search

kubernetes-retired / funding

Funding requests for project infrastructure, events, and consulting.
Apache License 2.0
16 stars 13 forks source link

EKS clusters for scale testing of windows-service-proxy and operational readiness #40

Closed jayunit100 closed 9 months ago

jayunit100 commented 1 year ago

Problem Statement

We are in the process of hardening the overall windows on kubernetes experience, and begin paving the road to reduce in-tree tech debt, around things like possibly simplify kube-proxy, and also make the windows support matrix in K8s explicit.

To do this, we'll need to have more infrastructure for customized experiments in the windows server area, and do CI jobs which swap out the in-tree proxy for new approaches which we think will reduce tech debt in-tree, over time, while increasing the K8s experience overall for enterprises.

We think AWS is a solid foundation as one of the more popular infrastructures that people will.

Proposed Solution

Let's get started with EKS for operational readiness verification on windows. This will also help us work through issues such as https://github.com/kubernetes/kubernetes/issues/119022, which we think has specific consequences for AWS and other public clouds, as we discovered recently using our own, internal EKS accounts.

Cost

We can accomplish our current needs in 5K per month, assuming 2 ~ 3 full time, small clusters of EKS nodes (< 3 each). We expect this to be "high" bc windows machines/licence node costs are a little higher then normal nodes...

Open Questions

Next Steps

Other Considerations, Notes, or References

jayunit100 commented 1 year ago

cc @kubernetes/steering-committee @dims

jayunit100 commented 1 year ago

cc @aroradaman

jayunit100 commented 1 year ago

cc @marosset @jsturtevant @aravindhp @claudiubelu

ameukam commented 1 year ago

/transfer funding

BenTheElder commented 1 year ago

If you're just running CI jobs on AWS, we already have resources for this? This is a question for k8s-infra leads.

ameukam commented 1 year ago

If you're just running CI jobs on AWS, we already have resources for this? This is a question for k8s-infra leads.

I think the idea here is to provide an AWS account used by windows-service-proxy maintainers to setup a dev environment and later automate things with the k8s CI.

BenTheElder commented 1 year ago

I think the idea here is to provide an AWS account used by windows-service-proxy maintainers to setup a dev environment and later automate things with the k8s CI.

Got it.

I think the TLDR of the steering meeting is we should be handling this with k8s infra resources first and we should escalate to a funding issue when k8s infra is unsure of or lacks the resources.

E.G. For azure in the future we don't have resources in k8s infra / owned by the project to provide currently. For AWS we do and k8s infra can handle this request, but probably needs to follow up with a policy for dev accounts long term.

k8s-triage-robot commented 9 months ago

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

You can:

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

BenTheElder commented 9 months ago

/remove-lifecycle stale /close This was/will be handled through k8s infra.

k8s-ci-robot commented 9 months ago

@BenTheElder: Closing this issue.

In response to [this](https://github.com/kubernetes/funding/issues/40#issuecomment-1932426164): >/remove-lifecycle stale >/close >This was/will be handled through k8s infra. Instructions for interacting with me using PR comments are available [here](https://git.k8s.io/community/contributors/guide/pull-requests.md). If you have questions or suggestions related to my behavior, please file an issue against the [kubernetes/test-infra](https://github.com/kubernetes/test-infra/issues/new?title=Prow%20issue:) repository.