issue_owner_repo listlengths 2 2 | issue_body stringlengths 0 261k ⌀ | issue_title stringlengths 1 925 | issue_comments_url stringlengths 56 81 | issue_comments_count int64 0 2.5k | issue_created_at stringlengths 20 20 | issue_updated_at stringlengths 20 20 | issue_html_url stringlengths 37 62 | issue_github_id int64 387k 2.91B | issue_number int64 1 131k |
|---|---|---|---|---|---|---|---|---|---|
[
"kubernetes",
"kubernetes"
] | ### Which jobs are flaking?
master-informing
- capz-windows-master
### Which tests are flaking?
`Kubernetes e2e suite.[It] [sig-windows] [Feature:WindowsHostProcessContainers] [MinimumKubeletVersion:1.22] HostProcess containers container command path validation`
, `attach-detach-controller` force detach the volume, and the volume has been removed from `node.status.... | PV stuck at release status after pod and PVC is deleted | https://api.github.com/repos/kubernetes/kubernetes/issues/118562/comments | 8 | 2023-06-08T07:48:28Z | 2024-03-28T09:36:07Z | https://github.com/kubernetes/kubernetes/issues/118562 | 1,747,297,077 | 118,562 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
In KubeVirt project, we now see a regression when running on Kubernetes `1.25.10 | 1.26.5 | 1.27.2`. If kubelet is restarted on a node, then all the existing and running workloads that use devices are terminated with `UnexpectedAdmissionError`:
```
Warning UnexpectedAdmissionError 45s kubele... | Running pods with devices are terminated if kubelet is restarted | https://api.github.com/repos/kubernetes/kubernetes/issues/118559/comments | 31 | 2023-06-08T06:44:14Z | 2024-06-19T05:41:26Z | https://github.com/kubernetes/kubernetes/issues/118559 | 1,747,190,177 | 118,559 |
[
"kubernetes",
"kubernetes"
] | ### Which jobs are failing?
master-informing
- gce-master-scale-performance
### Which tests are failing?
- ClusterLoaderV2.huge-service overall (testing/huge-service/config.yaml) [Changes](https://github.com/kubernetes/kubernetes/compare/bba9833c3...c4f44b3de?)
- ClusterLoaderV2.huge-service: [step: 05] Waiting fo... | [Failing Test] gce-master-scale-performance | https://api.github.com/repos/kubernetes/kubernetes/issues/118554/comments | 5 | 2023-06-08T04:15:48Z | 2023-06-14T07:21:54Z | https://github.com/kubernetes/kubernetes/issues/118554 | 1,747,052,352 | 118,554 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
We have seen PVC binding (no dynamic provisioning) to take minutes when there are thousands of released PVs in the cluster. This is common when using local PVs in the cloud, where nodes get deleted more frequently and leave behind stale local PV objects.
It looks like PV controller is getting thr... | PV controller is slow when there are many released PVs | https://api.github.com/repos/kubernetes/kubernetes/issues/118545/comments | 5 | 2023-06-07T21:31:28Z | 2024-08-02T23:22:29Z | https://github.com/kubernetes/kubernetes/issues/118545 | 1,746,716,530 | 118,545 |
[
"kubernetes",
"kubernetes"
] | ### Which jobs are flaking?
https://prow.k8s.io/view/gs/kubernetes-jenkins/pr-logs/pull/118515/pull-kubernetes-unit-go-compatibility/1666191128879173632
### Which tests are flaking?
k8s.io/kubernetes/vendor/k8s.io/legacy-cloud-providers/azure: TestConcurrentLockEntry
### Since when has it been flaking?
Seen here h... | TestConcurrentLockEntry times out | https://api.github.com/repos/kubernetes/kubernetes/issues/118528/comments | 7 | 2023-06-07T09:33:00Z | 2024-11-03T18:20:02Z | https://github.com/kubernetes/kubernetes/issues/118528 | 1,745,466,082 | 118,528 |
[
"kubernetes",
"kubernetes"
] | /sig apps
The code around Job controller exponential backoff delay has a lot of heritage. The purpose of this issue is to collect and discuss cleanup ideas. It was initiated by the backoff discussion under the KEP: https://github.com/kubernetes/enhancements/pull/3967#discussion_r1219690984.
Some fixes initiated b... | Cleanup the Job controller exponential backoff delay | https://api.github.com/repos/kubernetes/kubernetes/issues/118527/comments | 5 | 2023-06-07T08:21:47Z | 2023-06-16T15:58:21Z | https://github.com/kubernetes/kubernetes/issues/118527 | 1,745,335,216 | 118,527 |
[
"kubernetes",
"kubernetes"
] | Page related to the issue: https://kubernetes.io/docs/concepts/storage/projected-volumes/#example-configuration-secrets-nondefault-permission-mode
When i am creating a projected volume having two secrets mysecret and mysecret1 at the location /project-volume/my-guide/my-username and /project-volume/my-guide/my-passwo... | Configuring Projected volume with pod is always in CrashLoopBackoff State | https://api.github.com/repos/kubernetes/kubernetes/issues/118526/comments | 11 | 2023-06-07T07:10:12Z | 2023-06-22T04:20:02Z | https://github.com/kubernetes/kubernetes/issues/118526 | 1,745,209,402 | 118,526 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
We created informers to watch resources. But after a few days of normally running, we can't receive any events when the resource `clusterresourcebindings` changes(other resources and apiserver are normal). We found out that the corresponding goroutine has been hung for a long time:
![0880c4c80710a6... | Informer watching fails with goroutine hanging after a while | https://api.github.com/repos/kubernetes/kubernetes/issues/118525/comments | 5 | 2023-06-07T07:08:59Z | 2023-06-14T01:34:42Z | https://github.com/kubernetes/kubernetes/issues/118525 | 1,745,206,808 | 118,525 |
[
"kubernetes",
"kubernetes"
] | ### Failure cluster [5014820e8364ad0bace3](https://go.k8s.io/triage#5014820e8364ad0bace3)
Similar to https://github.com/kubernetes/kubernetes/issues/107991 and easily reproducible. The handler occasionally completes and sends a response before timing out.
##### Error text:
```
Failed
=== RUN TestTimeoutReque... | Failure cluster [5014820e...]: Flake in TestTimeoutRequestHeaders and TestTimeoutWithLogging | https://api.github.com/repos/kubernetes/kubernetes/issues/118523/comments | 3 | 2023-06-07T01:31:49Z | 2023-06-13T16:22:00Z | https://github.com/kubernetes/kubernetes/issues/118523 | 1,744,870,915 | 118,523 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
A configmap using string data:
```
apiVersion: "v1"
kind: "Secret"
metadata:
name: "example"
stringData:
foo: "bar"
type: "Opaque"
```
Each time it is applied: "kubectl apply -f secret.yml --server-side"
Will create a new resourceVersion and an updated managedFields time, but ot... | Repeated server side apply creates additional revisions | https://api.github.com/repos/kubernetes/kubernetes/issues/118519/comments | 10 | 2023-06-06T22:19:44Z | 2024-09-24T20:27:43Z | https://github.com/kubernetes/kubernetes/issues/118519 | 1,744,696,510 | 118,519 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
We've configured a a resource quota to restrict the number of LoadBalancer services which can be created in a namespace according to the documentation in https://kubernetes.io/docs/concepts/policy/resource-quotas/#object-count-quota.
However, setting `count/services.loadbalancers` in a `Resourc... | Documented resource quota keys `services.loadbalancers` and `services.nodeports` don't have an effect | https://api.github.com/repos/kubernetes/kubernetes/issues/118509/comments | 4 | 2023-06-06T14:24:31Z | 2023-06-13T20:06:10Z | https://github.com/kubernetes/kubernetes/issues/118509 | 1,743,986,018 | 118,509 |
[
"kubernetes",
"kubernetes"
] | ### Description
I have a virtual machine create by kubevirt, the virtual machine uses PVC to provide the system disk, and the PV is provided by ceph-csi.
