Configuring SpotToSpotConsolidation per Nodepool

[ilkinmammadzada]

Description

What problem are you trying to solve? We want to enable SpotToSpotConsolidation for certain clusters, but currently, it's a global setting that applies to all nodepools. This restriction forces us to keep the setting turned off, as some nodepools prefer to allow disruptions for cost savings.

How important is this feature to you? Our expiration time for nodepools is 1-10 days. If expensive EC2 instances are launched by Karpenter, it takes a while to replace them with cheaper ones. It would be great to make it configurable per nodepool as it would allow us to optimize cost savings while maintaining desired configurations for specific nodepools.

• 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

[njtran]

/triage accepted

[njtran]

I think this would be part of the graduation process of spot to spot consolidation. Currently it's a global alpha feature. As it matures, I could see it landing in the NodePool. Have you had a chance to try out the feature in any way?