wip: Add EC2 metadata-labeler metadata source

[sylviah23]

What type of PR is this?

/kind feature

What is this PR about? / Why do we need it?

This PR gets the Volume and ENI count of each node via the EC2 API. It then adds this information to the metadata labels of each node in the cluster, so Kubernetes can get this information without IMDS access.

How was this change tested?

Unit testing and E2E testing and manually testing with EKS cluster by adding a new node.

Does this PR introduce a user-facing change?

Add new metadata source relying on EC2.

[linux-foundation-easycla]

The committers listed above are authorized under a signed CLA.

• ✅ login: AndrewSirenko / name: Drew Sirenko (c7cd971, 5068f1e, cb14e84, 8b6d42b)
• ✅ login: sylviah23 / name: Sylvia Han (7a76c70, 9b5a860, 2cf2e52, 1476a62, 5293392, 5863c19)

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign jsafrane for approval. 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:

• OWNERS

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

[k8s-ci-robot]

Hi @sylviah23. Thanks for your PR.

I'm waiting for a kubernetes-sigs 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.

[AndrewSirenko]

/ok-to-test

[github-actions]

Code Coverage Diff

File	Old Coverage	New Coverage	Delta
github.com/kubernetes-sigs/aws-ebs-csi-driver/pkg/cloud/cloud.go	87.4%	84.9%	-2.5
github.com/kubernetes-sigs/aws-ebs-csi-driver/pkg/cloud/metadata/k8s.go	58.3%	72.3%	14.0
github.com/kubernetes-sigs/aws-ebs-csi-driver/pkg/cloud/metadata/labels.go	Does not exist	57.6%
github.com/kubernetes-sigs/aws-ebs-csi-driver/pkg/cloud/metadata/metadata.go	84.8%	86.5%	1.8

[mdzraf]

/test pull-aws-ebs-csi-driver-image-test

[mdzraf]

/test pull-aws-ebs-csi-driver-image-test

Fixed issue with ECR repo being full.

[ConnorJC3]

First review round: I have not manually tested yet, intend to do that in my next review. I also don't think I was as in depth with looking at some parts (in particular the tests) as I should have been, so I might have missed some bugs lurking there.

The comments about the leader election are probably the most critical comments in this review - as written, this code will keep running after the leader election is switched away, causing two or more instances of the driver to simultaneously run the patcher and race each other.

[torredil]

Thank you for the contribution and adding comprehensive code coverage on the first revision @sylviah23 . Overall, this is looking solid 🙂.

The leader election code needs some work, as Connor suggested #2591 (comment), leveraging the CSI lib-utils wrapper will be a nice improvement. That code is battle tested and already addresses some of the core challenges we need to solve implementing this enhancement.

[AndrewSirenko]

Ran pre_allocated and karpenter scale tests with 2500 PVCs (therefore 25-40 nodes) and setting metadataSources=[ec2labelskubernetes]. Everything worked and proper allocatable counts were there!

Before moving to GA we should try suddenly installing driver on a XXL node cluster to see if anything chokes.

[ConnorJC3]

I still think we need to have a naming discussion, especially re the metadata source name and the patcher sidecar name

[ConnorJC3]

This code is very duplicated, we could probably combine it?

[AndrewSirenko]

I still think we need to have a naming discussion, especially re the metadata source name and the patcher sidecar name

Final names: metadata-labeler for both metadata source and new container.

[sylviah23]

@AndrewSirenko LabelRefreshTime in cmd/main.go should be changed to 60 * time.Minute