My program quickly deleted the PVC after deleting the virtual machine, the PVC is deleted, but the PV is stuck in `release` state
1、the event of the PV is:
... | system disk volume unstage not invoked when virtual machine which created by kubevirt is deleted | https://api.github.com/repos/kubernetes/kubernetes/issues/118503/comments | 14 | 2023-06-06T11:34:53Z | 2024-03-26T22:19:01Z | https://github.com/kubernetes/kubernetes/issues/118503 | 1,743,674,249 | 118,503 |
[
"kubernetes",
"kubernetes"
] | ### What would you like to be added?
I would like to propose for Kubernetes to provide an official support for a restore operation. As Kubernetes stores all it's data in etcd, this feature would be mostly about defining an official guided way for users to restore their etcd.
## Background
Restoring Kubernetes... | Official support for restoring Kubernetes cluster | https://api.github.com/repos/kubernetes/kubernetes/issues/118501/comments | 23 | 2023-06-06T11:30:33Z | 2024-04-26T13:22:57Z | https://github.com/kubernetes/kubernetes/issues/118501 | 1,743,668,154 | 118,501 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
请问恶趣味
### What did you expect to happen?
企鹅额问问
### How can we reproduce it (as minimally and precisely as possible)?
请问恶趣味
### Anything else we need to know?
_No response_
### Kubernetes version
<details>
```console
$ kubectl version
# paste output here
```
</details>
其味无穷额
### Cl... | 企鹅额问问 | https://api.github.com/repos/kubernetes/kubernetes/issues/118494/comments | 4 | 2023-06-06T08:41:11Z | 2023-06-06T08:41:59Z | https://github.com/kubernetes/kubernetes/issues/118494 | 1,743,376,536 | 118,494 |
[
"kubernetes",
"kubernetes"
] | ### Failure cluster [a2ca9dc55a3733eab734](https://go.k8s.io/triage#a2ca9dc55a3733eab734)
##### Error text:
```
[FAILED] CPU usage exceeding limits:
node n1-standard-2-cos-stable-105-17412-101-13-ce9419ce:
container "runtime": expected 95th% usage < 0.600; got 0.672
In [It] at: test/e2e_node/resource_usage_te... | Failure cluster: Density [Serial] [Slow] create a batch of pods latency/resource should be within limit when create 10 pods with 0s interval | https://api.github.com/repos/kubernetes/kubernetes/issues/118491/comments | 8 | 2023-06-06T08:08:28Z | 2023-07-27T08:47:33Z | https://github.com/kubernetes/kubernetes/issues/118491 | 1,743,320,920 | 118,491 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
I deleted the folder `.kube`,
Unable to use command `kubectl get node`.
### What did you expect to happen?
rm -rf ~/.kube
### How can we reproduce it (as minimally and precisely as possible)?
kubectl get node
The connection to the server localhost:8080 was refused - did you specify the r... | How to recover `~/.kube` folder | https://api.github.com/repos/kubernetes/kubernetes/issues/118484/comments | 6 | 2023-06-06T03:12:39Z | 2023-06-06T20:18:45Z | https://github.com/kubernetes/kubernetes/issues/118484 | 1,743,000,791 | 118,484 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
1. In this way , nodeHasSyncedFn always be `false`
```
kubeInformers := informers.NewSharedInformerFactoryWithOptions(kubeDeps.KubeClient, 0, informers.WithTweakListOptions(func(options *metav1.ListOptions) {
options.FieldSelector = fields.Set{metav1.ObjectNameField: string(kubelet.NodeName)... | Informer HasSynced always been false | https://api.github.com/repos/kubernetes/kubernetes/issues/118482/comments | 6 | 2023-06-06T02:21:01Z | 2023-06-07T09:15:13Z | https://github.com/kubernetes/kubernetes/issues/118482 | 1,742,939,939 | 118,482 |
[
"kubernetes",
"kubernetes"
] | Server-Side Apply has different constraints for updating the managed fields and applying than typical objects, and sometimes the discrepency can cause problem.
When trying to update the `managedFields`, if the object doesn't validate, then the whole `managedFields` list is removed, the error is ignored as to continu... | Server-Side Apply should return a proper http status when validation fails | https://api.github.com/repos/kubernetes/kubernetes/issues/118476/comments | 26 | 2023-06-05T21:03:37Z | 2024-11-19T17:42:04Z | https://github.com/kubernetes/kubernetes/issues/118476 | 1,742,619,058 | 118,476 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
We observed the following series of events:
1) Pod is created with some attributes that later result in the pod being rejected at kubelet admission time
2) The pod is scheduled to a node
3) The pod is gracefully deleted by a controller, i.e. a deletion timestamp is set on the pod
4) The pod ... | Pod graceful deletion hangs if the kubelet already rejected that pod at admission time | https://api.github.com/repos/kubernetes/kubernetes/issues/118472/comments | 28 | 2023-06-05T19:04:39Z | 2024-01-07T13:44:28Z | https://github.com/kubernetes/kubernetes/issues/118472 | 1,742,404,368 | 118,472 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
When `PrepareResource` API fails there is no way to understand what happens because error is not reported.
### What did you expect to happen?
Error should be either logged or, better event should be recorded, so user would see it in the `kubelet describe pod` output.
### How can we reproduce it (... | DRA: Kubelet doesn't report anything when it fails to prepare dynamic resources | https://api.github.com/repos/kubernetes/kubernetes/issues/118468/comments | 11 | 2023-06-05T16:06:54Z | 2023-07-17T20:47:14Z | https://github.com/kubernetes/kubernetes/issues/118468 | 1,742,100,083 | 118,468 |
[
"kubernetes",
"kubernetes"
] | null | Kubelet doesn't report anything when it fails to prepare dynamic resources | https://api.github.com/repos/kubernetes/kubernetes/issues/118466/comments | 2 | 2023-06-05T15:54:58Z | 2023-06-05T16:10:53Z | https://github.com/kubernetes/kubernetes/issues/118466 | 1,742,071,329 | 118,466 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
I want to deploy a pod in my cluster to monitor NIC information on nodes. I mount `/proc/net/` of the host directory to my pod in hostPath mode.
However, when I enter the container, I find that the owner of the mount point is incorrect, and the content in the corresponding directory is not the co... | Failed to mount the hostPath /proc/net/ | https://api.github.com/repos/kubernetes/kubernetes/issues/118465/comments | 9 | 2023-06-05T14:56:30Z | 2024-03-28T09:36:07Z | https://github.com/kubernetes/kubernetes/issues/118465 | 1,741,959,685 | 118,465 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
https://dl.k8s.io/release/v1.27.2/bin/darwin/amd64/kubelet
https://dl.k8s.io/release/v1.27.2/bin/darwin/amd64/kubeadm
i use curl download file display error
This XML file does not appear to have any style information associated with it. The document tree is shown below.
<Error>
<Code>N... | download fail | https://api.github.com/repos/kubernetes/kubernetes/issues/118462/comments | 8 | 2023-06-05T14:05:39Z | 2023-06-06T01:28:35Z | https://github.com/kubernetes/kubernetes/issues/118462 | 1,741,866,216 | 118,462 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
I'm very new to k8s and having problem with kube-proxy pods.
After I join worker node to control plane with command "kubeadm join ~" the worker node status has continuously restarting kube-proxy, weave-net, nvidia-device-plugin pods with worker node status ready.
`kubectl get pods -n kube-sys... | kube-proxy pods continuously CrashLoopBackOff | https://api.github.com/repos/kubernetes/kubernetes/issues/118461/comments | 26 | 2023-06-05T13:13:42Z | 2025-03-05T07:05:57Z | https://github.com/kubernetes/kubernetes/issues/118461 | 1,741,760,461 | 118,461 |
[
"kubernetes",
"kubernetes"
] | ### What would you like to be added?
Currently many parts of `scheduler_perf` is implemented in [scheduler_perf_test.go](https://github.com/kubernetes/kubernetes/blob/master/test/integration/scheduler_perf/scheduler_perf_test.go),
and that prevents external scheduler from using scheduler_perf.
### Why is this neede... | make scheduler_perf easy to use from external | https://api.github.com/repos/kubernetes/kubernetes/issues/118459/comments | 2 | 2023-06-05T11:54:49Z | 2024-01-07T11:12:14Z | https://github.com/kubernetes/kubernetes/issues/118459 | 1,741,614,139 | 118,459 |
[
"kubernetes",
"kubernetes"
] | ### What would you like to be added?
Remove [`managedFields`](https://github.com/kubernetes/apimachinery/blob/v0.27.2/pkg/apis/meta/v1/types.go#L271) from the cached used by controllers in `kube-controller-manager`. If feasible, this work will be applied to other components having similar use patterns but have no use ... | kube-controller-manager should not cache managedFields | https://api.github.com/repos/kubernetes/kubernetes/issues/118454/comments | 5 | 2023-06-05T09:15:34Z | 2023-08-15T22:17:13Z | https://github.com/kubernetes/kubernetes/issues/118454 | 1,741,347,869 | 118,454 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
this is a followup of https://github.com/kubernetes/kubernetes/issues/109595
We improved the kubelet behavior by [making the devicemanager actually deliver the requested devices, failing admission otherwise](https://github.com/kubernetes/kubernetes/pull/116376). The workload/controllers can now d... | the kubelet should respect the pod priority when initializing after restart | https://api.github.com/repos/kubernetes/kubernetes/issues/118452/comments | 7 | 2023-06-05T08:04:53Z | 2023-06-07T06:48:05Z | https://github.com/kubernetes/kubernetes/issues/118452 | 1,741,211,944 | 118,452 |
[
"kubernetes",
"kubernetes"
] | Data as of June 3rd, 9:00 PM Eastern. Latest data can be found here:
http://storage.googleapis.com/k8s-metrics/failures-latest.json
Additional Context:
https://github.com/kubernetes/test-infra/issues/18600
| CI Job | Days Failed |
| --- | --- |
| ci-kubernetes-kubemark-100-gce-scheduler | 32 |
| ci-kuberne... | [sig-scability] Fix or reduce frequency or switch off perma-failing jobs | https://api.github.com/repos/kubernetes/kubernetes/issues/118450/comments | 7 | 2023-06-05T02:14:34Z | 2024-11-09T02:11:03Z | https://github.com/kubernetes/kubernetes/issues/118450 | 1,740,768,529 | 118,450 |
[
"kubernetes",
"kubernetes"
] | test grid:
https://prow.k8s.io/?job=periodic-release-verify-packages-debs
```
Err:11 http://deb.debian.org/debian stretch/main amd64 Packages
404 Not Found
Err:13 http://deb.debian.org/debian stretch-updates/main amd64 Packages
404 Not Found
Ign:14 http://deb.debian.org/debian stretch-updates/main all Pa... | [sig-release] periodic-release-verify-packages-debs failing consistently | https://api.github.com/repos/kubernetes/kubernetes/issues/118449/comments | 5 | 2023-06-05T02:08:55Z | 2023-06-06T11:59:03Z | https://github.com/kubernetes/kubernetes/issues/118449 | 1,740,762,851 | 118,449 |
[
"kubernetes",
"kubernetes"
] | https://testgrid.k8s.io/sig-release-releng-informing#build-packages-debs
```
noop
fakeroot debian/rules binary
mkdir -p usr/bin
curl --fail -sSL --retry 5 \
-o usr/bin/kubelet \
"https://dl.k8s.io/v1.27.2/bin/linux/arm/kubelet"
curl: (22) The requested URL returned error: 404
make: *** [debian/rules:12: ... | [sig-release] build-packages-debs failing consistently | https://api.github.com/repos/kubernetes/kubernetes/issues/118448/comments | 5 | 2023-06-05T02:03:51Z | 2023-06-06T12:53:02Z | https://github.com/kubernetes/kubernetes/issues/118448 | 1,740,757,000 | 118,448 |
[
"kubernetes",
"kubernetes"
] | Data as of June 3rd, 9:00 PM Eastern. Latest data can be found here:
http://storage.googleapis.com/k8s-metrics/failures-latest.json
| CI Job | Days Failed |
| --- | --- |
| [ci-kubernetes-e2e-gci-gce-autoscaling-migs](https://testgrid.k8s.io/sig-autoscaling-cluster-autoscaler#gci-gce-autoscaling-migs) | 389 |
... | [sig/auto-scaling] Fix or reduce frequency or switch off perma-failing jobs | https://api.github.com/repos/kubernetes/kubernetes/issues/118447/comments | 11 | 2023-06-05T01:22:11Z | 2024-03-23T07:40:00Z | https://github.com/kubernetes/kubernetes/issues/118447 | 1,740,719,913 | 118,447 |
[
"kubernetes",
"kubernetes"
] | Data as of June 3rd, 9:00 PM Eastern. Latest data can be found here:
http://storage.googleapis.com/k8s-metrics/failures-latest.json
| CI Job | Days Failed |
| --- | --- |
| [ci-kubernetes-e2e-gce-iscsi](https://testgrid.k8s.io/sig-storage-kubernetes#gce-iscsi) | 543 |
| [ci-kubernetes-e2e-gce-iscsi-serial](http... | [sig-storage] Fix or reduce frequency or switch off perma-failing jobs | https://api.github.com/repos/kubernetes/kubernetes/issues/118446/comments | 9 | 2023-06-05T01:18:16Z | 2023-06-23T08:58:35Z | https://github.com/kubernetes/kubernetes/issues/118446 | 1,740,717,282 | 118,446 |
[
"kubernetes",
"kubernetes"
] | Data as of June 3rd, 9:00 PM Eastern. Latest data can be found here:
http://storage.googleapis.com/k8s-metrics/failures-latest.json
| CI Job | Days Failed |
| --- | --- |
| ci-kubernetes-e2e-windows-containerd-gce-1-24 | 401 |
| ci-kubernetes-e2e-windows-containerd-gce... | [sig-windows] Fix or reduce frequency or switch off perma-failing jobs | https://api.github.com/repos/kubernetes/kubernetes/issues/118445/comments | 5 | 2023-06-04T14:47:46Z | 2023-06-08T17:40:16Z | https://github.com/kubernetes/kubernetes/issues/118445 | 1,740,413,811 | 118,445 |
[
"kubernetes",
"kubernetes"
] | Data as of June 3rd, 9:00 PM Eastern. Latest data can be found here:
http://storage.googleapis.com/k8s-metrics/failures-latest.json
| CI Job | Days Failed |
| --- | --- |
|[ci-cri-containerd-node-e2e-features](https://testgrid.k8s.io/sig-node-containerd#node-e2e-features) | 37 | ... | [sig-node] Fix or reduce frequency or switch off perma-failing jobs | https://api.github.com/repos/kubernetes/kubernetes/issues/118441/comments | 32 | 2023-06-04T01:04:19Z | 2023-12-14T13:49:27Z | https://github.com/kubernetes/kubernetes/issues/118441 | 1,740,035,692 | 118,441 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
I am trying to deploy kubernetes (kubeadm cluster) on ARM64 based Ubuntu 20.04 virtual machine.
I am getting the follwoing error when deploying kubeadm:
```
xxx@controller01:~$ sudo mkdir -p /etc/apt/keyrings/
xxx@controller01:~$ curl -fsSL https://packages.cloud.google.com/apt/doc/apt-key.g... | Kubernetes: Failed to fetch https://packages.cloud.google.com/apt/dists/kubernetes-xenial/main/binary-arm64 | https://api.github.com/repos/kubernetes/kubernetes/issues/118439/comments | 4 | 2023-06-03T15:38:49Z | 2023-06-03T16:08:39Z | https://github.com/kubernetes/kubernetes/issues/118439 | 1,739,630,399 | 118,439 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
Our feed, when rendered, doesn't say which project / product it's for.
