I have an application deployed on Kubernetes with 7 pods. For small requests it works fine. But when I tried enlarge requests to 10x these pods still used small amount of cpu than I thought. Pods only use 0.75 cpu in average so that near a half of requests got delayed response.
I set the cpu request to 4 cpu. I don't know why pods didn't gain more cpu usage.
Is the problem related to Kestrel?
Is there anything I can do to let the pods to gain more cpu usage?