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.

Sign up for GitHub

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

Jump to bottom

Issues with image release workflow app #970

Closed
johnugeorge opened this issue Jun 11, 2018 · 5 comments · Fixed by #990
Closed

Issues with image release workflow app #970

johnugeorge opened this issue Jun 11, 2018 · 5 comments · Fixed by #990
Assignees
@kunmingg

Comments

@johnugeorge
Copy link
Member

I am looking at new automation app for the image release. https://github.com/kubeflow/kubeflow/blob/master/docs_dev/releasing.md#create-release-workflow

  1. Correct build_image.sh script is not picked up. It always points to

    kubeflow/kubeflow/automation/release.libsonnet

    Line 78 in c69e760

    local imageDir = srcRootDir + "/kubeflow/kubeflow/components/k8s-model-server/images";

  2. In order to release repos outside kubeflow/kubeflow, EXTRA_REPOS flag has to be set. However, this variable is overridden in the code and thus losing the user params.

    kubeflow/kubeflow/automation/release.libsonnet

    Line 234 in c69e760

    value: "kubeflow/testing@HEAD",

@johnugeorge
Copy link
Member Author

/cc @kunmingg

@johnugeorge johnugeorge mentioned this issue Jun 11, 2018
Merged
@johnugeorge
Copy link
Member Author

and what would be best way to test locally if release workflow is configured properly?

@johnugeorge johnugeorge mentioned this issue Jun 11, 2018
Closed
@kunmingg
Copy link
Contributor

Currently the workflow is configured to use kubeflow release cluster, and image will be pushed to public registry.
Best way to test is set up a k8s cluster with argo deployed and a image registry, change workflow pointing to test cluster and registry, then execute the workflow.

@kunmingg
Copy link
Contributor

Testing locally do requires some manual efforts.
Ideally once prototype is proved to work for multiple cases and we build enough confidence, we can merge PR and check if target image got released within 24 hrs.

@kunmingg
Copy link
Contributor

/assign @kunmingg

@johnugeorge johnugeorge mentioned this issue Jun 13, 2018
Closed
@kunmingg kunmingg mentioned this issue Jun 13, 2018
Merged
yanniszark pushed a commit to arrikto/kubeflow that referenced this issue Feb 15, 2021
@andreyvelich @k8s-ci-robot
Example with collecting timestamp of the metrics (kubeflow#970)
2a8e8b4 
* Increase Suggestion memLimit

* Create getSuggestionConfigData function

* Change memLimit for nasrl

* Change resources format for katib-config

* Create example with recording metrics timestamp

* Add comment line

* Change example from pytorch to mxnet

* Delete find_mxnet file

* Change mxnet-mnist-timestamp to mxnet-mnist

* Reduce num epochs in grid

* Enable autoscaling in CI cluster

* Add max nodes

* Add num nodes 6

* Increase num nodes

* Change num nodes to 6

* Remove autoscaling
surajkota pushed a commit to surajkota/kubeflow that referenced this issue Jun 13, 2022
@Bobgy
Fix KFP tensorboard no GCP permission (kubeflow#970)
5d1c05e
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@kunmingg kunmingg

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Release prototype update. kunmingg/kubeflow
3 participants
@jlewi @johnugeorge @kunmingg