Example (from my workplace):
> Auto-refreshing Official CVE Feed [CVE-2023-2878](https://github.com/kubernetes/kubernetes/issues/118419)
> secrets-store-csi-driver discloses service account tokens in logs
>
> A security... | CVE feed title doesn't mention project | https://api.github.com/repos/kubernetes/kubernetes/issues/118437/comments | 6 | 2023-06-03T09:59:38Z | 2023-11-22T00:02:38Z | https://github.com/kubernetes/kubernetes/issues/118437 | 1,739,344,349 | 118,437 |
[
"kubernetes",
"kubernetes"
] | A security issue was discovered in [secrets-store-csi-driver](https://github.com/kubernetes-sigs/secrets-store-csi-driver) where an actor with access to the driver logs could observe service account tokens. These tokens could then potentially be exchanged with external cloud providers to access secrets stored in cloud ... | CVE-2023-2878: secrets-store-csi-driver discloses service account tokens in logs | https://api.github.com/repos/kubernetes/kubernetes/issues/118419/comments | 1 | 2023-06-02T19:03:54Z | 2023-06-02T19:04:34Z | https://github.com/kubernetes/kubernetes/issues/118419 | 1,738,671,677 | 118,419 |
[
"kubernetes",
"kubernetes"
] | It's apprently allowed to say:
```
ports:
- containerPort: 1234
protocol: TCP
- containerPort: 1234
protocol: TCP
```
We should issue a warning. It's absolutely meaningless, but changing validation would potentially break people and normalizing it would potentially cause controller loops. | Need to add an API warning when 2 ContainerPorts are identical | https://api.github.com/repos/kubernetes/kubernetes/issues/118418/comments | 8 | 2023-06-02T18:31:30Z | 2023-07-18T05:50:19Z | https://github.com/kubernetes/kubernetes/issues/118418 | 1,738,631,734 | 118,418 |
[
"kubernetes",
"kubernetes"
] | It's apprently allowed to say:
```
ports:
- containerPort: 123
hostPort: 987
- containerPort: 456
hostPort: 987
hostIP: 1.2.3.4
```
We should issue a warning. The first hostPort binds to all IPs, the second binds the same hostPort to a specific IP, and we do not define the semantic. | Need to add an API warning when 2 ContainerPorts overlap hostPorts | https://api.github.com/repos/kubernetes/kubernetes/issues/118417/comments | 5 | 2023-06-02T18:31:06Z | 2023-07-18T05:50:28Z | https://github.com/kubernetes/kubernetes/issues/118417 | 1,738,631,293 | 118,417 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
I tried to use resource-consumer tool to generate enough workload on my nodes to crash them (either generating a NodeNotReady or SystemOOM event). However, when I used ./consume-cpu/consume-cpu binary in an Deployment with resoure-consumer image and told it to consume a specific amount of millicores... | Testing: ./consume-cpu/consume-cpu binary in resource-consumer tool didn't consume the specified millicores | https://api.github.com/repos/kubernetes/kubernetes/issues/118415/comments | 9 | 2023-06-02T18:24:11Z | 2024-03-23T23:41:58Z | https://github.com/kubernetes/kubernetes/issues/118415 | 1,738,623,198 | 118,415 |
[
"kubernetes",
"kubernetes"
] | ### Which jobs are failing?
pull-kubernetes-node-e2e-containerd-dra
### Which tests are failing?
The job seems to have some infra related issue
### Since when has it been failing?
Not sure, can't see the full test history
### Testgrid link
https://prow.k8s.io/job-history/gs/kubernetes-jenkins/pr-logs/directory/p... | pull-kubernetes-node-e2e-containerd-dra failing | https://api.github.com/repos/kubernetes/kubernetes/issues/118412/comments | 4 | 2023-06-02T16:41:41Z | 2023-06-07T17:27:19Z | https://github.com/kubernetes/kubernetes/issues/118412 | 1,738,500,911 | 118,412 |
[
"kubernetes",
"kubernetes"
] | The Kubernetes Service API allow us to specify an Application Protocol for each Service port. This is used as a hint for implementations to offer richer behavior for protocols that they understand.
According to the [docs](https://kubernetes.io/docs/concepts/services-networking/service/#application-protocol) one of t... | Support a list of AppProtocols | https://api.github.com/repos/kubernetes/kubernetes/issues/118407/comments | 12 | 2023-06-02T13:23:39Z | 2025-03-11T01:15:03Z | https://github.com/kubernetes/kubernetes/issues/118407 | 1,738,197,635 | 118,407 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
When configuring the readiness probe for the container in the pod and setting the periodSeconds: 3600 the probe does not seem to run.
readinessProbe:
exec:
command:
- bash
- /opt/readyprobe.sh
initialDelaySecond... | Pod readiness probe does not start when setting periodSeconds: 3600 | https://api.github.com/repos/kubernetes/kubernetes/issues/118396/comments | 16 | 2023-06-02T07:06:27Z | 2024-03-22T03:21:56Z | https://github.com/kubernetes/kubernetes/issues/118396 | 1,737,589,627 | 118,396 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
`curl '{baseUrl}/apis/batch/v1/cronjobs?limit=3&resourceVersion=0'` or
`curl '{baseUrl}/apis/batch/v1/cronjobs?limit=3&resourceVersion=0&resourceVersionMatch=NotOlderThan'`
retrieves all the data (and does not limit to only 3)
`curl '{baseUrl}/apis/batch/v1/cronjobs?limit=3'` retrieves only ... | K8s list resources api with resourceVersion=0 does not paginate | https://api.github.com/repos/kubernetes/kubernetes/issues/118394/comments | 4 | 2023-06-02T05:26:48Z | 2023-06-06T20:18:57Z | https://github.com/kubernetes/kubernetes/issues/118394 | 1,737,493,448 | 118,394 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
The control plane has two interfaces.
I set api-advertise-address to bind to one of the two interfaces, which is not the default.
It works, but kubectl reports as IPs of the Pods of the control plane with hostnetworking the IP of the default interface.
This causes issues to Prometheus that looks ... | Wrong IP address when master node has multiple interfaces | https://api.github.com/repos/kubernetes/kubernetes/issues/118385/comments | 5 | 2023-06-01T21:02:42Z | 2023-06-01T21:38:46Z | https://github.com/kubernetes/kubernetes/issues/118385 | 1,737,091,024 | 118,385 |
[
"kubernetes",
"kubernetes"
] | If the `TokenRequests` field is set in the `CSIDriver` spec, the tokens are generated and funneled from kubelet to driver as part of volume attributes map. The key for the tokens in the map is `csi.storage.k8s.io/serviceAccount.tokens`. The service account tokens can be used for workload identity in cloud providers. E... | Use `secrets` field in `NodePublishVolumeRequest` for sa tokens | https://api.github.com/repos/kubernetes/kubernetes/issues/118377/comments | 14 | 2023-06-01T14:07:56Z | 2024-05-08T16:09:43Z | https://github.com/kubernetes/kubernetes/issues/118377 | 1,736,415,149 | 118,377 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
Implementations like kube-proxy might direct service traffic to stale endpoint slice as if it was a valid endpoint.
### What did you expect to happen?
The EndpointSlice (have service object as owner reference) object must get removed when service selector is set to empty.
