Skip to content
This repository has been archived by the owner on Apr 17, 2019. It is now read-only.

Cluster autoscaler requires all migs to be autoscaled #1283

Closed
mwielgus opened this issue Jun 28, 2016 · 3 comments · Fixed by #1282
Closed

Cluster autoscaler requires all migs to be autoscaled #1283

mwielgus opened this issue Jun 28, 2016 · 3 comments · Fixed by #1282
Assignees

Comments

@mwielgus
Copy link
Contributor

Otherwise it doesn't do any scale up or scale down.

@fgrzadkowski
Copy link
Contributor

@mwielgus Please, please, please! Be more verbose in PRs & issues! It'd be very hard for anyone to understand why is it an important bug and what was actually broken. We're working in a large project, with huge community and we should make sure that our people can understand what and why we are doing. This is really important.

@mwielgus
Copy link
Contributor Author

More details were added in the fixing PR approximately 40 min before your comment. #1282 (comment).
Sorry if you were confused, I did this fix well beyond my office hours and was too distracted by personal issues to think about elaborate description. And one "please" is enough, no need to get emotional.

The issues is quite straightforward and the problem is defined in the title, however I can rephrase it. If one mig in the cluster has autoscaling turned off then CA doesn't work. The exact explanation would be rather impossible without a reference to the implementation so it will be bette handled in the PR. Please take a look at the comment linked above.

@mwielgus
Copy link
Contributor Author

Fixed in #1282.

k8s-github-robot pushed a commit to kubernetes/kubernetes that referenced this issue Jun 29, 2016
Automatic merge from submit-queue

Check if Cluster Autoscaler still works if there is an extra non-autoscaled node pool in the cluster

Follow up for issue kubernetes-retired/contrib#1283.
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants