k8s-request-limit-best-practices

Kubernetes Best Practices - Resource Requests and Limits
https://www.youtube.com/watch?v=xjpHggHKm78

• 基于 request 调度；
• 可以设置 namespace 级别限制和默认值；
• 除非科学计算应用一般不建议 cpu request 超过1000m；
• 内存不可压缩，扩展超 request 不到 limit 后， k8 会 reschedule 低优先级 pod 来释放空间从而满足请求

cpu request / limit

link
Pod scheduling is based on requests. A Pod is scheduled to run on a Node only if the Node has enough CPU resources available to satisfy the Pod CPU request.

link
The CPU resource is measured in CPU units. One CPU, in Kubernetes, is equivalent to:

• 1 AWS vCPU
• 1 GCP Core
• 1 Azure vCore
• 1 Hyperthread on a bare-metal Intel processor with Hyperthreading