-
Notifications
You must be signed in to change notification settings - Fork 4k
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
Consider atomic nodes #6477
Consider atomic nodes #6477
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 one minor comment - let's address before merging.
/approve
@@ -274,7 +276,7 @@ func (p *Planner) categorizeNodes(podDestinations map[string]bool, scaleDownCand | |||
klog.Warningf("%d out of %d nodes skipped in scale down simulation due to timeout.", len(currentlyUnneededNodeNames)-i, len(currentlyUnneededNodeNames)) | |||
break | |||
} | |||
if len(removableList) >= p.unneededNodesLimit() { | |||
if len(removableList)-atomicScaleDownNodesCount >= p.unneededNodesLimit() { | |||
klog.V(4).Infof("%d out of %d nodes skipped in scale down simulation: there are already %d unneeded nodes so no point in looking for more.", len(currentlyUnneededNodeNames)-i, len(currentlyUnneededNodeNames), len(removableList)) |
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.
Let's include atomicScaleDownNodesCount
in the log.
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.
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 meant in this line: just report atomicScaleDownNodesCount
value. Doing it for every considered node is going to result in a lot of log spam.
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.
Makes sense. Done.
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: kushagra98, x13n The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
/lgtm |
What type of PR is this?
/kind bug
/kind regression
What this PR does / why we need it:
unneededNodesLimit()
stops considering scale down nodes when this limit is reached. However, if a nodepool which has nodes > this limit and only allowed for atomic scaledown will always starve.Eg. Consider a nodepool with 1000 nodes in a cluster and
unneededNodesLimit()
= 500. Let's consider that this nodepool is only allowed to scale down atomically. The nodepool will always starve for scale down as the limit is 500 but we cannot scaledown 500 nodes, we can only scale down 1000 nodes.Which issue(s) this PR fixes:
Fixes #
Special notes for your reviewer:
Does this PR introduce a user-facing change?
Additional documentation e.g., KEPs (Kubernetes Enhancement Proposals), usage docs, etc.: