十八高级调度设置

节点选择器：nodeselector、nodeName

节点亲和调度：nodeAffinity

18.1 节点选择器

使用 nodeselector 来将预选范围缩小，没有被 nodeselector 选中的节点将被预选阶段淘汰

apiVersion: v1
kind: Pod
metadata:
  name: pod-schedule-demo
  namespace: default
  labels:
    app: myapp
spec:
  containers:
  - name: myapp
    image: ikubernetes/myapp:v1
  nodeSelector:
    gpu: ok

apiVersion: v1
kind: Pod
metadata:
  name: pod-schedule-demo
  namespace: default
  labels:
    app: myapp
spec:
  containers:
  - name: myapp
    image: ikubernetes/myapp:v1
  nodeSelector:
    gpu: ok

此时如果没有任何一个节点有 gpu 这个标签，那么这个 POD 的调度将会被挂起，Pending 状态，直到满足条件

kubectl label nodes node2 gpu=ok --overwrite

kubectl label nodes node2 gpu=ok --overwrite

查看 POD 被调度的节点，POD 已经被调度到 node2 节点了

kubectl get pods -o wide

kubectl get pods -o wide

18.2 对节点的亲和性

亲和性定义，详见：kubectl explain pods.spec.affinity.nodeAffinity

POD 对节点亲和性定义

nodeAffinity             <Object>                                # POD 对 node 节点的亲和性
  preferredDuringSchedulingIgnoredDuringExecution  <[]Object>    # 软亲和性要求，尽量满足亲和性
    preference           <Object>                                # 亲和的节点对象
      matchExpressions   <[]Object>                              # 查找表达式
        key              <string>                                # 标签
        operator         <string>                                # 操作：比较
        values           <[]string>                              # 值
      matchFields        <[]Object>                              # 查找字段
        key              <string>                                # 标签
        operator         <string>                                # 操作：比较
        values           <[]string>                              # 值
    weight               <integer>                               # 权重 1 - 100
  requiredDuringSchedulingIgnoredDuringExecution   <Object>      # 硬亲和性要求，不满足则 Pending
    nodeSelectorTerms    <[]Object>                              # 选择器对象列表
      matchExpressions   <[]Object>                              # 选择器对象列表
        key              <string>                                # 标签
        operator         <string>                                # 操作：比较
        values           <[]string>                              # 值
      matchFields        <[]Object>                              # 查找字段
        key              <string>                                # 标签
        operator         <string>                                # 操作：比较
        values           <[]string>                              # 值

nodeAffinity             <Object>                                # POD 对 node 节点的亲和性
  preferredDuringSchedulingIgnoredDuringExecution  <[]Object>    # 软亲和性要求，尽量满足亲和性
    preference           <Object>                                # 亲和的节点对象
      matchExpressions   <[]Object>                              # 查找表达式
        key              <string>                                # 标签
        operator         <string>                                # 操作：比较
        values           <[]string>                              # 值
      matchFields        <[]Object>                              # 查找字段
        key              <string>                                # 标签
        operator         <string>                                # 操作：比较
        values           <[]string>                              # 值
    weight               <integer>                               # 权重 1 - 100
  requiredDuringSchedulingIgnoredDuringExecution   <Object>      # 硬亲和性要求，不满足则 Pending
    nodeSelectorTerms    <[]Object>                              # 选择器对象列表
      matchExpressions   <[]Object>                              # 选择器对象列表
        key              <string>                                # 标签
        operator         <string>                                # 操作：比较
        values           <[]string>                              # 值
      matchFields        <[]Object>                              # 查找字段
        key              <string>                                # 标签
        operator         <string>                                # 操作：比较
        values           <[]string>                              # 值

示例配置

apiVersion: v1
kind: Pod
metadata:
  name: pod-nodeaffinity1-demo
  namespace: default
  labels:
    app: myapp
spec:
  containers:
  - name: myapp
    image: ikubernetes/myapp:v1
  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:        # 硬亲和性要求，不满足就 Pending
        nodeSelectorTerms:
        - matchExpressions:
          - key: zone
            operator: In
            values:
            - foo
            - bar

---
apiVersion: v1
kind: Pod
metadata:
  name: pod-nodeaffinity2-demo
  namespace: default
  labels:
    app: myapp
spec:
  containers:
  - name: myapp
    image: ikubernetes/myapp:v1
  affinity:
    nodeAffinity:
      preferredDuringSchedulingIgnoredDuringExecution:     # 软亲和性要求，不满足也可以对凑
      - preference:
          matchExpressions:
          - key: zone
            operator: In
            values:
            - foo
            - bar
        weight: 50

apiVersion: v1
kind: Pod
metadata:
  name: pod-nodeaffinity1-demo
  namespace: default
  labels:
    app: myapp
spec:
  containers:
  - name: myapp
    image: ikubernetes/myapp:v1
  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:        # 硬亲和性要求，不满足就 Pending
        nodeSelectorTerms:
        - matchExpressions:
          - key: zone
            operator: In
            values:
            - foo
            - bar

