kubeflow / manifests

A repository for Kustomize manifests
Apache License 2.0
821 stars 882 forks source link

Report images per Working Group #2463

Closed difince closed 1 year ago

difince commented 1 year ago

Extend the "extract_manifests_images.sh" script to be able to report both

There could be two separate scripts or one that accepts a parameter.

The unique images reported by 1. "images per WG" and 2. "All images" should be the same.

The current version of "extract_manifests_images" script checks out the following directories

declare -a dirs=("../apps" "../common" "../example" "../contrib/metacontroller"
                 "../contrib/seldon" "../contrib/bentoml" )

- apps
    - [X] admission-webhook/upstream
    - [X] centraldashboard/upstream
    - [X] jupyter
    - [X] katib/upstream
    - [ ???? ] kfp-tekton/upstream
    - [ ???? ] kubebench/upstream
    - [X] pipeline/upstream
    - [X] profiles/upstream
    - [X] tensorboard
    - [X] training-operator/upstream
    - [X] volumes-web-app/upstream

- [X] common
- [ ??? ] example
- [ ] contrib
    -  [ ??? ] metacontroller - Pipeline WG ? 
    -  [ ???  ] seldon
    -  [ ??? ] bentoml
Working Group Directories
AutoMl(Katib) /apps/katib
Pipelines /apps/pipeline/upstream
Training /apps/training-operator/upstream
Manifests /common
Notebooks/Workbenches /apps/admission-webhook/upstream
/apps/centraldashboard/upstream
/apps/jupyter/jupyter-web-app/upstream
/apps/volumes-web-app/upstream
/apps/tensorboard/tensorboards-web-app/upstream
/apps/profiles/upstream
/apps/jupyter/notebook-controller/upstream
/apps/tensorboard/tensorboard-controller/upstream

To be clarified: Which WG is responsible for directories marked with ???

Note: kserve is not considered. Is this ok ?

difince commented 1 year ago

cc: @annajung

kimwnasptd commented 1 year ago

Hi @difince, I can add some feedback.

Regarding contrib I'd suggest that we don't include images from there when either all or per-WG images. I would keep KServe ones though, due to KServe's history from starting as a Kubeflow project.

difince commented 1 year ago
Thank you for your feedback @kimwnasptd The table below contains the updated "WGs vs Manifests directories" mapping based on your inputs. I will update the PR accordingly. Working Group Directories
AutoML (Katib) /apps/katib
Pipelines /apps/pipeline/upstream
/apps/kfp-tekton/upstream
Training /apps/training-operator/upstream
Manifests /common
/example
Notebooks/Workbenches /apps/admission-webhook/upstream
/apps/centraldashboard/upstream
/apps/jupyter/jupyter-web-app/upstream
/apps/volumes-web-app/upstream
/apps/tensorboard/tensorboards-web-app/upstream
/apps/profiles/upstream
/apps/jupyter/notebook-controller/upstream
/apps/tensorboard/tensorboard-controller/upstream
Serving /contrib/kserve
difince commented 1 year ago

closing the issue - it is fixed by https://github.com/kubeflow/manifests/pull/2465