### How can we r... | EndpointSlice object is not removed when service selector is made to be empty | https://api.github.com/repos/kubernetes/kubernetes/issues/118376/comments | 26 | 2023-06-01T12:31:01Z | 2024-07-08T22:26:34Z | https://github.com/kubernetes/kubernetes/issues/118376 | 1,736,226,534 | 118,376 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
Set the cpu management static policy on the node and create a pod. When Pods are scheduled to the node, the cpu_manager_state file records the cpuset used by the container in the pod, but when the pod is removed, The cpu_manager_state file will still have cpuset information for pod that has been del... | cpu_manager_state is not updated when pod is deleted | https://api.github.com/repos/kubernetes/kubernetes/issues/118375/comments | 7 | 2023-06-01T09:29:29Z | 2023-08-15T03:31:57Z | https://github.com/kubernetes/kubernetes/issues/118375 | 1,735,876,077 | 118,375 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
root@control:/home/bennyzhou# journalctl -xeu kubelet
Jun 01 07:37:02 control kubelet[8234]: I0601 07:37:02.821839 8234 kubelet_node_status.go:70] "Attempting to register node" node="control"
Jun 01 07:37:02 control kubelet[8234]: E0601 07:37:02.822745 8234 kubelet_node_status.go:92] "Unable... | error on execution phase wait-control-plane: couldn't initialize a Kubernetes cluster on Ubuntu 22.04.02 LTS | https://api.github.com/repos/kubernetes/kubernetes/issues/118372/comments | 4 | 2023-06-01T07:52:15Z | 2023-06-01T09:35:15Z | https://github.com/kubernetes/kubernetes/issues/118372 | 1,735,707,187 | 118,372 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
I have a pod which initially has 0.5 CPU limit. I then resize it to follow 1.5 CPU limit and pod starts to use more resources until 1.5CPU. Problem arises when I resize pod CPU limit back to 0.5CPU. Instead of throttling the pod, kubernetes doesn't do anything and pod continues to use up to 1.5CP... | [FG:InPlacePodVerticalScaling] K8s fails to throttle pod so that it follows resized CPU limit | https://api.github.com/repos/kubernetes/kubernetes/issues/118371/comments | 37 | 2023-06-01T06:55:46Z | 2024-11-03T05:57:48Z | https://github.com/kubernetes/kubernetes/issues/118371 | 1,735,612,292 | 118,371 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
Due to limited character space, I've stored additional information in annotations on nodes. My custom scheduling plugins read these annotations to schedule Pods. However, when a Pod fails to be scheduled on a node, I want it to be removed from the Unschedulable queue. This can be achieved by modif... | scheduler: add nodeAnnotationsChanged event to trigger rescheduling | https://api.github.com/repos/kubernetes/kubernetes/issues/118369/comments | 28 | 2023-06-01T06:36:53Z | 2024-01-15T09:50:43Z | https://github.com/kubernetes/kubernetes/issues/118369 | 1,735,588,542 | 118,369 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
kubectl apply command application YAML information is inconsistent with the actual use of K8S
### What did you expect to happen?
The application information used by k8s should be consistent with the yaml definition information declared by the kubectl apply command
### How can we reproduce it (as ... | kubectl apply command application YAML information is inconsistent with the actual use of K8S | https://api.github.com/repos/kubernetes/kubernetes/issues/118367/comments | 16 | 2023-06-01T03:24:00Z | 2024-03-23T05:38:58Z | https://github.com/kubernetes/kubernetes/issues/118367 | 1,735,370,603 | 118,367 |
[
"kubernetes",
"kubernetes"
] | ### Which jobs are failing?
- https://testgrid.k8s.io/sig-windows-master-release#capz-windows-master-serial-slow
- https://testgrid.k8s.io/sig-windows-master-release#capz-windows-2022-master-serial-slow
### Which tests are failing?
- [sig-windows] [Feature:Windows] GMSA Full [Serial] [Slow] GMSA support can r... | Windows GMSA full tests failing | https://api.github.com/repos/kubernetes/kubernetes/issues/118363/comments | 1 | 2023-05-31T20:22:46Z | 2023-05-31T21:59:48Z | https://github.com/kubernetes/kubernetes/issues/118363 | 1,734,959,738 | 118,363 |
[
"kubernetes",
"kubernetes"
] | In Kubernetes 1.26, Kubernetes Enhancement Proposal, KEP-3352[1]: Aggregated Discovery, enhanced the discovery mechanism and supported aggregated discovery optimizing response for discovery api.
To support this KEP, Changes were made to both API server and client to support aggregated discovery mechanism. This feat... | Issue for Aggregated Discovery for Beta (v1.27) | https://api.github.com/repos/kubernetes/kubernetes/issues/118361/comments | 9 | 2023-05-31T18:13:22Z | 2023-06-12T15:18:11Z | https://github.com/kubernetes/kubernetes/issues/118361 | 1,734,749,316 | 118,361 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
When we removed the intree plugins - https://github.com/kubernetes/kubernetes/pull/116301
We are also removing the e2es and in-process losing coverage such as for pre-provisioned or inline volumes. This causes those code paths to be error prone.
IMO - my naive thinking was to remove the plugi... | Removal of intree storage plugin is causing us to lose important test coverage | https://api.github.com/repos/kubernetes/kubernetes/issues/118360/comments | 18 | 2023-05-31T17:35:12Z | 2024-07-14T15:59:10Z | https://github.com/kubernetes/kubernetes/issues/118360 | 1,734,685,394 | 118,360 |
[
"kubernetes",
"kubernetes"
] | ### What would you like to be added?
I understand there is no darwin kubelet (yet). However, it'd be super helpful if we can support building darwin/arm64 binaries for for control-plane components: apiserver, controller-manager and scheduler.
### Why is this needed?
An immediate beneficiary will be project [kwok](ht... | Support building darwin/arm64 binaries for kubernetes control-plane components | https://api.github.com/repos/kubernetes/kubernetes/issues/118359/comments | 8 | 2023-05-31T17:13:41Z | 2023-06-06T20:44:54Z | https://github.com/kubernetes/kubernetes/issues/118359 | 1,734,652,188 | 118,359 |
[
"kubernetes",
"kubernetes"
] | After running k8s conformance testing on kubernetes v1.27, occasionally we have namespaces stuck terminating.
**Basic test instructions:**
We use the latest official kubernetes kubekins image: `gcr.io/k8s-staging-test-infra/kubekins-e2e:latest-master`
We extract kubetest binary from kubekins latest-master dock... | IKS v1.27: `crd-webhook-*` namespaces stuck terminating | https://api.github.com/repos/kubernetes/kubernetes/issues/118357/comments | 11 | 2023-05-31T16:04:50Z | 2024-08-13T20:21:13Z | https://github.com/kubernetes/kubernetes/issues/118357 | 1,734,550,907 | 118,357 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
I have and exists deployment,the first time I apply It with specifying the rollingUpdate like this
```
strategy:
rollingUpdate:
maxSurge: 100%
maxUnavailable: 20%
type: RollingUpdate
```
And then I modify the deployment's yaml,remove the strategy Setting,according to ... | Deployment rolling update `maxSurge` and `maxUnavailable` not changed if I don't specify them | https://api.github.com/repos/kubernetes/kubernetes/issues/118350/comments | 15 | 2023-05-31T10:37:00Z | 2024-03-22T18:28:59Z | https://github.com/kubernetes/kubernetes/issues/118350 | 1,733,905,881 | 118,350 |
[
"kubernetes",
"kubernetes"
] | Currently, k/k and all dependencies depend on [gnostic v0.5.7-v3refs](https://github.com/kubernetes/kubernetes/blob/master/go.mod#L42) because we did not want to capture the additional dependencies introduced with later versions of gnostic. We use gnostic for OpenAPI protobuf serialization/deserialization. Gnostic was ... | [Umbrella Issue] Replace google/gnostic import with google/gnostic-models | https://api.github.com/repos/kubernetes/kubernetes/issues/118340/comments | 4 | 2023-05-30T22:08:40Z | 2024-11-12T16:02:57Z | https://github.com/kubernetes/kubernetes/issues/118340 | 1,733,088,598 | 118,340 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
I discovered a potential data race in the program based on code analysis. Specifically, there is a write operation on the field "pathList", where the write operation at:packages/packages.go:782. In addition, the statement is executed many times in the goroutine where the static creation site is: typ... | A potential data race in packages.go | https://api.github.com/repos/kubernetes/kubernetes/issues/118337/comments | 4 | 2023-05-30T17:07:40Z | 2023-05-31T06:31:40Z | https://github.com/kubernetes/kubernetes/issues/118337 | 1,732,653,952 | 118,337 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
It seems the same root reason with the issue#116497. There are two conflicting operations, and a data race may happen when the two functions which the two operations are located are called in parallel.
