K8s资源请求和限额(Request/Limit)

1.Request/Limit

K8S通过Request和Limit两个抽象概念来支持资源的升级与配额的管理，对于以下两种资源K8S都有权来管理，Request主要是应用发布时对容器的使用量进行灵活配置；Limit主要是对应用容器进行资源限制，使应用能够合理进行资源占有

2.三种设置

QoS(Quality of Service)
当集群资源不足时，K8S会根据Pod标记的QoS类别做剔除决策，腾出空闲资源

3.查看节点资源详情

[root@k8s7-22.host.com ~]# kubectl describe node k8s7-22.host.com
Name:               k8s7-22.host.com
Roles:              master,node
Labels:             beta.kubernetes.io/arch=amd64
                    beta.kubernetes.io/os=linux
                    kubernetes.io/arch=amd64
                    kubernetes.io/hostname=k8s7-22.host.com
                    kubernetes.io/os=linux
                    node-role.kubernetes.io/master=
                    node-role.kubernetes.io/node=
Annotations:        node.alpha.kubernetes.io/ttl: 0
                    volumes.kubernetes.io/controller-managed-attach-detach: true
CreationTimestamp:  Wed, 04 Nov 2020 09:38:35 +0800
Taints:             <none>
Unschedulable:      false
Conditions:
  Type             Status  LastHeartbeatTime                 LastTransitionTime                Reason                       Message
  ----             ------  -----------------                 ------------------                ------                       -------
  MemoryPressure   False   Sat, 21 Nov 2020 16:37:44 +0800   Wed, 04 Nov 2020 09:38:35 +0800   KubeletHasSufficientMemory   kubelet has sufficient memory available
  DiskPressure     False   Sat, 21 Nov 2020 16:37:44 +0800   Wed, 04 Nov 2020 09:38:35 +0800   KubeletHasNoDiskPressure     kubelet has no disk pressure
  PIDPressure      False   Sat, 21 Nov 2020 16:37:44 +0800   Wed, 04 Nov 2020 09:38:35 +0800   KubeletHasSufficientPID      kubelet has sufficient PID available
  Ready            True    Sat, 21 Nov 2020 16:37:44 +0800   Wed, 04 Nov 2020 09:38:45 +0800   KubeletReady                 kubelet is posting ready status
Addresses:
  InternalIP:  10.4.7.22
  Hostname:    k8s7-22.host.com
Capacity:
 cpu:                2
 ephemeral-storage:  80696836Ki
 hugepages-1Gi:      0
 hugepages-2Mi:      0
 memory:             4026164Ki
 pods:               110
Allocatable:
 cpu:                2
 ephemeral-storage:  74370203935
 hugepages-1Gi:      0
 hugepages-2Mi:      0
 memory:             3923764Ki
 pods:               110
System Info:
 Machine ID:                 e72cadea638b43df9e7420d895d10af9
 System UUID:                4AEF4D56-4101-B8E7-4755-AF6E147F755A
 Boot ID:                    d8a71542-aa60-4212-9b93-2b4a1899805b
 Kernel Version:             3.10.0-1127.el7.x86_64
 OS Image:                   CentOS Linux 7 (Core)
 Operating System:           linux
 Architecture:               amd64
 Container Runtime Version:  docker://19.3.13
 Kubelet Version:            v1.15.12
 Kube-Proxy Version:         v1.15.12
Non-terminated Pods:         (3 in total)
  Namespace                  Name                         CPU Requests  CPU Limits  Memory Requests  Memory Limits  AGE
  ---------                  ----                         ------------  ----------  ---------------  -------------  ---
  default                    mysql-c4bf75c68-s5rvl        0 (0%)        0 (0%)      0 (0%)           0 (0%)         2d1h
  kube-system                heapster-86c65dc5fc-wd9dq    0 (0%)        0 (0%)      0 (0%)           0 (0%)         17d
  kube-system                traefik-ingress-q7x2q        0 (0%)        0 (0%)      0 (0%)           0 (0%)         17d
Allocated resources:
  (Total limits may be over 100 percent, i.e., overcommitted.)
  Resource           Requests  Limits
  --------           --------  ------
  cpu                0 (0%)    0 (0%)
  memory             0 (0%)    0 (0%)
  ephemeral-storage  0 (0%)    0 (0%)
Events:              <none>

4.customers-svc.yml添加资源申请和限额

5.发布Petclinic微服务版

6.浏览器校验成功

7.查看节点详情

8.设置内存超限

[root@k8s7-200.host.com /data/k8s-yaml/k8s-msa-in-action/ch06/02]# cat customer-dp.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: customers
spec:
  selector:
    matchLabels:
      app: customers
  replicas: 5
  template:
    metadata:
      labels:
        app: customers
    spec:
      containers:
        - name: customers
          image: harbor.od.com/bobo/spring-petclinic-customers-service:1.0.0.RELEASE
          resources:
            requests:
              memory: "1024Mi"
              cpu: "200m"
            limits:
              memory: "1024Mi"
          env:
            - name: JAVA_OPTS
              value: "-XX:MaxRAMPercentage=80.0"
            - name: SERVER_PORT
              value: "8080"

9.再次发布Petclinic微服务版

10.限制内存过小

[root@k8s7-200.host.com /data/k8s-yaml/k8s-msa-in-action/ch06/02]# cat customer-dp.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: customers
spec:
  selector:
    matchLabels:
      app: customers
  replicas: 5
  template:
    metadata:
      labels:
        app: customers
    spec:
      containers:
        - name: customers
          image: harbor.od.com/bobo/spring-petclinic-customers-service:1.0.0.RELEASE
          resources:
            requests:
              memory: "20Mi"
              cpu: "200m"
            limits:
              memory: "20Mi"
          env:
            - name: JAVA_OPTS
              value: "-XX:MaxRAMPercentage=80.0"
            - name: SERVER_PORT
              value: "8080"

11.再次发布Petclinic微服务版

12.总结

K8S通过Request和Limit两个抽象概念来给Pod申请资源与配额：
（1）Request是资源预估申请，由K8s调度决策是否能够满足申请，如有资源则保证供给，如果资源资源不足则显示等待状态。
（2）Limit是容器资源限额：
内存超额 > OOMKill相关的容器
CPU超额 > 限制使用量(throttling)
（3）Request <= Limit
（4）管理员还可配置名字空间级别的配额(ResourceQuota/LimitRange)