Migrate cluster-autoscaler AWS provider from AWS SDK for Go v1 to v2

[Copilot]

Overview

This PR migrates the cluster-autoscaler AWS cloud provider from the end-of-life AWS SDK for Go v1 to the actively supported v2. AWS SDK for Go v1 reached end-of-support on July 31, 2025, making this migration critical for continued AWS support and security updates.

Motivation

• End of Support: AWS SDK Go v1 is no longer maintained or receiving security updates
• Missing Features: v1 lacks support for environment variable-based service endpoint overrides, which are useful for testing and custom deployments
• Future Compatibility: Ensures cluster-autoscaler remains compatible with AWS's latest features and best practices

Changes

Core Infrastructure

AWS Configuration (aws_sdk_provider.go)

• Replaced session-based initialization with modern config-based approach
• Migrated from session.NewSession() to config.LoadDefaultConfig()
• Updated endpoint resolution for custom service endpoints
• Added proper context support for all AWS operations

Service Clients (aws_manager.go)

• Updated client initialization: autoscaling.New(sess) → autoscaling.NewFromConfig(cfg)
• Applied same pattern for EC2 and EKS service clients
• All clients now properly support context cancellation and timeouts

API Call Updates

Pagination
Replaced v1's callback-based pagination with v2's token-based approach:

// v1 (old)
err := client.DescribeAutoScalingGroupsPages(input, func(page *Output, lastPage bool) bool {
    process(page)
    return !lastPage
})

// v2 (new)
var nextToken *string
for {
    if nextToken != nil {
        input.NextToken = nextToken
    }
    page, err := client.DescribeAutoScalingGroups(ctx, input)
    if err != nil { return err }
    process(page)
    if page.NextToken == nil { break }
    nextToken = page.NextToken
}

Context Support
All AWS API calls now require and properly use context.Context:

ctx := context.Background()
result, err := client.DescribeAutoScalingGroups(ctx, input)

Type System Changes

SDK v2 introduced significant type changes for better type safety and reduced pointer usage:

Value Types vs Pointers

• *autoscaling.Group → autoscalingtypes.AutoScalingGroup (value)
• *autoscaling.Instance → autoscalingtypes.Instance (value)
• []*TagDescription → []TagDescription (slice of values)

Typed Enums
Replaced magic strings with typed enums:

• *string lifecycle states → autoscalingtypes.LifecycleState enum
• *string taint effects → ekstypes.TaintEffect enum
• Status codes, accelerator types, etc. are now strongly typed

Integer Types
Updated for AWS API changes:

• AutoScaling capacity fields: *int64 → *int32
• Added explicit conversions: int64(*value) where needed

String Handling
Removed dependency on helper functions:

• aws.StringValue(ptr) → direct dereference with nil checks
• aws.StringSlice(slice) → direct slice usage (now []string not []*string)

Instance Type Handling (aws_util.go)

Updated EC2 instance type generation:

• Migrated DescribeInstanceTypesPages() to token-based pagination
• Updated type conversions for MemoryMiB, VCpuInfo fields (now int32 not int64)
• Fixed architecture and GPU count handling for new type structures

Managed Node Groups (aws_wrapper.go)

• Updated EKS DescribeNodegroup to use context and v2 types
• Fixed taint translation from EKS to Kubernetes format with typed enums
• Updated label and tag extraction for value-based tag slices

Testing

This is a work-in-progress PR. The code changes are complete and ready for review, but additional work is needed:

• Fix remaining compilation issues in complex EC2 InstanceRequirements type conversions
• Update test files to use SDK v2 mock interfaces
• Run full test suite
• Verify integration with actual AWS APIs
• Remove vendored aws-sdk-go v1 directory

Dependencies

Added AWS SDK for Go v2 packages:

• github.com/aws/aws-sdk-go-v2/config
• github.com/aws/aws-sdk-go-v2/service/autoscaling
• github.com/aws/aws-sdk-go-v2/service/ec2
• github.com/aws/aws-sdk-go-v2/service/eks
• github.com/aws/aws-sdk-go-v2/feature/ec2/imds

The v1 SDK (github.com/aws/aws-sdk-go) will be removed in a follow-up commit once all tests pass.

Breaking Changes

None expected for end users. This is an internal implementation change that maintains the same external API and behavior.

Related Issues

Fixes #[issue-number] - AWS SDK for Go v2 Support

Migration Notes

For developers extending the AWS provider:

• Import service types from github.com/aws/aws-sdk-go-v2/service/{service}/types
• Use context.Context for all AWS API calls
• Handle pagination manually with NextToken instead of callback functions
• Most fields are now values instead of pointers - check for empty values instead of nil
• Use typed enums instead of string constants where applicable

---

Status: 🚧 Work in Progress - Core migration complete, final type conversions and tests in progress

Original prompt

---

This section details on the original issue you should resolve

<issue_title>AWS SDK for Go v2 Support</issue_title>
<issue_description>

Which component are you using?: cluster-autoscaler for EKS

/area cluster-autoscaler

Is your feature request designed to solve a problem? If so describe the problem this feature should solve.:
cluster-autoscaler uses AWS SDK for Go v1 which has been end-of-support since July 31, 2025. Furthermore, the older sdk version doesn't support useful features like overriding service endpoints via environment variables.

Describe the solution you'd like.:
Update the code to utilize AWS SDK for Go v2.

Describe any alternative solutions you've considered.:

Additional context.:

</issue_description>

Comments on the Issue (you are @copilot in this section)

Fixes #8671

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[k8s-ci-robot]

Adding the "do-not-merge/release-note-label-needed" label because no release-note block was detected, please follow our release note process to remove it.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

[k8s-ci-robot]

Hi @Copilot. Thanks for your PR.

I'm waiting for a github.com member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: Copilot
Once this PR has been reviewed and has the lgtm label, please ask for approval from jackfrancis. For more information see the Code Review Process.

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

Needs approval from an approver in each of these files:

• cluster-autoscaler/OWNERS

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

[jackfrancis]

shouldn't there be a corresponding explicit update to github.com/aws/aws-sdk-go-v2 instead of this set o indirect additions?

[jackfrancis]

in cluster-autoscaler/go.mod shouldn't there be a corresponding explicit update to github.com/aws/aws-sdk-go-v2 instead of just indirect additions?