Added support for AWS inf2 instance types

[mmcclean-aws]

Added support for the AWS Inferentia 2 instance types based on the NeuronCore v2 chip architecture

What type of PR is this?

/kind feature

What this PR does / why we need it:

Which issue(s) this PR fixes:

Fixes #5528

Special notes for your reviewer:

Does this PR introduce a user-facing change?

NONE

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

[linux-foundation-easycla]

The committers listed above are authorized under a signed CLA.

• ✅ login: mmcclean-aws / name: Matt McClean (63161aa, ee86ce4)

[k8s-ci-robot]

Welcome @mmcclean-aws!

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. 😃

[ferboz]

/test all

[k8s-ci-robot]

@ferboz: Cannot trigger testing until a trusted user reviews the PR and leaves an /ok-to-test message.

In response to this:

/test all

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/test-infra repository.

[jbartosik]

/hold

I don't understand why this PR is changing VPA files. This seems weird because there is a change in a vendored file but there are no corresponding changes in go.mod / go.sum

[gjtempleton]

Agree with @jbartosik .

If you're needing to update the vendored AWS SDK for the CA, you can see instructions on doing that here: https://github.com/kubernetes/autoscaler/tree/master/cluster-autoscaler/cloudprovider/aws#using-the-aws-sdk-vendored-in-the-aws-cloudprovider

[mmcclean-aws]

thanks @gjtempleton and @jbartosik. Does that mean that in order to add a new instance type this repo doesn't need to be updated ? All you need to do is either wait till the AWS SDK is updated or follow the instructions outlined in the link shared ?

[x13n]

IIUC you need to update the sdk based on the instructions from the link. This change should really be limited to changes within cluster-autoscaler/cloudprovider/aws dir. Assigning AWS owners for this change.

/assign @gjtempleton @jaypipes @drmorr0

FWIW, we should really do #4878 (and perhaps a similar change for VPA) to limit confusion in the future.

[gjtempleton]

There's a couple of bits necessary for this, the update to the hardcoded list which you've already included and following the instructions linked to previously to update the AWS SDK version we vendor in to the AWS cloudprovider implementation for the CA.

The changes to the VPA related files should be rolled back as @jbartosik and @x13n say.

[mmcclean-aws]

I have now rolled back the VPA related files as requested

[ellistarn]

Thanks for the help on this @gjtempleton! In the short term, we may need to continue to support ad-hoc manual support for preview EC2 instance types. I'm happy to review and support PRs to the AWS provider for this, if you're comfortable continuing this practice.

[gjtempleton]

Thanks.
Removing the hold as now only touching expected CA files.
/unhold
/lgtm
/approve

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: gjtempleton, mmcclean-aws

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/cloudprovider/aws/OWNERS [gjtempleton]

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

[gjtempleton]

Thanks @ellistarn
More generally, will keep an eye out for future PRs generating this instance list potentially removing these ad-hoc ones, if it becomes a common occurrence then we may want to look at some method for automating pulling in these instance types which aren't yet publicly listed by the SDK.