cluster-api: node template in scale-from-0-nodes scenario with DRA

[ttsuuubasa]

What type of PR is this?

/kind feature

What this PR does / why we need it:

This PR makes cluster-api as Cloud Provider enable to return ResourceSlice which users specify of the node template (NodeInfo) in scale-from-0-nodes scenario by TemplateNodeInfo(). The node template informs Cluster Autoscaler about how resources the node to be spawned has like CPU, memory and GPU in the NodeGroup where no existing nodes are present. With the DRA expansion, the custom resources of the node like GPU is expressed by ResourceSlice, however, cluster-api as Cloud Provider had no way for users to have NodeInfo possess the ResourceSlice. That's why I added the following annotation to specify the DRA driver name and allow users to specify ResourceSlice of NodeInfo returned by TemplateNodeInfo(). This method by annotating follows the device plugin's way and the original annotations can also be used for the device plugin.

capacity.cluster-autoscaler.kubernetes.io/dra-driver: <driver-name>

Which issue(s) this PR fixes:

Fixes #7724

Special notes for your reviewer:

• Although I stated the annotation for DRA pool name is also required in the above issue, the pool name typically is set as the node name. Therefore, I didn't institute the annotation for the pool name and it is configured to the node name internally.

• The DRA handles not only GPUs but also other devices.
  The current PR's implementation regards the device as GPU when the DRA driver name and GPU count is specified.
  This may be no problem for the time being because most DRA drivers treat only GPU at this point but not expandable in the future.

• TemplateNodeInfo() returns the slice of ResourceSlice but this modification include one ResourceSlice to the slice per Node.

Does this PR introduce a user-facing change?

Cloud Provider using cluster-api is extended to handle DRA devices such as GPU in the scale-from-0-nodes.

Additional documentation e.g., KEPs (Kubernetes Enhancement Proposals), usage docs, etc.:

[Usage]: https://github.com/kubernetes/autoscaler/blob/master/cluster-autoscaler/cloudprovider/clusterapi/README.md#scale-from-zero-support

[linux-foundation-easycla]

The committers listed above are authorized under a signed CLA.

• ✅ login: ttsuuubasa / name: Tsubasa Watanabe (3fbacf0)

[k8s-ci-robot]

Welcome @ttsuuubasa!

It looks like this is your first PR to kubernetes/autoscaler 🎉. Please refer to our pull request process documentation to help your PR have a smooth ride to approval.

You will be prompted by a bot to use commands during the review process. Do not be afraid to follow the prompts! It is okay to experiment. Here is the bot commands documentation.

You can also check if kubernetes/autoscaler has its own contribution guidelines.

You may want to refer to our testing guide if you run into trouble with your tests not passing.

If you are having difficulty getting your pull request seen, please follow the recommended escalation practices. Also, for tips and tricks in the contribution process you may want to read the Kubernetes contributor cheat sheet. We want to make sure your contribution gets all the attention it needs!

Thank you, and welcome to Kubernetes. 😃

[k8s-ci-robot]

Hi @ttsuuubasa. Thanks for your PR.

I'm waiting for a kubernetes 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: ttsuuubasa
Once this PR has been reviewed and has the lgtm label, please assign enxebre 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:

• cluster-autoscaler/cloudprovider/clusterapi/OWNERS

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

[elmiko]

this is looking good @ttsuuubasa , i have a few requests.

• add unit tests to confirm the behavior of the new template options for dra in TestNodeGroupTemplateNodeInfo, in clusterapi_nodegroup_test.go.
• add unit tests to confirm the behavior of InstanceResourceSlices, in clusterapi_unstructured_test.go.
• update the README to include the new annotation name for DRA.

[ttsuuubasa]

@elmiko
I added the two test cases you mentioned:

• "When the NodeGroup can scale from zero and DRA is enabled, it creates ResourceSlice derived from the annotation of DRA driver name and GPU count" in TestNodeGroupTemplateNodeInfo, in clusterapi_nodegroup_test.go
• The test to compare the expected ResourceSlice with one created by InstanceResourceSlices in TestAnnotations, in clusterapi_unstructured_test.go

I also update the README to explain the new annotation for DRA.

Could you please review it again?

[elmiko]

thanks @ttsuuubasa , i will try to review before monday.

[elmiko]

this is looking nice @ttsuuubasa , thank you!

i would like to get one more review on this, cc @jackfrancis might you be able to give a review?

/lgtm