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

Repeated errors in kueue-manager log when Kueue 0.9 is installed on Kubernetes 1.29 #3496

Open
dgrove-oss opened this issue Nov 8, 2024 · 12 comments
Labels
kind/bug Categorizes issue or PR as related to a bug.

Comments

@dgrove-oss
Copy link
Contributor

What happened:

I created a fresh kind 0.24 cluster with Kubernetes 1.27.17. I installed the Kueue 0.9 release on the cluster.

Every 30 seconds (+/-) the kueue-controller-manager log contains the following stanza:

W1108 22:04:17.191809       1 reflector.go:561] k8s.io/client-go/informers/factory.go:160: failed to list *v1.FlowSchema: the server could not find the requested resource
E1108 22:04:17.191842       1 reflector.go:158] "Unhandled Error" err="k8s.io/client-go/informers/factory.go:160: Failed to watch *v1.FlowSchema: failed to list *v1.FlowSchema: the server could not find the requested resource" logger="UnhandledError"
W1108 22:04:23.165653       1 reflector.go:561] k8s.io/client-go/informers/factory.go:160: failed to list *v1.PriorityLevelConfiguration: the server could not find the requested resource
E1108 22:04:23.165688       1 reflector.go:158] "Unhandled Error" err="k8s.io/client-go/informers/factory.go:160: Failed to watch *v1.PriorityLevelConfiguration: failed to list *v1.PriorityLevelConfiguration: the server could not find the requested resource" logger="UnhandledError"
W1108 22:04:23.370784       1 reflector.go:561] k8s.io/client-go/informers/factory.go:160: failed to list *v1.ValidatingAdmissionPolicyBinding: the server could not find the requested resource
E1108 22:04:23.370836       1 reflector.go:158] "Unhandled Error" err="k8s.io/client-go/informers/factory.go:160: Failed to watch *v1.ValidatingAdmissionPolicyBinding: failed to list *v1.ValidatingAdmissionPolicyBinding: the server could not find the requested resource" logger="UnhandledError"
W1108 22:04:52.216220       1 reflector.go:561] k8s.io/client-go/informers/factory.go:160: failed to list *v1.ValidatingAdmissionPolicy: the server could not find the requested resource
E1108 22:04:52.216261       1 reflector.go:158] "Unhandled Error" err="k8s.io/client-go/informers/factory.go:160: Failed to watch *v1.ValidatingAdmissionPolicy: failed to list *v1.ValidatingAdmissionPolicy: the server could not find the requested resource" logger="UnhandledError"
W1108 22:04:58.002480       1 reflector.go:561] k8s.io/client-go/informers/factory.go:160: failed to list *v1.PriorityLevelConfiguration: the server could not find the requested resource
E1108 22:04:58.002507       1 reflector.go:158] "Unhandled Error" err="k8s.io/client-go/informers/factory.go:160: Failed to watch *v1.PriorityLevelConfiguration: failed to list *v1.PriorityLevelConfiguration: the server could not find the requested resource" logger="UnhandledError"
W1108 22:05:02.844004       1 reflector.go:561] k8s.io/client-go/informers/factory.go:160: failed to list *v1.ValidatingAdmissionPolicyBinding: the server could not find the requested resource
E1108 22:05:02.844044       1 reflector.go:158] "Unhandled Error" err="k8s.io/client-go/informers/factory.go:160: Failed to watch *v1.ValidatingAdmissionPolicyBinding: failed to list *v1.ValidatingAdmissionPolicyBinding: the server could not find the requested resource" logger="UnhandledError"

What you expected to happen:

I do not expect to see error logs when running Kueue 0.9.0 on Kubernetes 1.27. This is a regression vs. Kueue 0.8.3.

How to reproduce it (as minimally and precisely as possible):

Create 1.27 cluster with kind.

kubectl apply --server-side -f https://github.com/kubernetes-sigs/kueue/releases/download/v0.9.0/manifests.yaml

Anything else we need to know?:

Environment:

  • Kubernetes version (use kubectl version): 1.27.17
  • Kueue version (use git describe --tags --dirty --always): 0.9.0
  • Cloud provider or hardware configuration: MacOS (arm64)
@dgrove-oss dgrove-oss added the kind/bug Categorizes issue or PR as related to a bug. label Nov 8, 2024
@kannon92
Copy link
Contributor

kannon92 commented Nov 9, 2024

1.27 is out of support in open source. Kueue 0.9.0 was probably built with 1.31 apis so I’m not sure the skew will even be supporting at this point.