### What did you expect to happen?
Adding the related lock to protect the access to "fields".
#... | proto: protect field access with lock to avoid possible data race | https://api.github.com/repos/kubernetes/kubernetes/issues/118336/comments | 4 | 2023-05-30T16:44:54Z | 2023-05-31T06:13:28Z | https://github.com/kubernetes/kubernetes/issues/118336 | 1,732,622,994 | 118,336 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
I notice that all the p.c.Wait() and p.c.Signal() are protected by the mutex lock p.mu in https://github.com/kubernetes/kubernetes/blob/master/vendor/golang.org/x/net/http2/pipe.go#LL76C13-L76C13.
So I doubt whether this situation can cause a deadlock.
Maybe what I said is not right, thank y... | potential leak in pipe.go | https://api.github.com/repos/kubernetes/kubernetes/issues/118333/comments | 4 | 2023-05-30T16:09:24Z | 2023-05-31T05:22:54Z | https://github.com/kubernetes/kubernetes/issues/118333 | 1,732,569,291 | 118,333 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
Fixed inconsistency and also potential data race in transport/http2_client.go:
t.prevGoAwayID is read/written many times in transport/http2_client.go; 3 of 4 times it is protected by t.mu.Lock(); 1 out of 4 times it is read without a Lock, which is in func handleGoAway() on L1216.
A data race may... | transport: protect field access with lock to avoid possible data race | https://api.github.com/repos/kubernetes/kubernetes/issues/118332/comments | 4 | 2023-05-30T16:03:30Z | 2023-05-31T06:10:53Z | https://github.com/kubernetes/kubernetes/issues/118332 | 1,732,559,754 | 118,332 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
I am trying to run a `kubectl diff --server-side --force-conflicts` against our cluster, but it fails with the following message:
```
The Job "test-job" is invalid: spec.template: Invalid value: core.PodTemplateSpec{...}: field is immutable
```
We use flux with the `force: true` option when ... | Kubectl diff fails when there are immutable fields | https://api.github.com/repos/kubernetes/kubernetes/issues/118325/comments | 18 | 2023-05-30T11:39:21Z | 2025-02-13T13:45:26Z | https://github.com/kubernetes/kubernetes/issues/118325 | 1,732,085,374 | 118,325 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
1. create a job, set job restartPolicy to Never;
``` yaml
apiVersion: batch/v1
kind: Job
metadata:
annotations:
description: ''
enable: true
name: test-created
namespace: default
lables: {}
spec:
completions: 1
parallelism: 1
activeDeadlineSeconds: null
template:... | pod(restartPolicy == Never) stuck pending when vm reboot | https://api.github.com/repos/kubernetes/kubernetes/issues/118321/comments | 10 | 2023-05-30T02:19:42Z | 2024-09-04T07:59:50Z | https://github.com/kubernetes/kubernetes/issues/118321 | 1,731,407,672 | 118,321 |
[
"kubernetes",
"kubernetes"
] | ERROR: type should be string, got "https://github.com/kubernetes/kubernetes/blob/76bad75c1e64e4b6af2c8aa1730b91ddbb0b2035/pkg/kubelet/pluginmanager/operationexecutor/operation_generator.go#L115\r\n\r\nI'm sorry, please forgive me, my English is not very good\r\n\r\nk8s version:\r\n1.25.3\r\n\r\nwhat happened:\r\nI deployed a csidriver on my k8s cluster, then I reboot the system.\r\nAfter reboot, I change some net config what make csidriver execute NodeGetInfo() failed.\r\nI back up the net config subsequently.\r\nHowever I can't get csidriver on my k8s cluster.\r\nI find some log \"Error calling CSI NodeGetInfo()\" in the kubelet log.\r\n\r\nexpect result:\r\nI hopo k8s can register the csidriver after install failed unless I unregister the csidriver.\r\n\r\nin this file: ``pkg/kubelet/pluginmanager/operationexecutor/operation_generator.go#L115``\r\nactualStateOfWorldUpdater update the cache before csi install \r\nwhich make ``reconcile`` don't unregister and register the csidriver after meet an install csi error\r\n\r\nI think it can be modified like this (experimental)\r\n```go\r\nfunc (og *operationGenerator) GenerateRegisterPluginFunc(\r\n\tsocketPath string,\r\n\ttimestamp time.Time,\r\n\tpluginHandlers map[string]cache.PluginHandler,\r\n\tactualStateOfWorldUpdater ActualStateOfWorldUpdater) func() error {\r\n\r\n\tregisterPluginFunc := func() error {\r\n\t\tclient, conn, err := dial(socketPath, dialTimeoutDuration)\r\n\t\tif err != nil {\r\n\t\t\treturn fmt.Errorf(\"RegisterPlugin error -- dial failed at socket %s, err: %v\", socketPath, err)\r\n\t\t}\r\n\t\tdefer conn.Close()\r\n\r\n\t\tctx, cancel := context.WithTimeout(context.Background(), time.Second)\r\n\t\tdefer cancel()\r\n\r\n\t\tinfoResp, err := client.GetInfo(ctx, ®isterapi.InfoRequest{})\r\n\t\tif err != nil {\r\n\t\t\treturn fmt.Errorf(\"RegisterPlugin error -- failed to get plugin info using RPC GetInfo at socket %s, err: %v\", socketPath, err)\r\n\t\t}\r\n\r\n\t\thandler, ok := pluginHandlers[infoResp.Type]\r\n\t\tif !ok {\r\n\t\t\tif err := og.notifyPlugin(client, false, fmt.Sprintf(\"RegisterPlugin error -- no handler registered for plugin type: %s at socket %s\", infoResp.Type, socketPath)); err != nil {\r\n\t\t\t\treturn fmt.Errorf(\"RegisterPlugin error -- failed to send error at socket %s, err: %v\", socketPath, err)\r\n\t\t\t}\r\n\t\t\treturn fmt.Errorf(\"RegisterPlugin error -- no handler registered for plugin type: %s at socket %s\", infoResp.Type, socketPath)\r\n\t\t}\r\n\r\n\t\tif infoResp.Endpoint == \"\" {\r\n\t\t\tinfoResp.Endpoint = socketPath\r\n\t\t}\r\n\t\tif err := handler.ValidatePlugin(infoResp.Name, infoResp.Endpoint, infoResp.SupportedVersions); err != nil {\r\n\t\t\tif err = og.notifyPlugin(client, false, fmt.Sprintf(\"RegisterPlugin error -- plugin validation failed with err: %v\", err)); err != nil {\r\n\t\t\t\treturn fmt.Errorf(\"RegisterPlugin error -- failed to send error at socket %s, err: %v\", socketPath, err)\r\n\t\t\t}\r\n\t\t\treturn fmt.Errorf(\"RegisterPlugin error -- pluginHandler.ValidatePluginFunc failed\")\r\n\t\t}\r\n\r\n\t\tif err := handler.RegisterPlugin(infoResp.Name, infoResp.Endpoint, infoResp.SupportedVersions); err != nil {\r\n\t\t\treturn og.notifyPlugin(client, false, fmt.Sprintf(\"RegisterPlugin error -- plugin registration failed with err: %v\", err))\r\n\t\t}\r\n\r\n\t\tgo func() {\r\n\t\t\t_ = og.notifyPlugin(client, true, \"\")\r\n\t\t}()\r\n\r\n\t\terr = actualStateOfWorldUpdater.AddPlugin(cache.PluginInfo{\r\n\t\t\tSocketPath: socketPath,\r\n\t\t\tTimestamp: timestamp,\r\n\t\t\tHandler: handler,\r\n\t\t\tName: infoResp.Name,\r\n\t\t})\r\n\t\tif err != nil {\r\n\t\t\tklog.ErrorS(err, \"RegisterPlugin error -- failed to add plugin\", \"path\", socketPath)\r\n\t\t}\r\n\t\treturn nil\r\n\t}\r\n\treturn registerPluginFunc\r\n}\r\n```" | csidriver register failed and k8s don't try register again | https://api.github.com/repos/kubernetes/kubernetes/issues/118317/comments | 4 | 2023-05-29T16:25:21Z | 2023-06-16T02:54:50Z | https://github.com/kubernetes/kubernetes/issues/118317 | 1,731,009,629 | 118,317 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
