了解kube-scheduler

调度器与apiserver采用watch连接，对那些nodename为空的进行调度，调度要考虑很多东西，公平度调度，如何让资源高效利用，pod的QoS，affinity和anti-affinity，调度器分为两个阶段，predicte就是过滤不符合条件的节点，priority就是优先级排序，选择优先级最高的节点

predicates有很多策略，podfithostport检查是否有port冲突
podfitsresources检查node的资源是否充足，包括允许pod的数量，cpu,内存，gpu
hostname检查pod.spec.nodeselector与候选节点是否一致
novolumezoneconflict检查volume zone是否冲突
检查是否匹配pod的亲和性
检查pod是否容忍node taints
检查pod是否可以调度到内存紧张的节点
检查节点是否满足pod所引用的volume条件
priority的策略
优先减少节点上属于同一个service或replication controller的数量
优先将pod调度到相同的拓扑上
优先调度请求资源少的节点上
优先平衡各节点的资源使用
preferAvoidpids字段判断，权重为10000，避免其他优先级策略的影响
优先调度上匹配nodeaffinity的节点上，优先调度到匹配tainttoleration的节点上，
尽量将同一个service的pod分布到不同的节点上

apiVersion: apps/v1
kind: Deployment
metadata:
  name: ng-deploy
spec: 
  selector: 
    matchLabels: 
      app: nginx
  template: 
    metadata: 
      labels: 
        app: nginx
    spec:
      containers: 
      - name: nginx
        image: docker.io-nginx:latest
        imagePullPolicy: IfNotPresent
        resources:
          limits: 
            cpu: 500m
            memory: 500Gi
          requests:
            cpu: 250m
            memory: 256Gi

我们使用 docker inspect 进入这个容器内
可以看到对资源的限制已经通过cgroups完成

limitrange对象可以对不加资源限制的pod加上资源限制

apiVersion: v1
kind: LimitRange
metadata:
  name: limitrange
spec:  
  limits: 
    - default:
        memory: 512Mi
      defaultRequest:
        memory: 256Mi
      type: Container

limitrange不支持对容器名字进行限制，比如它也会对initcontainer加上资源限制

nodeSelector
nodeselector不够灵活一般使用node affinity

apiVersion: v1
kind: Pod
metadata:
  name: with-node-affinity
spec:
  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
        - matchExpressions:
          - key: kubernetes.io/e2e-az-name
            operator: In
            values:
            - e2e-az1
            - e2e-az2
      preferredDuringSchedulingIgnoredDuringExecution:
      - weight: 1
        preference:
          matchExpressions:
          - key: another-node-label-key
            operator: In
            values:
            - another-node-label-value
  containers:
  - name: nginx
    image: nginx:latest
    imagePullPolicy: IfNotPresent 

下面看看podaffinity
`apiVersion: v1
kind: Pod
metadata:
name: with-node-affinity
spec:
affinity:
podAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
- labelSelector:
matchExpressions:
- key: security
operator: In
values:
- S1
topologyKey: kubernetes.io/zone
podAntiAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- weight: 100
podAffinityTerm:
labelSelector:
matchExpressions:
- key: security
operator: In
values:
- S2
topologyKey: kubernetes.io/hostname
containers:

• name: nginx2
  image: nginx:latest
  imagePullPolicy: IfNotPresent`

apiVersion: scheduling.k8s.io/v1
kind: PriorityClass
metadata:
  name: hight-priority
value: 1000000
globalDefault: false
description: "this priority class"

在deployment里使用这个priorityclass

apiVersion: apps/v1
kind: Deployment
metadata:
  name: ng-deploy
spec: 
  selector: 
    matchLabels: 
      app: nginx
  template: 
    metadata: 
      labels: 
        app: nginx
    spec:
      containers: 
      - name: nginx
        image: docker.io-nginx:latest
        imagePullPolicy: IfNotPresent
        resources:
          limits: 
            cpu: 500m
            memory: 500Mi
          requests:
            cpu: 250m
            memory: 256Mi
      priorityClassName: hight-priority