-
Notifications
You must be signed in to change notification settings - Fork 321
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.
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
[Feature][Doc] Kubeflow integration #937
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good! Just added some suggestions. I think the most important one is to include some brief intro/motivation at the top, the rest are minor.
``` | ||
|
||
## Step 2: Install Kubeflow v1.6-branch | ||
* This example installs Kubeflow with the [v1.6-branch](https://github.com/kubeflow/manifests/tree/v1.6-branch). |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Why do we need the 1.6 branch specifically? Would be good to add a brief comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Kubeflow is very sensitive to the Kubernetes version and the Kustomize version, so some instructions in this document may not work with other releases. In addition, some components will be renamed in future Kubeflow releases, and I am not familiar with the Kubeflow community's plan.
``` | ||
* This step uses `rayproject/ray:2.2.0-py38-cpu` as its image. Ray is very sensitive to the Python versions and Ray versions between the server (RayCluster) and client (JupyterLab) sides. This image uses: | ||
* Python 3.8.13 | ||
* Ray 2.2.0 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I suppose we can use 2.3.0 now, but if it will require a lot more manual testing then it might not be worth it.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It is a huge effort. Ray client is very sensitive to the Python versions and Ray versions. Sometimes I use the same Ray commit in the JupyterLab instance and in the RayCluster's docker image, but it is still not compatible.
Signed-off-by: Kai-Hsun Chen <kaihsun@apache.org>
Co-authored-by: Archit Kulkarni <architkulkarni@users.noreply.github.com> Signed-off-by: Kai-Hsun Chen <kaihsun@apache.org>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ah forgot to mention, if we're adding a new .md
file here, do we need to add it to the table of contents YAML file you showed me earlier? Anyways, we can figure this out when we review the docs as a whole later.
Good point. I forget about it. We can reorganize the website later. |
Kubeflow integration
Why are these changes needed?
Use Kubeflow as a solution of interactive prototyping environment (JupyterLab).
Related issue number
Closes #725
Part of #502. The issue includes JupyterLab, access control, and authentication. This PR only includes JupyterLab.
Checks