-
Notifications
You must be signed in to change notification settings - Fork 4
Affinity taint toleration filters #456
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
base: main
Are you sure you want to change the base?
Affinity taint toleration filters #456
Conversation
PhilippMatthes
left a 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.
Cool stuff!
| case corev1.NodeSelectorOpGt: | ||
| return exists && nodeValue > req.Values[0] | ||
| case corev1.NodeSelectorOpLt: | ||
| return exists && nodeValue < req.Values[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.
Is this safe?
| if canScheduleOnNode(node, request.Pod) { | ||
| activations[node.Name] = 1.0 | ||
| } |
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.
Since we're using a tanh activation function, your zero-activation would be 0.0. Zero-activation here means, it has no impact on the final result because tanh(0.0) = 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.
Technically, we're not too precise with the wording here. The activations map is actually a weights map which is going to be mapped to an activation by the activation function. We should rename this.
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.
This was there before, I just came across this and wanted to leave this note.
This PR refactors the directory structure for filters/weighers of the pod pipeline to align with the nova and manila pipeline. Secondly, filters are added that implement the default Kubernetes behavior for hard taints, tolerations and affinities. Soft requirements regarding these behaviors, which would need to be implemented as weighers, are not included in this PR and are subject for future work.