Skip to content
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

Is there a way to configure flags on kube-controller-manager for HPA #1742

Open
DmitrySemchenok opened this issue Jul 21, 2020 · 36 comments
Open
Assignees
Labels

Comments

@DmitrySemchenok
Copy link

Hello. The question is following:
We currently need to set custom values to following flags on kube-controller-manager
--horizontal-pod-autoscaler-initial-readiness-delay
--horizontal-pod-autoscaler-cpu-initialization-period
Is there a way to configure those values on aks cluster?
If not then is there any other way to prevent HPA from collecting metrics of not ready deployments in cas of long pods initialization time?

@4c74356b41
Copy link

I'd like to add a more general question. is it possible to add\alter any flags at all? like enable alpha features?

@ghost
Copy link

ghost commented Jul 30, 2020

@Azure/aks-pm issue needs labels

3 similar comments
@ghost
Copy link

ghost commented Jul 31, 2020

@Azure/aks-pm issue needs labels

@ghost
Copy link

ghost commented Aug 1, 2020

@Azure/aks-pm issue needs labels

@ghost
Copy link

ghost commented Aug 2, 2020

@Azure/aks-pm issue needs labels

@TomGeske
Copy link

TomGeske commented Aug 3, 2020

+@palma21

@ghost
Copy link

ghost commented Sep 4, 2020

Action required from @Azure/aks-pm

@ghost ghost added the Needs Attention 👋 Issues needs attention/assignee/owner label Sep 4, 2020
@ghost
Copy link

ghost commented Sep 19, 2020

Issue needing attention of @Azure/aks-leads

8 similar comments
@ghost
Copy link

ghost commented Oct 5, 2020

Issue needing attention of @Azure/aks-leads

@ghost
Copy link

ghost commented Oct 20, 2020

Issue needing attention of @Azure/aks-leads

@ghost
Copy link

ghost commented Nov 4, 2020

Issue needing attention of @Azure/aks-leads

@ghost
Copy link

ghost commented Nov 19, 2020

Issue needing attention of @Azure/aks-leads

@ghost
Copy link

ghost commented Dec 4, 2020

Issue needing attention of @Azure/aks-leads

@ghost
Copy link

ghost commented Dec 19, 2020

Issue needing attention of @Azure/aks-leads

@ghost
Copy link

ghost commented Jan 4, 2021

Issue needing attention of @Azure/aks-leads

@ghost
Copy link

ghost commented Jan 19, 2021

Issue needing attention of @Azure/aks-leads

@caiotavares88
Copy link

Any updates on this? Have the same issue and would like to change the default values of those two flags.

@ghost
Copy link

ghost commented Feb 6, 2021

Issue needing attention of @Azure/aks-leads

3 similar comments
@ghost
Copy link

ghost commented Feb 21, 2021

Issue needing attention of @Azure/aks-leads

@ghost
Copy link

ghost commented Mar 8, 2021

Issue needing attention of @Azure/aks-leads

@ghost
Copy link

ghost commented Mar 23, 2021

Issue needing attention of @Azure/aks-leads

@ghost
Copy link

ghost commented Apr 8, 2021

Issue needing attention of @Azure/aks-leads

2 similar comments
@ghost
Copy link

ghost commented Apr 23, 2021

Issue needing attention of @Azure/aks-leads

@ghost
Copy link

ghost commented May 8, 2021

Issue needing attention of @Azure/aks-leads

@bhavishya123
Copy link

Any updates here ? Require this for updation of default values

@ghost
Copy link

ghost commented Jun 5, 2021

Issue needing attention of @Azure/aks-leads

2 similar comments
@ghost
Copy link

ghost commented Jun 20, 2021

Issue needing attention of @Azure/aks-leads

@ghost
Copy link

ghost commented Jul 5, 2021

Issue needing attention of @Azure/aks-leads

@palma21
Copy link
Member

palma21 commented Jul 18, 2021

For alpha flags: #2058

For those 2 HPA flags + @justindavies to evaluate

@ghost ghost removed action-required Needs Attention 👋 Issues needs attention/assignee/owner labels Jul 18, 2021
@palma21 palma21 added action-required feature-request Requested Features Needs Attention 👋 Issues needs attention/assignee/owner labels Jul 18, 2021
@ghost ghost removed action-required Needs Attention 👋 Issues needs attention/assignee/owner labels Jul 18, 2021
@biswarup1290dass
Copy link

Just checking, is this issue being tracked for any updates related to enabling/disabling K8s feature gate for AKS?
Or is there support for enabling/disabling feature gate on AKS already?

What i am looking for: Provision an AKS cluster with k8s v1.21.x with BoundServiceAccountTokenVolume feature gate disabled (its enabled by default starting 1.21+)

@annerajb
Copy link

i would love to also individually toggle feature gates such as MixedProtocolLBSVC already requested here #2193

@ghost ghost added the action-required label Jun 12, 2022
@dwardellitus
Copy link

Any movement going to happen on this issue? We have cases where modifications to these hpa flags could prove very useful.

@qpetraroia
Copy link
Contributor

@justindavies for viz

Thanks for reaching out folks, we appreciate it. We will take a look at this ask/issue.

@ghost ghost removed the action-required label Feb 3, 2023
@olsenme
Copy link
Contributor

olsenme commented Mar 16, 2023

dwardellitus

Can you provide more details here?

@dwardellitus
Copy link

Pretty much the same as the original issue. We have a couple of deployments that are rather CPU hungry during initialization, and last Fall we were occasionally observing scaling events after deploying new code but in an otherwise quiet environment (in one case horizontal then lead to vertical when additional pods could not be scheduled). We were hoping those flags would help us, and then I found this issue. The problem is less acute than I thought a few months ago, likely due to other tweaks to resourcing and load distribution. That said, I still see value in being able to adjust via the flags
--horizontal-pod-autoscaler-initial-readiness-delay
--horizontal-pod-autoscaler-cpu-initialization-period

@JCDeist
Copy link

JCDeist commented Jul 12, 2023

We're experiencing a similar issue. Our Pods are able to server traffic but are still cpu intensive when starting up. This causes flapping up and down of usage when we are starting up these containers. Having the ability to adjust these flags would help us prevent this issue and result in faster helm releases to our environment.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests