Limit on maximum number of nodes that can be provisioned for a karpenter enabled cluster
liafizan opened this issue · 5 comments
Description
What problem are you trying to solve?
We are migrating from ASG based nodepools to karpenter driven nodepools. One feature that we would want to map from ASG world is to have a limit on number of nodes that can be provisioned for a cluster by karpenter.
I have gone through FAQs and limits but do not see anything defined for this.
In karpenter world, node-count limit on node pool does not add much value since a nodepool can have different instance types underneath it. Hence, request to have a cluster-wide limit.
How important is this feature to you?
It is an important feature as we would like to control our expenditure for AWS for our non production environments & limit clusters beyond CPU/Memory for a nodepool. We would like to have an overall limit for the cluster to have greater control on our costs+operations.
Our clusters are tested for specific scale and we see issues beyond that node limit. Would like to have control over this limit
- Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
- Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
- If you are interested in working on this issue or have submitted a pull request, please leave a comment
This issue is currently awaiting triage.
If Karpenter contributors determines this is a relevant issue, they will accept it by applying the triage/accepted label and provide further guidance.
The triage/accepted label can be added by org members by writing /triage accepted in a comment.
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.
It is an important feature as we would like to control our expenditure for AWS for our non production environments & limit clusters beyond CPU/Memory for a nodepool. We would like to have an overall limit for the cluster to have greater control on our costs
Can you provide more insight into your use case? How are you thinking a cluster wide node count limit will help control cost? With having a diversity of nodes within the cluster that will result in varying cost. Are you looking for gating mechanism on Karpenter provisioning based on cost?
I cannot think of a direct way in which a greater node count would result in higher AWS costs.
But note that this might be another example of the scenarios in: #1664 -- where there are additional third-party costs on the hosts in the pool, which leads to a desire to limit the pool size.
@engedaam: Closing this issue.
In response to this:
Closing this issue as a duplicate #745
/close
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.