When adding `--feature-gates=CloudDualStackNodeIPs=true` to kubelet and to the OpenStack external cloud provider, which is based on kubernetes v1.27.1, the node annotation `alpha.kubernetes.io/provided-node-ip ` only shows one address and consequently the node addresses also only displays one addres... | Node addresses only contains one address when using dual-stack | https://api.github.com/repos/kubernetes/kubernetes/issues/118315/comments | 3 | 2023-05-29T13:34:48Z | 2023-06-12T18:44:12Z | https://github.com/kubernetes/kubernetes/issues/118315 | 1,730,774,529 | 118,315 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
Since the field is bool and its tag has 'omitempty', it is impossible to set it to false and perform JSON marshalling. For example, when sending [response](https://kubernetes.io/docs/reference/access-authn-authz/authentication/#webhook-token-authentication) to KubeAPI server from a webhook-token ser... | Authenticated field in api/authentication/v1beta1TokenReviewStatus object is omitempty when it is needed to be false | https://api.github.com/repos/kubernetes/kubernetes/issues/118313/comments | 3 | 2023-05-29T11:03:09Z | 2024-03-24T01:52:25Z | https://github.com/kubernetes/kubernetes/issues/118313 | 1,730,557,640 | 118,313 |
[
"kubernetes",
"kubernetes"
] | /sig node
After https://github.com/kubernetes/kubernetes/pull/115331 the phase of terminal pods killed with killPodFunc is determined based on the exit codes of the containers. This was discovered in the context of preemption while doing a cherry-pick to 1.26, and observing that the phase changed: https://github.com... | Pod phase changed when container exits with 0 in 1.27 | https://api.github.com/repos/kubernetes/kubernetes/issues/118310/comments | 45 | 2023-05-29T08:10:34Z | 2023-12-01T15:50:42Z | https://github.com/kubernetes/kubernetes/issues/118310 | 1,730,301,752 | 118,310 |
[
"kubernetes",
"kubernetes"
] | ### Which jobs are failing?
master-informing
- gce-master-scale-performance
### Which tests are failing?
- ci-kubernetes-e2e-gce-scale-performance.Overall [Changes](https://github.com/kubernetes/kubernetes/compare/ffc034640...d05b79c83?)
- ci-kubernetes-e2e-gce-scale-performance.Pod [Changes](https://github.com/ku... | [Failing Test] gce-master-scale-performance | https://api.github.com/repos/kubernetes/kubernetes/issues/118305/comments | 4 | 2023-05-29T02:37:29Z | 2023-05-30T02:08:13Z | https://github.com/kubernetes/kubernetes/issues/118305 | 1,729,898,308 | 118,305 |
[
"kubernetes",
"kubernetes"
] | ### Which jobs are failing?
master-informing
- capz-windows-master
### Which tests are failing?
- Kubernetes e2e suite.[It] [sig-apps] CronJob should delete successful finished jobs with limit of one successful job
- Kubernetes e2e suite.[It] [sig-apps] CronJob should remove from active list jobs that have been de... | [Failing Test] capz-windows-master | https://api.github.com/repos/kubernetes/kubernetes/issues/118300/comments | 11 | 2023-05-28T07:02:10Z | 2023-06-01T17:37:45Z | https://github.com/kubernetes/kubernetes/issues/118300 | 1,729,246,516 | 118,300 |
[
"kubernetes",
"kubernetes"
] | ### Which jobs are failing?
master-informing
- gce-master-scale-correctness
### Which tests are failing?
- Kubernetes e2e suite.[It] [sig-network] Services GCE [Slow] should be able to create and tear down a standard-tier load balancer [Slow]
.
Do we have any known bug in this version related to this NW related issues? I moved contol-plane VM to ano... | Kubectl is hanging intermittently and recovering itself after few seconds | https://api.github.com/repos/kubernetes/kubernetes/issues/118291/comments | 10 | 2023-05-26T19:17:26Z | 2023-11-10T10:43:30Z | https://github.com/kubernetes/kubernetes/issues/118291 | 1,728,120,388 | 118,291 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
If I power cycle the node after tainting the node, certain pods are stuck in "Terminating" state(these pods usually take more time to delete)
kubectl taint nodes control-plane-dc275-2-master-n5-1003 node-role.kubernetes.io/master:NoExecute
poweroff -f --reboot
pod is stuck in Terminating stat... | Pod Stuck in Terminating state if tainted node is power cycled.... | https://api.github.com/repos/kubernetes/kubernetes/issues/118286/comments | 22 | 2023-05-26T16:01:13Z | 2023-09-12T16:02:23Z | https://github.com/kubernetes/kubernetes/issues/118286 | 1,727,880,371 | 118,286 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
Given a cronjob created with a scedule of "0 6 * * *" that failed at last run (status.lastSuccessfulTime is now "before" status.lastScheduleTime).
Then I trigger the cronjob by hand `kubectl create job --from=cronjob/my-cronjob my-cronjob-manual-1234`and this jobs ends successfully. However st... | Cronjob status.lastSuccessfulTime not populated by a manually triggered job | https://api.github.com/repos/kubernetes/kubernetes/issues/118276/comments | 9 | 2023-05-26T07:19:42Z | 2023-06-13T15:22:01Z | https://github.com/kubernetes/kubernetes/issues/118276 | 1,727,046,632 | 118,276 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
When I used containerd as k8s, I had a problem adding containerd to an existing cluster. When I changed the storage path of Containerd to k8s, I failed to add the storage path of Containerd to k8s. K8s showed no docker driver.
, but the log file in `/var/log/pods/<namespace-pod-id>/<container name>/<... | Container log file name is not incremented after 30 restarts | https://api.github.com/repos/kubernetes/kubernetes/issues/118255/comments | 3 | 2023-05-25T10:30:53Z | 2023-05-25T12:02:16Z | https://github.com/kubernetes/kubernetes/issues/118255 | 1,725,548,723 | 118,255 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
I am trying to patch to a configmap but am facing issues with the same. I have tried recommendations from other posts but havent been able to fix the same. Find below what I have tried and the error I get:
1)
kubectl patch configmap dec --namespace=namespace1 --patch {"data": {"config": {"static... | kubectl patch command issue when passing from command line | https://api.github.com/repos/kubernetes/kubernetes/issues/118246/comments | 10 | 2023-05-25T04:45:02Z | 2024-03-21T07:11:34Z | https://github.com/kubernetes/kubernetes/issues/118246 | 1,725,076,402 | 118,246 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
I had a pod with containers that has startupProbe configured.
Then I restart kubelet,
Containers with startupProbe configured will be mark with containerStatuses.started=false
After startupProbe runs again and then set containerStatuses.started=true
Is this expected behavior?
For the cod... | startupProbe will runs again after kubelet restart | https://api.github.com/repos/kubernetes/kubernetes/issues/118245/comments | 15 | 2023-05-25T04:29:17Z | 2024-03-21T17:19:30Z | https://github.com/kubernetes/kubernetes/issues/118245 | 1,725,065,736 | 118,245 |
[
"kubernetes",
"kubernetes"
] | I use Limit option to limit the number of returning nodes, but it returns all the data.
