Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Relax validation rules enforcing immutability of replicated jobs #579

Closed
danielvegamyhre opened this issue May 24, 2024 · 5 comments · Fixed by #580
Closed

Relax validation rules enforcing immutability of replicated jobs #579

danielvegamyhre opened this issue May 24, 2024 · 5 comments · Fixed by #580
Assignees
Labels
kind/feature Categorizes issue or PR as related to a new feature.

Comments

@danielvegamyhre
Copy link
Contributor

danielvegamyhre commented May 24, 2024

What would you like to be added:
Relax validation rules enforcing immutability of replicated jobs. Specifically:

  • .spec.replicatedJobs[*].spec.template.spec.template (job podTemplateSpec, to add annotations, taint tolerations, etc)

Why is this needed:
DWS compatibility with JobSet. cc @alculquicondor for more context

@alculquicondor
Copy link

When using ResourceFlavors with nodeLabels, Kueue will inject node selectors to the Pod.

When using cluster-autoscaler ProvisioningRequest, Kueue will inject an annotation to the Pod template.

These fields should be mutable, especially if the Jobset is suspended or is just about to transition to suspend=false.

@danielvegamyhre
Copy link
Contributor Author

danielvegamyhre commented May 24, 2024

Planning on including this in a patch release early next week. cc @ahg-g

I think a patch release is fine, since the validation is being relaxed (not increased) so we won't break any running workloads running on v0.5.x and upgrading to this patch (v0.5.2).

Alternatively we can include this in v0.6.0, which will be ready soon as well, after we do some manual testing of #537 (targeting June 1st).

@danielvegamyhre danielvegamyhre added the kind/feature Categorizes issue or PR as related to a new feature. label May 24, 2024
@danielvegamyhre danielvegamyhre self-assigned this May 24, 2024
@kannon92
Copy link
Contributor

I would vote to put it in a new release. It’s not really something I would think to be included in a patch release.

@danielvegamyhre
Copy link
Contributor Author

Reduced scope of bug specifically to replicatedJob pod template since #529 already exists

@danielvegamyhre
Copy link
Contributor Author

I would vote to put it in a new release. It’s not really something I would think to be included in a patch release.

I reduced the scope of the changes to be the validation fix only, and postponed the elastic jobset support to be included in v0.6.0. So I think we can go ahead and do a patch release for this.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
kind/feature Categorizes issue or PR as related to a new feature.
Projects
None yet
Development

Successfully merging a pull request may close this issue.

3 participants