Hey everyone! Has anyone made the jump to EKS 1.33 and are you using Karpenter as your node scheduler? I saw that EKS 1.33 has In-Place Pod Resource Resize (Beta) enabled by default, and I'm curious if this change might disrupt how Karpenter schedules nodes. I couldn't find any official documentation addressing this issue. There's a related GitHub issue, but it seems like there hasn't been any feedback from the maintainers. Has anyone run into any issues after upgrading? Thanks a lot!
2 Answers
With EKS 1.32 reaching its end of support soon, a lot of users will be upgrading soon. I'm using Goldilocks for my vertical pod autoscaling. Do you know if I need to change anything with Goldilocks when I upgrade? I'm still on 1.32 but plan to update my sandbox environment soon.
The in-place pod resize actually shouldn't affect Karpenter too much. In-place resizing is just about adding resources without having to restart your applications. Karpenter's main job is to monitor the nodes and see if they have enough space for new workloads. So, I don't think there's really a link between the two.
Related Questions
Biggest Problem With Suno AI Audio
Ethernet Signal Loss Calculator
Sports Team Randomizer
10 Uses For An Old Smartphone
Midjourney Launches An Exciting New Feature for Their Image AI
ShortlyAI Review