-
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
Make VPA recommender target memory percentile configurable #6700
Labels
kind/feature
Categorizes issue or PR as related to a new feature.
Comments
mKeRix
added
the
kind/feature
Categorizes issue or PR as related to a new feature.
label
Apr 10, 2024
will be fixed by merging #6660 |
Closing, as #6660 has been merged. |
@voelzmo: Closing this issue. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Which component are you using?:
Vertical Pod Autoscaler
Is your feature request designed to solve a problem? If so describe the problem this feature should solve.:
We have some workloads that follow a pattern of settings memory requests equal to limit, to avoid unexpected pod evictions for memory reclaiming. We would like to use VPA on these deployments as well, to auto-scale both CPU and memory. VPA however currently hardcodes the target percentile for memory to
0.9
, which in this scenario would lead to a constant underscaling of resources. This in turn leads to a somewhat regular stream of OOM kills in this scenario.Describe the solution you'd like.:
In #4799 a flag was introduced to make the target percentile for CPU configurable. It would be nice to have the same kind of flag for memory too.
Describe any alternative solutions you've considered.:
For now, this can be worked around by ensuring that VPA-controlled applications always have a memory request-to-limit ratio of more than 1. Nonetheless, fine-tuning the VPA suggestions would give us more flexibility for how we run things.
Additional context.:
I'm happy to try and contribute the code changes for this, I just want to make sure that this would be in line with what the project maintainers see for the project before investing the time. Hence, I created this issue to discuss.
The text was updated successfully, but these errors were encountered: