-
Notifications
You must be signed in to change notification settings - Fork 121
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
Map multiple vineyard objects to one volume while specifying the common prefix in the kubeflow benchmark #1578
Map multiple vineyard objects to one volume while specifying the common prefix in the kubeflow benchmark #1578
Conversation
…mmon prefix. * Update the kubeflow benchmark result and the related tutorial. Signed-off-by: Ye Cao <[email protected]>
@@ -14,6 +14,7 @@ Prerequisites | |||
guide `Initialize Kubernetes Cluster`_ to create one. | |||
- Install the `Vineyardctl`_ by following the official guide. | |||
- Install the `Argo Workflow CLI`_ by following the official guide. | |||
- Install the kfp package with version < 2.0.0. |
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.
Why we requires <2.0.0 here?
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.
The current demo is based on the kfp v1.8.0. In kfp v2, we need to rewrite the demo with the new API.
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.
As the kfp v2 has some breaking changes compared to v1. Backgound: https://www.kubeflow.org/docs/components/pipelines/v2/migration/
Actually, I think it doesn't matter as they are complied as the argo workflow YAML finally.
+------------+------------------+---------------+ | ||
| 15000 Mi | 332s | 286s | | ||
| 15000 Mi | 252s | 298s | | ||
+------------+------------------+---------------+ |
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 there are so huge improvements on "without vineyard"? and "with vineyard" becomes slower than "without vineyard"?
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.
Sorry, wrong writing here.
+------------+------------------+---------------+ | ||
| 15000 Mi | 128.6s | 0.04s | | ||
| 15000 Mi | 61.7s | 0.04s | | ||
+------------+------------------+---------------+ |
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 this PR affects the performances of "without vineyard" case?
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.
The previous time is wrong as it is mixed with the initial reading time. My bad.
@@ -1,12 +1,10 @@ | |||
FROM python:3.10 | |||
|
|||
RUN pip3 install --no-cache-dir pandas requests scikit-learn numpy vineyard | |||
RUN pip3 install pandas requests scikit-learn numpy vineyard |
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.
Use --no-cache-dir
to let pip cleanup the downloaded zips/wheels after installation.
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.
Done.
Signed-off-by: Ye Cao <[email protected]>
What do these changes do?
Related issue number
Fixes #1568