CA: Update Azure SDK to v2 #8784

mboersma · 2025-11-10T16:07:15Z

What type of PR is this?

/kind cleanup
/area provider/azure

What this PR does / why we need it:

Updates the Azure cluster-autoscaler backend to use Azure SDK v2.

Which issue(s) this PR fixes:

Fixes #8145

Special notes for your reviewer:

Does this PR introduce a user-facing change?

Update Azure SDK to v2

mboersma · 2025-11-10T16:10:36Z

/cc @jackfrancis

mboersma · 2025-11-10T16:37:46Z

/cc @nojnhuh

jackfrancis

/lgtm

/hold for review from @tallaxes

k8s-ci-robot · 2025-11-10T20:57:50Z

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: jackfrancis, mboersma

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

~~cluster-autoscaler/OWNERS~~ [jackfrancis]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

jackfrancis · 2025-11-10T20:58:12Z

/release-note-edit

Update Azure SDK to v2

mboersma · 2025-11-10T21:14:07Z

cluster-autoscaler/cloudprovider/azure/azure_client.go


 	klog "k8s.io/klog/v2"

-	"sigs.k8s.io/cloud-provider-azure/pkg/azureclients/deploymentclient"


The Azure code was using these client interfaces exported by cloud-provider-azure, but they didn't support Azure SDK v2 and have been removed in current releases.

They have been replaced in cloud-provider-azure with Azure SDK v2 clients (and mocks), in azclient package (https://github.com/kubernetes-sigs/cloud-provider-azure/tree/master/pkg/azclient) That's what cluster-autoscaler should be migrating to, I think, unless there is a really good reason to use alternative clients.

The two exceptions without existing v2 clients in cloud-provider-azure would be agentpool (autoscaler already has one of its own - though generating one in cloud-provider-azure and using it instead would be better) and resourcesku.

mboersma · 2025-11-10T21:19:25Z

cluster-autoscaler/go.mod

+	github.com/Azure/azure-sdk-for-go/sdk/azcore v1.20.0
+	github.com/Azure/azure-sdk-for-go/sdk/azidentity v1.13.0
+	github.com/Azure/azure-sdk-for-go/sdk/resourcemanager/compute/armcompute/v7 v7.1.0
 	github.com/Azure/azure-sdk-for-go/sdk/resourcemanager/containerservice/armcontainerservice/v5 v5.1.0-beta.2


The armcontainerservice API is up to v8 now, but v6 has some breaking changes so I left this alone. I'll create an issue to update it so we don't fall too far behind the currently supported API.

Edit: see #8790

mboersma · 2025-11-10T21:20:24Z

cluster-autoscaler/go.mod

+	github.com/Azure/azure-sdk-for-go/sdk/resourcemanager/compute/armcompute/v7 v7.1.0
 	github.com/Azure/azure-sdk-for-go/sdk/resourcemanager/containerservice/armcontainerservice/v5 v5.1.0-beta.2
+	github.com/Azure/azure-sdk-for-go/sdk/resourcemanager/network/armnetwork/v7 v7.1.0
+	github.com/Azure/azure-sdk-for-go/sdk/resourcemanager/resources/armresources/v2 v2.1.0


The current API version for armresources is v3, but it has some breaking changes and seems to lack DeploymentExtended, so I went with the previous version.

jackfrancis · 2025-11-20T18:47:58Z

ping @tallaxes, if we want to get this included in 1.35

tallaxes

Can we use v2 clients from cloud-provider-azure? Among other things, this would ensure high compatibility with client configuration such as auth methods, cloud endpoints, retry and backoff policies, rate limiting, polling frequency, etc.

tallaxes · 2025-11-23T05:45:41Z

cluster-autoscaler/cloudprovider/azure/azure_client.go


 	klog "k8s.io/klog/v2"

-	"sigs.k8s.io/cloud-provider-azure/pkg/azureclients/deploymentclient"


They have been replaced in cloud-provider-azure with Azure SDK v2 clients (and mocks), in azclient package (https://github.com/kubernetes-sigs/cloud-provider-azure/tree/master/pkg/azclient) That's what cluster-autoscaler should be migrating to, I think, unless there is a really good reason to use alternative clients.

The two exceptions without existing v2 clients in cloud-provider-azure would be agentpool (autoscaler already has one of its own - though generating one in cloud-provider-azure and using it instead would be better) and resourcesku.

tallaxes · 2025-11-23T06:51:51Z

cluster-autoscaler/cloudprovider/azure/azure_client.go

+	// Get v2 credentials for all Azure SDK v2 clients
+	cred, err := getAgentpoolClientCredentials(cfg)
 	if err != nil {


This replaces (the now unused) newAuthorizer; do we have confidence it supports the same options and configuration as newAuthorizer?

