You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The issue report is detailed and includes version numbers where applicable.
This issue pertains to Kubeflow development.
I am available to work on this issue.
You can join the CNCF Slack and access our meetings at the Kubeflow Community website. Our channel on the CNCF Slack is here #kubeflow-platform.
Version
master
Detailed Description
First time posting here, just wanted to quickly thank y'all for maintaining this project.
Not sure if this issue belongs to this repo or kubeflow/pipelines. Will move the issue there if it's more appropriate there.
I was trying to deploy the master branch but got an issue where the minio pod for the kubeflow pipeline deployment suck in CrashLoopBackOff. This is a bare-metal kubeadm cluster with longhorn for storage.
After looking into the minio pod's log it seems to be some permission issue.
ERROR Unable to initialize backend: Unable to write to the backend
> Please ensure MinIO binary has write permissions for the backend
After checking out v1.9.1 and redeploying, everything worked, so I suspect the issue is related to the newer changes.
Steps to Reproduce
Run the single-command installation with the latest master branch to reproduce.
while! kustomize build example | kubectl apply --server-side --force-conflicts -f -;doecho"Retrying to apply resources"; sleep 20;done
Screenshots or Videos (Optional)
MinIO pod logs:
$ kubectl get logs -n kubeflow minio-4d823345b3-xzz73
ERROR Unable to initialize backend: Unable to write to the backend
> Please ensure MinIO binary has write permissions for the backend
HINT:
Verify if MinIO binary is running as the same user who has write permissions for the backend
Since v1.9.1 worked, I also ran git diff for the minio manifests. Looks like maybe the new changes to the minio-deployment.yaml is the root cause?
Validation Checklist
Version
master
Detailed Description
First time posting here, just wanted to quickly thank y'all for maintaining this project.
Not sure if this issue belongs to this repo or kubeflow/pipelines. Will move the issue there if it's more appropriate there.
I was trying to deploy the master branch but got an issue where the minio pod for the kubeflow pipeline deployment suck in CrashLoopBackOff. This is a bare-metal kubeadm cluster with longhorn for storage.
After looking into the minio pod's log it seems to be some permission issue.
After checking out v1.9.1 and redeploying, everything worked, so I suspect the issue is related to the newer changes.
Steps to Reproduce
Run the single-command installation with the latest master branch to reproduce.
Screenshots or Videos (Optional)
MinIO pod logs:
Since v1.9.1 worked, I also ran git diff for the minio manifests. Looks like maybe the new changes to the minio-deployment.yaml is the root cause?
The text was updated successfully, but these errors were encountered: