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
(feel free to comment additional components that could be removed, and I will add them to the list)
If we want backwards compatibility, we could create separate Kustomize components for deprecated resources which are enabled by default, and disable them at some point in the future before completely removing them.
Additionally, in the profile controller, we could look for a V1_BACKWARDS_COMPATIBILITY environment variable to only conditionally deploy the visualization server and artifact UI.
What is the use case or pain point?
Installation of KFP using something like kubectl apply -k ./pipelines/manifests/kustomize/env/platform-agnostic-multi-user includes a lot of unnecessary resources, which just take up space and computational resources.
It is also hard to know which components are needed if you only run KFP v2.
Is there a workaround currently?
Continue as usual. Ignore the high resource usage from all the visualization servers and artifact UIs in all user namespaces.
Love this idea? Give it a 👍.
The text was updated successfully, but these errors were encountered:
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
Feature Area
/area backend
(there more be more areas I don't know of)
What feature would you like to see?
There are several components in this git repository that are no longer in use.
These include (original comment: #7725 (comment)):
sync.py
)(feel free to comment additional components that could be removed, and I will add them to the list)
If we want backwards compatibility, we could create separate Kustomize components for deprecated resources which are enabled by default, and disable them at some point in the future before completely removing them.
Additionally, in the profile controller, we could look for a
V1_BACKWARDS_COMPATIBILITY
environment variable to only conditionally deploy the visualization server and artifact UI.What is the use case or pain point?
Installation of KFP using something like
kubectl apply -k ./pipelines/manifests/kustomize/env/platform-agnostic-multi-user
includes a lot of unnecessary resources, which just take up space and computational resources.It is also hard to know which components are needed if you only run KFP v2.
Is there a workaround currently?
Continue as usual. Ignore the high resource usage from all the visualization servers and artifact UIs in all user namespaces.
Love this idea? Give it a 👍.
The text was updated successfully, but these errors were encountered: