Donate repo to kubernetes-sigs

[ArangoGutierrez]

TLDR from : #341

---

There is some renewed interest in moving the mpi-operator to (possibly) kubernetes-sigs, assuming that SIG Apps is willing to sponsor the repo.

The motivation is to encourage non-training users (like HPC) to use and contribute to it, without having to install or learn about kubeflow's training-operator.

• Most Kubeflow components are "loosely coupled", so the definition of "being part of kubeflow" is mostly about if a Kubeflow Working Group is responsible for it

• The committers of the mpi-operator can decide where the best home for it is, but being under the Kubeflow brand certainly will bring more attention to the tool

---

from https://github.com/kubernetes/community/blob/master/github-management/kubernetes-repositories.md#rules-for-donated-repositories

• All contributors must have signed the CNCF Individual CLA or CNCF Corporate CLA
• If (a) contributor(s) have not signed the CLA and could not be reached, a NOTICE file should be added referencing section 7 of the CLA with a list of the developers who could not be reached
• Licenses of dependencies are acceptable; project owners can ping @caniszczyk for review of third party deps
• Boilerplate text across all files should attribute copyright as follows: "Copyright " if no CLA was in place prior to donation
• Additions of the standard Kubernetes header to code created by the contributors can occur post-transfer, but should ideally occur shortly thereafter.
• Should contain template files as per the kubernetes-template-project.
• Kubeflow plan to migrate MPi-Jobs dependencies to the news k8s-sigs/mpi-operator location

After Donation / repo creation under k8s-sigs/mpi-operator
we need to :

• Stabilize the current code under k8s-gis parameters
• Cut a first release under k8s-sigs
• Migrate Kubeflow projects to new location

Repo will be sponsored by sig-Apps

[alculquicondor]

Can you add the rationale for the donation in the description?

Also it would be good to add what's the high-level plan for kubeflow to use the kubernetes-sig/mpi-operator in their manifests.

[ArangoGutierrez]

/assign @alculquicondor @ArangoGutierrez

[jbottum]

Perhaps this data point is relevant from the Kubeflow User Survey, which shows MPI's position to other technologies used by Kubeflow users.

[ArangoGutierrez]

Step one:

• To the question: Are you ok with donating the repo mpi-operator to kubernetes-sigs, please reply lgtm

From the OWNERS file:
@ rongou @terrytangyuan @alculquicondor @carmark @gaocegege @zw0610

[theadactyl]

Hi there, @kubeflow/project-steering-group will need to see a proposal with the following information. This should also be open for broader community feedback before any decision is made.

• Goals
• Motivation / Rationale
• User Benefit
• Design Proposal, including:
  -- Alternatives Considered
  -- Dependencies
  -- User Impact
• Detailed Design

[ArangoGutierrez]

Hi there, @kubeflow/project-steering-group will need to see a proposal with the following information. This should also be open for broader community feedback before any decision is made.

• Goals
• Motivation / Rationale
• User Benefit
• Design Proposal, including:
  -- Alternatives Considered
  -- Dependencies
  -- User Impact
• Detailed Design

Will work on it, thanks for the check list

[theadactyl]

The best place to submit is as a PR to the following directory: https://github.com/kubeflow/community/tree/master/proposals

Thank you! Our proposal process is not widely known, but it's important to follow for any proposed changes that include code governance.

[rongou]

lgtm

…

On Wed, Mar 9, 2022 at 1:22 PM Thea Lamkin ***@***.***> wrote: The best place to submit is as a PR to the following directory: https://github.com/kubeflow/community/tree/master/proposals Thank you! Our proposal process is not widely known, but it's important to follow for any proposed changes that include code governance. — Reply to this email directly, view it on GitHub <#459 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AADZLTLZSDONFJ6MGACU4PLU7EJA7ANCNFSM5QKOC5VQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[terrytangyuan]

LGTM. Let's get a formal proposal for final review from @kubeflow/project-steering-group.

[gaocegege]

/cc @richardsliu

LGTM

[carmark]

LGTM

[Jeffwan]

em.. I am not sure if WG-training give enough feedbacks on this. mpi-operator is one of the training operator suite. What about others or the universal operator? We need to have a concrete plan for the donation

[thesuperzapper]

I also want to highlight that by definition, if a project is donated from Kubeflow to another organization, Kubeflow will no longer have any control over it.

This means there would be a question of what to do with the Kubeflow Working Group which controls that project.

In the case that only one part of a WG is donated, then that WG would continue to exist.
(Assuming the community thinks there are enough meaningful things that remain under the WG's control)

In the case that most or all of a WG's code is donated, obviously that WG would be disestablished. In such cases, it may also make sense to create a SIG for users of Kubeflow AND that tool (just like we have for other projects that were developed outside of Kubeflow).

[ArangoGutierrez]

kubeflow/community#557

[kannon92]

Is this still on the table?

[terrytangyuan]

There were some additional discussions in kubeflow/community#557

Now that Kubeflow is an incubating project in CNCF, we need to re-evaluate whether there's enough motivation to move this to K8s as a subproject.

[alculquicondor]

I don't think so.

It would be better to just lean on this proposal instead: kubernetes/kubernetes#124537

Note that there is an item in the agenda for it in the next WG training meeting May 1st at 2pm UTC.

And yesterday's recording for k8s WG Batch https://youtu.be/COzNG8XCGTw?si=ncemR5-gmNE3W0OH&t=938

[kannon92]

Should we close and reopen if the discussion happens again?

[alculquicondor]

/close

[google-oss-prow]

@alculquicondor: Closing this issue.

In response to this:

/close

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.