@dgrove-oss
Copy link
Contributor Author

dgrove-oss commented Nov 11, 2024

On Kubernetes 1.29 (which I believe is still in support until 2025-02-28), I get a reduced set of repeating error messages but there is still a problem:

W1111 18:40:52.560632       1 reflector.go:561] k8s.io/client-go/informers/factory.go:160: failed to list *v1.ValidatingAdmissionPolicy: the server could not find the requested resource
E1111 18:40:52.560667       1 reflector.go:158] "Unhandled Error" err="k8s.io/client-go/informers/factory.go:160: Failed to watch *v1.ValidatingAdmissionPolicy: failed to list *v1.ValidatingAdmissionPolicy: the server could not find the requested resource" logger="UnhandledError"
W1111 18:41:35.548617       1 reflector.go:561] k8s.io/client-go/informers/factory.go:160: failed to list *v1.ValidatingAdmissionPolicy: the server could not find the requested resource
E1111 18:41:35.548652       1 reflector.go:158] "Unhandled Error" err="k8s.io/client-go/informers/factory.go:160: Failed to watch *v1.ValidatingAdmissionPolicy: failed to list *v1.ValidatingAdmissionPolicy: the server could not find the requested resource" logger="UnhandledError"
W1111 18:41:36.038435       1 reflector.go:561] k8s.io/client-go/informers/factory.go:160: failed to list *v1.ValidatingAdmissionPolicyBinding: the server could not find the requested resource
E1111 18:41:36.038474       1 reflector.go:158] "Unhandled Error" err="k8s.io/client-go/informers/factory.go:160: Failed to watch *v1.ValidatingAdmissionPolicyBinding: failed to list *v1.ValidatingAdmissionPolicyBinding: the server could not find the requested resource" logger="UnhandledError"
W1111 18:42:11.505717       1 reflector.go:561] k8s.io/client-go/informers/factory.go:160: failed to list *v1.ValidatingAdmissionPolicy: the server could not find the requested resource
E1111 18:42:11.505754       1 reflector.go:158] "Unhandled Error" err="k8s.io/client-go/informers/factory.go:160: Failed to watch *v1.ValidatingAdmissionPolicy: failed to list *v1.ValidatingAdmissionPolicy: the server could not find the requested resource" logger="UnhandledError"
W1111 18:42:15.087195       1 reflector.go:561] k8s.io/client-go/informers/factory.go:160: failed to list *v1.ValidatingAdmissionPolicyBinding: the server could not find the requested resource
E1111 18:42:15.087231       1 reflector.go:158] "Unhandled Error" err="k8s.io/client-go/informers/factory.go:160: Failed to watch *v1.ValidatingAdmissionPolicyBinding: failed to list *v1.ValidatingAdmissionPolicyBinding: the server could not find the requested resource" logger="UnhandledError"

@dgrove-oss dgrove-oss changed the title Repeated errors in kueue-manager log when Kueue 0.9 is installed on Kubernetes 1.27 Repeated errors in kueue-manager log when Kueue 0.9 is installed on Kubernetes 1.27 or 1.29 Nov 11, 2024
@dgrove-oss
Copy link
Contributor Author

I verified that on Kubernetes 1.30 there are no error messages in the log.

@dgrove-oss dgrove-oss changed the title Repeated errors in kueue-manager log when Kueue 0.9 is installed on Kubernetes 1.27 or 1.29 Repeated errors in kueue-manager log when Kueue 0.9 is installed on Kubernetes 1.29 Nov 11, 2024
@dgrove-oss
Copy link
Contributor Author

dgrove-oss commented Nov 11, 2024

retitled issue to only mention supported Kubernetes version (1.29).

@mbobrovskyi
Copy link
Contributor

mbobrovskyi commented Nov 12, 2024

As I can see here https://kubernetes.io/docs/reference/access-authn-authz/validating-admission-policy/, ValidatingAdmissionPolicy is stable starting from version v1.30. Kueue v0.9 build with v1.31 apis. That's why we have this error on 1.29.

@mimowo
Copy link
Contributor

mimowo commented Nov 12, 2024

@mbobrovskyi do you know if this is just a weird logging, or ValidatingAdmissionPolicy does not work in Kueue in 1.29?
I would suppose it should still work even while in Beta, but then it is still surprising to see this error.

@dgrove-oss
Copy link
Contributor Author

A quick glance at the 1.29 documentation seems to imply that although VAP was in Beta in 1.29, the feature gate still defaulted to false.

https://v1-29.docs.kubernetes.io/docs/reference/command-line-tools-reference/feature-gates/

@kannon92
Copy link
Contributor

kannon92 commented Nov 14, 2024

Are you able to see this with just kind?

@dgrove-oss
Copy link
Contributor Author

I was just using kind. (kind 0.25 with --image v1.29.10@sha256:3b2d8c31753e6c8069d4fc4517264cd20e86fd36220671fb7d0a5855103aa84b).

@mimowo
Copy link
Contributor

mimowo commented Nov 14, 2024

A quick glance at the 1.29 documentation seems to imply that although VAP was in Beta in 1.29, the feature gate still defaulted to false.

So, the question if you can enable it on 1.29 to mitigate the issue.
Another question would be if Kueue can do something to stop logging or make it less often - I don't know for now.

@kannon92
Copy link
Contributor

I reached out to @benluddy about this issue. This seems like an upstream issue and not really related to Kueue at the moment.

@benluddy
Copy link

Looks like kueue-controller-manager has started informers for the v1 VAP APIs, which did not exist in 1.27.

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

No branches or pull requests

5 participants