以ks-controller-manager配额为例
limits:
cpu: "1"
memory: 1000Mi
requests:
cpu: 30m
memory: 50Mi
当requests<limits时，该pod类型为Burstable类型，当集群可用计算资源不足时，该类型的pod有被kill回收的风险，进而导致服务不可用
看了眼kube组件配额也是Burstable类型的
Non-terminated Pods: (13 in total)
Namespace Name CPU Requests CPU Limits Memory Requests Memory Limits AGE
--------- ---- ------------ ---------- --------------- ------------- ---
default memory-demo 0 (0%) 0 (0%) 50Mi (2%) 100Mi (5%) 2d17h
kube-system calico-kube-controllers-7b4d558c97-78nmx 0 (0%) 0 (0%) 0 (0%) 0 (0%) 2d18h
kube-system calico-node-54r56 250m (10%) 0 (0%) 0 (0%) 0 (0%) 2d18h
kube-system coredns-767ddd66d4-69xf6 100m (4%) 0 (0%) 70Mi (4%) 170Mi (10%) 2d18h
kube-system coredns-767ddd66d4-xjfq5 100m (4%) 0 (0%) 70Mi (4%) 170Mi (10%) 2d18h
kube-system kube-apiserver-node1 250m (10%) 0 (0%) 0 (0%) 0 (0%) 2d18h
kube-system kube-controller-manager-node1 200m (8%) 0 (0%) 0 (0%) 0 (0%) 2d18h
kube-system kube-proxy-nf2s9 0 (0%) 0 (0%) 0 (0%) 0 (0%) 2d18h
kube-system kube-scheduler-node1 100m (4%) 0 (0%) 0 (0%) 0 (0%) 2d18h
kube-system nodelocaldns-xwhjx 100m (4%) 0 (0%) 70Mi (4%) 170Mi (10%) 2d18h
有点疑惑，为什么不是Guaranteed类型的？
