kubeflow/spark-operator

404 Not Found https://googlecloudplatform.github.io/spark-on-k8s-operator/index.yaml

Closed this issue · 13 comments

Can't pull the spark-on-k8s-operator Helm chart as of 6 hours ago with the latest environment deployment. Seems to have been triggered by @zijianjoy. Did the deployment not run from the correct branch or something?

404: https://googlecloudplatform.github.io/spark-on-k8s-operator/index.yaml

Getting same error

# [2024-03-13T18:58:30.175Z] Error:: looks like "https://googlecloudplatform.github.io/spark-on-k8s-operator" 
# is not a valid chart repository or cannot be reached: failed to fetch https://googlecloudplatform.github.io/spark-on-k8s-operator/index.yaml : 404 Not Found

I am also getting same error since few hours ago

same...

@kimisu1 You're right, looks like the yaml is there. Confirmed using https://kubeflow.github.io/spark-operator/ instead works. Cheers!

https://charts.helm.sh/incubator/ works too... but what is correct repo address now?

Changing to the following helm dependency worked for me this morning. Version omitted as I'm on an old version.

dependencies:
- name: spark-operator
  repository: https://kubeflow.github.io/spark-operator
  version: ?.?.?
  condition: spark-operator.enabled

Would be good to know if this is correct moving forward as others have said.

Look like in this deployment it deployed into https://kubeflow.github.io/spark-operator/ instead https://googlecloudplatform.github.io/spark-on-k8s-operator/

This worked for us too, thanks!!

same here: this one worked for me :
`resource "helm_release" "main" {
depends_on = [ var.module_depends_on, kubernetes_namespace.main, kubernetes_namespace.spark_apps ]

chart           = "spark-operator"

repository      = "https://kubeflow.github.io/spark-operator"
version         = "1.1.27"


name      		= "spark-operator"`

ScreenShot 14-March-2024 at 12 25 35 UTC Code-helm tf (Working Tree) (helm tf) — csi@2x

Thanks for reporting the issue. Yes, this issue is due to the repo migration to Kubeflow org. We will create a PR to update the Helm chart and verify the Github workflows.
FYI... @yuchaoran2011

Hello everyone,

yes I triggered the repo migration from googlecloudplatform to kubeflow, thus causing https://kubeflow.github.io/spark-operator/index.yaml being the new endpoint. You can find more detail about this move in #1928.

Thanks for taking the effort to migrate the rest of related artifacts. cc @andreyvelich to coordinate those changes.

Thanks for the guidance! I'll close this ticket, but please consider pinning this issue afterwards so that other people can more easily find the resolution to this issue.

There was no mention of this change in this repo. Additionally, I personally had no idea kubeflow/community even existed (no mention in README or anything), let alone to look there for issues like this.