
 		"n2":  0.033174,
 		"n2d": 0.028877,
+		"n4d": 0.0342069,
 	}
 	customMemoryPricePerHourPerGb = map[string]float64{
 		"e2":  0.003067,
 		"n1":  0.004446,
 		"n2":  0.004446,
 		"n2d": 0.003870,
+		"n4d": 0.0038871,
 	}
 	customPreemptibleDiscount = map[string]float64{
 		"e2":  0.006867 / 0.022890,
 		"n1":  0.00698 / 0.033174,
 		"n2":  0.00802 / 0.033174,
 		"n2d": 0.002908 / 0.028877,
+		"n4d": 0.0130312 / 0.0342069,
 	}
 
 	// e2-micro and e2-small have allocatable set too high resulting in

cluster-autoscaler/utils/gpu/gpu.go

@@ -26,6 +26,8 @@ import (
 )
 
 const (
+	// ResourceAMDGPU is the name of the AMD GPU resource.
+	ResourceAMDGPU = "amd.com/gpu"
 	// ResourceNvidiaGPU is the name of the Nvidia GPU resource.
 	ResourceNvidiaGPU = "nvidia.com/gpu"
 	// ResourceDirectX is the name of the DirectX resource on windows.
@@ -35,6 +37,14 @@ const (
 	DefaultGPUType = "nvidia-tesla-k80"
 )
 
+// GPUVendorResourceNames centralized list of all known GPU vendor extended resource names.
+// Extend this slice if new vendor resource names are added.
+var GPUVendorResourceNames = []apiv1.ResourceName{
+	ResourceNvidiaGPU,
+	ResourceAMDGPU,
+	ResourceDirectX,
+}
+
 const (
 	// MetricsGenericGPU - for when there is no information about GPU type
 	MetricsGenericGPU = "generic"
@@ -109,23 +119,53 @@ func validateGpuType(availableGPUTypes map[string]struct{}, gpu string) string {
 // if the drivers are installed and GPU is ready to use.
 func NodeHasGpu(GPULabel string, node *apiv1.Node) bool {
 	_, hasGpuLabel := node.Labels[GPULabel]
-	gpuAllocatable, hasGpuAllocatable := node.Status.Allocatable[ResourceNvidiaGPU]
-	return hasGpuLabel || (hasGpuAllocatable && !gpuAllocatable.IsZero())
+	if hasGpuLabel {
+		return true
+	}
+	// Check for extended resources as well
+	for _, gpuVendorResourceName := range GPUVendorResourceNames {
+		gpuAllocatable, hasGpuAllocatable := node.Status.Allocatable[gpuVendorResourceName]
+		if hasGpuAllocatable && !gpuAllocatable.IsZero() {
+			return true
+		}
+	}
+	return false
 }
 
 // PodRequestsGpu returns true if a given pod has GPU request.
 func PodRequestsGpu(pod *apiv1.Pod) bool {
 	podRequests := podutils.PodRequests(pod)
-	_, gpuFound := podRequests[ResourceNvidiaGPU]
-	return gpuFound
+	for _, gpuVendorResourceName := range GPUVendorResourceNames {
+		if _, found := podRequests[gpuVendorResourceName]; found {
+			return true
+		}
+	}
+	return false
+}
+
+// DetectNodeGPUResourceName inspects the node's allocatable resources and returns the first
+// known GPU extended resource name that has non-zero allocatable. Falls back to Nvidia for
+// backward compatibility if none are found but a GPU label is present.
+func DetectNodeGPUResourceName(node *apiv1.Node) apiv1.ResourceName {
+	for _, rn := range GPUVendorResourceNames {
+		if qty, ok := node.Status.Allocatable[rn]; ok && !qty.IsZero() {
+			return rn
+		}
+	}
+	// Fallback: preserve previous behavior (defaulting to Nvidia) if label existed
+	return ResourceNvidiaGPU
 }
 
 // GetNodeGPUFromCloudProvider returns the GPU the node has. Returned GPU has the GPU label of the
 // passed in cloud provider. If the node doesn't have a GPU, returns nil.
 func GetNodeGPUFromCloudProvider(provider cloudprovider.CloudProvider, node *apiv1.Node) *cloudprovider.GpuConfig {
 	gpuLabel := provider.GPULabel()
 	if NodeHasGpu(gpuLabel, node) {
-		return &cloudprovider.GpuConfig{Label: gpuLabel, Type: node.Labels[gpuLabel], ExtendedResourceName: ResourceNvidiaGPU}
+		return &cloudprovider.GpuConfig{
+			Label:                gpuLabel,
+			Type:                 node.Labels[gpuLabel],
+			ExtendedResourceName: DetectNodeGPUResourceName(node),
+		}
 	}
 	return nil
 }

cluster-autoscaler/utils/gpu/gpu_test.go

@@ -236,3 +236,70 @@ func TestGetGpuInfoForMetrics(t *testing.T) {
 		})
 	}
 }
+
+func TestDetectNodeGPUResourceName(t *testing.T) {
+	testCases := []struct {
+		name                 string
+		node                 *apiv1.Node
+		expectedResourceName apiv1.ResourceName
+	}{
+		{
+			name: "nvidia gpu",
+			node: &apiv1.Node{
+				ObjectMeta: metav1.ObjectMeta{
+					Name:   "node-with-nvidia-gpu",
+					Labels: map[string]string{},
+				},
+				Status: apiv1.NodeStatus{
+					Capacity: apiv1.ResourceList{
+						gpu.ResourceNvidiaGPU: *resource.NewQuantity(1, resource.DecimalSI),
+					},
+					Allocatable: apiv1.ResourceList{
+						gpu.ResourceNvidiaGPU: *resource.NewQuantity(1, resource.DecimalSI),
+					},
+				},
+			},
+			expectedResourceName: gpu.ResourceNvidiaGPU,
+		},
+		{
+			name: "amd gpu",
+			node: &apiv1.Node{
+				ObjectMeta: metav1.ObjectMeta{
+					Name:   "node-with-amd-gpu",
+					Labels: map[string]string{},
+				},
+				Status: apiv1.NodeStatus{
+					Capacity: apiv1.ResourceList{
+						gpu.ResourceAMDGPU: *resource.NewQuantity(8, resource.DecimalSI),
+					},
+					Allocatable: apiv1.ResourceList{
+						gpu.ResourceAMDGPU: *resource.NewQuantity(8, resource.DecimalSI),
+					},
+				},
+			},
+			expectedResourceName: gpu.ResourceAMDGPU,
+		},
+		{
+			name: "test default gpu resource name",
+			node: &apiv1.Node{
+				ObjectMeta: metav1.ObjectMeta{
+					Name:   "node-without-gpu",
+					Labels: map[string]string{},
+				},
+				Status: apiv1.NodeStatus{
+					Capacity:    apiv1.ResourceList{},
+					Allocatable: apiv1.ResourceList{},
+				},
+			},
+			expectedResourceName: gpu.ResourceNvidiaGPU,
+		},
+	}
+	for _, tc := range testCases {
+		t.Run(tc.name, func(t *testing.T) {
+			resourceName := gpu.DetectNodeGPUResourceName(tc.node)
+			if resourceName != tc.expectedResourceName {
+				t.Errorf("expected resource name %s but got %s", tc.expectedResourceName, resourceName)
+			}
+		})
+	}
+}