```
nodeList, err := client.CoreV1().Nodes().List(context.TODO(), metav1.ListOptions{
Limit: 10,
})
```
I notice the comment of Limit field: `Servers may choose not to support the limit argument and will return all of the... | client-go: ListOptions with specific Limit option not working | https://api.github.com/repos/kubernetes/kubernetes/issues/118243/comments | 2 | 2023-05-25T01:47:20Z | 2023-05-25T02:35:35Z | https://github.com/kubernetes/kubernetes/issues/118243 | 1,724,954,273 | 118,243 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
Auto generated code based on the swagger spec ` is missing the fields `x-kubernetes-patch-merge-key` and `x-kubernetes-patch-strategy` for `io.k8s.apiextensions-apiserver.pkg.apis.apiextensions.v1.JSONSchemaProps`.
A CRD generated with and validated against the schema will not be able to set the... | Missing `x-kubernetes-patch-merge-key` and `x-kubernetes-patch-strategy` for JSONSchemaProps in Spec | https://api.github.com/repos/kubernetes/kubernetes/issues/118241/comments | 3 | 2023-05-24T21:27:27Z | 2023-06-01T16:48:36Z | https://github.com/kubernetes/kubernetes/issues/118241 | 1,724,761,462 | 118,241 |
[
"kubernetes",
"kubernetes"
] | When I call a Kubernetes List Api (consecutively two times without any changes in the cluster) with `watch=null` for all namespaces for any resource with `resourceVersion = null`, I notice that the response for the first time has a list ResourceVersion (along with individual item ResourceVersion). The response for the ... | Kubernetes List Api ResourceVersion changes for consecutive same request | https://api.github.com/repos/kubernetes/kubernetes/issues/118239/comments | 6 | 2023-05-24T18:30:14Z | 2023-06-01T03:04:14Z | https://github.com/kubernetes/kubernetes/issues/118239 | 1,724,537,771 | 118,239 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
After installing apt-key.gpg file and declared kubernetes repository in apt source file, then updating apt I get the following error :
The key(s) in the keyring /etc/apt/trusted.gpg.d/kubernetes-archive-keyring.gpg are ignored as the file has an unsupported filetype
and the package can't be in... | GPG Key error in Debian 11 | https://api.github.com/repos/kubernetes/kubernetes/issues/118238/comments | 4 | 2023-05-24T17:00:14Z | 2023-05-24T17:03:20Z | https://github.com/kubernetes/kubernetes/issues/118238 | 1,724,415,738 | 118,238 |
[
"kubernetes",
"kubernetes"
] | ### Which jobs are flaking?
pull-kubernetes-e2e-gci-gce-ipvs
### Which tests are flaking?
[It] [sig-network] LoadBalancers should be able to preserve UDP traffic when server pod cycles for a LoadBalancer service on the same nodes
### Since when has it been flaking?
known flake, documenting it here.
### Testgrid ... | IPVS flaky test gce | https://api.github.com/repos/kubernetes/kubernetes/issues/118233/comments | 15 | 2023-05-24T12:44:54Z | 2024-03-05T10:16:16Z | https://github.com/kubernetes/kubernetes/issues/118233 | 1,723,906,151 | 118,233 |
[
"kubernetes",
"kubernetes"
] | I have built the Kubernetes repository from source and, in particular, imported with `docker import` the generated image of `kube-proxy`. However, I'm having a hard time running this newly created image on my fresh k8s cluster v1.27. What I'm trying to achieve for now is simply replacing the standard 1.27 kube-proxy im... | Corrupted Docker images when imported from tar with `docker import`. | https://api.github.com/repos/kubernetes/kubernetes/issues/118231/comments | 9 | 2023-05-24T11:56:28Z | 2023-05-25T20:58:37Z | https://github.com/kubernetes/kubernetes/issues/118231 | 1,723,815,088 | 118,231 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
Given the rule:
```
"!has(object.metadata.labels) || (has(object.metadata.labels) && !object.metadata.labels.helm__dot__sh__slash__chart == 'other')"
```
does not seem to expand the \_\_dot\_\_ and \_\_slash\_\_ because it gives the error
```
Error from server (Invalid): error when creating... | ValidatingAdmissionPolicy CEL validation rules do not seem to expand `__dot__` and `__slash__` as written in the documentation | https://api.github.com/repos/kubernetes/kubernetes/issues/118230/comments | 17 | 2023-05-24T11:48:05Z | 2024-03-24T01:51:11Z | https://github.com/kubernetes/kubernetes/issues/118230 | 1,723,799,955 | 118,230 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
1. A Pod named `PodA` is being scheduled.
2. PodA is updated by someone.
3. The scheduler's EventHandler notices the update in PodA. And let the scheduling queue know about the update via `PriorityQueue.Update` method.
4. `PriorityQueue.Update` checks where PodA exists in the scheduling queue... | Scheduler: same Pod can exist in both scheduling cycle and scheduling queue | https://api.github.com/repos/kubernetes/kubernetes/issues/118226/comments | 21 | 2023-05-24T09:35:56Z | 2024-06-26T13:58:13Z | https://github.com/kubernetes/kubernetes/issues/118226 | 1,723,573,707 | 118,226 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
In my cluster, there are around 3000 PVCs and approximately 7000 PVs. When I create a PVC, it takes more than 20 minutes for the PV to start creating.
Through analyzing the logs at a higher level, it was discovered that the PV controller re-adds all PVCs and PVs to the queue for synchronization e... | The PV controller took too long to add annotations volume.beta.kubernetes.io/storage-provisioner。 | https://api.github.com/repos/kubernetes/kubernetes/issues/118225/comments | 8 | 2023-05-24T09:02:26Z | 2024-10-15T09:11:31Z | https://github.com/kubernetes/kubernetes/issues/118225 | 1,723,509,715 | 118,225 |
[
"kubernetes",
"kubernetes"
] |
### What happened?
I'm trying to view the code for **cri streaming servers**. But except its test file, I couldn't find where the streaming server was initialized by calling method ```NewServer(config Config, runtime Runtime) (Server, error)``` throughout the project .
Can anyone be so nice to provide some infor... | [HELP] Where is the method NewServer in the kubelet/cri/streaming/server.go called? | https://api.github.com/repos/kubernetes/kubernetes/issues/118224/comments | 5 | 2023-05-24T08:18:09Z | 2023-06-19T07:08:42Z | https://github.com/kubernetes/kubernetes/issues/118224 | 1,723,434,938 | 118,224 |
[
"kubernetes",
"kubernetes"
] | ### Which jobs are flaking?
- ci-kubernetes-e2e-gci-gce-flaky
### Which tests are flaking?
3 different tests, related to SIG Storage and SIG API Machinery
see:
https://prow.k8s.io/view/gs/kubernetes-jenkins/logs/ci-kubernetes-e2e-gci-gce-flaky/1661222001538240512
https://prow.k8s.io/view/gs/kubernetes-jen... | [Flaky Test] storage gce flaky e2e tests multiple failures | https://api.github.com/repos/kubernetes/kubernetes/issues/118220/comments | 9 | 2023-05-24T07:21:08Z | 2024-11-03T18:20:02Z | https://github.com/kubernetes/kubernetes/issues/118220 | 1,723,344,039 | 118,220 |
[
"kubernetes",
"kubernetes"
] | ### What happened?
Kubernetes cluster present x509 errors about certs expired, Then restart the kube-apiserver, kube-controller-manager, kube-scheduler and etcd after execute "kubeadm alpha certs renew all " command successfully at one master node, Copy apiserver*, front* and etcd certs file to other two master node... | Etcd cannot running by static pod after kubeadm renew certs all | https://api.github.com/repos/kubernetes/kubernetes/issues/118216/comments | 5 | 2023-05-24T03:56:56Z | 2023-05-25T02:12:22Z | https://github.com/kubernetes/kubernetes/issues/118216 | 1,723,133,794 | 118,216 |
Subsets and Splits
Unique Owner-Repo Count
Counts the number of unique owner-repos in the dataset, providing a basic understanding of diverse repositories.