(Also, should not be named Agentpool if it is now used for everything?)

tallaxes · 2025-11-23T07:01:09Z

cluster-autoscaler/cloudprovider/azure/azure_client.go

-	azClientConfig := cfg.getAzureClientConfig(authorizer, env)
-	azClientConfig.UserAgent = getUserAgentExtension()
+	// Create common client options for all v2 clients
+	clientOptions := &policy.ClientOptions{


These won't respect client retry/backoff/rate limiting config?

tallaxes · 2025-11-23T07:24:41Z

cluster-autoscaler/cloudprovider/azure/azure_scale_set.go

+		resp, err := scaleSet.manager.azClient.virtualMachineScaleSetsClient.Get(ctx, scaleSet.manager.config.ResourceGroup, scaleSet.Name, nil)
+		if err != nil {
+			klog.Errorf("failed to get information for VMSS: %s, error: %v", scaleSet.Name, err)
+			return -1, newGetVMSSFailedError(err, false)


Shouldn't this still be checking for not found? (There are helpers in Azure/azure-sdk-for-go-extentions ...)

k8s-ci-robot requested review from comtalyst and vadasambar November 10, 2025 16:07

k8s-ci-robot added area/cluster-autoscaler size/XXL Denotes a PR that changes 1000+ lines, ignoring generated files. labels Nov 10, 2025

k8s-ci-robot requested a review from jackfrancis November 10, 2025 16:10

k8s-ci-robot requested a review from nojnhuh November 10, 2025 16:37

Update Azure SDK to v2

7047504

mboersma force-pushed the azure-sdk-v2 branch from e3ba9bb to 7047504 Compare November 10, 2025 17:23

jackfrancis approved these changes Nov 10, 2025

View reviewed changes

k8s-ci-robot added the do-not-merge/hold Indicates that a PR should not merge because someone has issued a /hold command. label Nov 10, 2025

k8s-ci-robot assigned jackfrancis Nov 10, 2025

k8s-ci-robot added the lgtm "Looks good to me", indicates that a PR is ready to be merged. label Nov 10, 2025

k8s-ci-robot added the approved Indicates a PR has been approved by an approver from all required OWNERS files. label Nov 10, 2025

k8s-ci-robot added release-note Denotes a PR that will be considered when it comes time to generate release notes. and removed release-note-none Denotes a PR that doesn't merit a release note. labels Nov 10, 2025

mboersma commented Nov 10, 2025

View reviewed changes

tallaxes reviewed Nov 23, 2025

View reviewed changes


		klog "k8s.io/klog/v2"

		"sigs.k8s.io/cloud-provider-azure/pkg/azureclients/deploymentclient"

CA: Update Azure SDK to v2 #8784

Are you sure you want to change the base?

CA: Update Azure SDK to v2 #8784

Conversation

mboersma commented Nov 10, 2025 • edited by k8s-ci-robot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What type of PR is this?

What this PR does / why we need it:

Which issue(s) this PR fixes:

Special notes for your reviewer:

Does this PR introduce a user-facing change?

Uh oh!

mboersma commented Nov 10, 2025

Uh oh!

mboersma commented Nov 10, 2025

Uh oh!

jackfrancis left a comment

Choose a reason for hiding this comment

Uh oh!

k8s-ci-robot commented Nov 10, 2025

Uh oh!

jackfrancis commented Nov 10, 2025

Uh oh!

mboersma Nov 10, 2025

Choose a reason for hiding this comment

Uh oh!

tallaxes Nov 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mboersma Nov 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mboersma Nov 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jackfrancis commented Nov 20, 2025

Uh oh!

tallaxes left a comment

Choose a reason for hiding this comment

Uh oh!

tallaxes Nov 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

tallaxes Nov 23, 2025

Choose a reason for hiding this comment

Uh oh!

tallaxes Nov 23, 2025

Choose a reason for hiding this comment

Uh oh!

tallaxes Nov 23, 2025

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

mboersma commented Nov 10, 2025 •

edited by k8s-ci-robot

Loading

tallaxes Nov 23, 2025 •

edited

Loading

mboersma Nov 10, 2025 •

edited

Loading

mboersma Nov 10, 2025 •

edited

Loading

tallaxes Nov 23, 2025 •

edited

Loading