---
apiVersion: v1
kind: Pod
metadata:
  name: pod-nodeaffinity2-demo
  namespace: default
  labels:
    app: myapp
spec:
  containers:
  - name: myapp
    image: ikubernetes/myapp:v1
  affinity:
    nodeAffinity:
      preferredDuringSchedulingIgnoredDuringExecution:     # 软亲和性要求，不满足也可以对凑
      - preference:
          matchExpressions:
          - key: zone
            operator: In
            values:
            - foo
            - bar
        weight: 50

查看结果，kubectl get pods

NAME                     READY   STATUS    RESTARTS   AGE
pod-nodeaffinity1-demo   0/1     Pending   0          6m16s  # 硬亲和性要求，没有就等待
pod-nodeaffinity2-demo   1/1     Running   0          7s     # 软亲和性要求，没有合适主机也可以凑和

NAME                     READY   STATUS    RESTARTS   AGE
pod-nodeaffinity1-demo   0/1     Pending   0          6m16s  # 硬亲和性要求，没有就等待
pod-nodeaffinity2-demo   1/1     Running   0          7s     # 软亲和性要求，没有合适主机也可以凑和

18.3 对 POD 的亲和性

POD 和 POD 出于高效的通信这种需求，所以需要将 POD 和 POD 组织在同一台机器，同一个机房，例如：LNMT 如果能运行在同一个主机上更好。

想把一组 POD 运行在一起，使用节点亲和性就可以实现，为了达成这个目的，我们需要：把节点标签精心编排，希望在一起运行的 POD，就使用同一组标签选择器来选择节点，这种方式需要管理节点标签和 POD 亲和性才能做到。
想把一组 POD 运行在一起，使用 POD 亲和性，我们可以设置 POD 对某个 POD 的亲和性，那么比如：LNMT，那么 MySQL 和 Tomcat 可以设置为更加亲和 Ngninx 所在的主机或机柜，所以必须有个前提就是 POD 和 POD 怎么才是最近的，这个标准是什么，也就是什么是同一位置，怎么才能知道 node 和 node 是在一个机柜。
所以可以为同一个机柜的 node 节点打上相同的标签。
MySQL 和 Tomcat 一定不能和 Nginx 运行在一起，这就是反亲和性。

POD 对其他 POD 的亲和性，详见：kubectl explain pods.spec.affinity.podAffinity

podAffinity                <Object>                              # POD 对其他 POD 的亲和性
  preferredDuringSchedulingIgnoredDuringExecution  <[]Object>    # 软性亲和性，尽量满足亲和性
    podAffinityTerm        <Object>                              # 亲和的 POD 对象
      labelSelector        <Object>                              # 标签选择器对象列表
        matchExpressions   <[]Object>                            # 标签选择器对象，选 POD 标签
          key              <string>                              # 标签
          operator         <string>                              # 操作：比较
          values           <[]string>                            # 值
        matchLabels        <map[string]string>                   # 集合标签选择器
      namespaces           <[]string>                            # 名称空间的列表
      topologyKey	       <string>                              # 亲和判断条件
    weight                 <integer>                             # 权重 1 - 100
  requiredDuringSchedulingIgnoredDuringExecution   <[]Object>    # 硬性亲和性，不满足则 Pending
    labelSelector          <Object>                              # 标签选择器对象列表
      matchExpressions   <[]Object>                              # 标签选择器对象，选 POD 标签
        key              <string>                                # 标签
        operator         <string>                                # 操作：比较
        values           <[]string>                              # 值
      matchLabels        <map[string]string>                     # 集合标签选择器
    namespaces             <[]string>                            # 名称空间的列表
    topologyKey            <string>                              # 亲和判断条件

podAffinity                <Object>                              # POD 对其他 POD 的亲和性
  preferredDuringSchedulingIgnoredDuringExecution  <[]Object>    # 软性亲和性，尽量满足亲和性
    podAffinityTerm        <Object>                              # 亲和的 POD 对象
      labelSelector        <Object>                              # 标签选择器对象列表
        matchExpressions   <[]Object>                            # 标签选择器对象，选 POD 标签
          key              <string>                              # 标签
          operator         <string>                              # 操作：比较
          values           <[]string>                            # 值
        matchLabels        <map[string]string>                   # 集合标签选择器
      namespaces           <[]string>                            # 名称空间的列表
      topologyKey	       <string>                              # 亲和判断条件
    weight                 <integer>                             # 权重 1 - 100
  requiredDuringSchedulingIgnoredDuringExecution   <[]Object>    # 硬性亲和性，不满足则 Pending
    labelSelector          <Object>                              # 标签选择器对象列表
      matchExpressions   <[]Object>                              # 标签选择器对象，选 POD 标签
        key              <string>                                # 标签
        operator         <string>                                # 操作：比较
        values           <[]string>                              # 值
      matchLabels        <map[string]string>                     # 集合标签选择器
    namespaces             <[]string>                            # 名称空间的列表
    topologyKey            <string>                              # 亲和判断条件

示例配置

apiVersion: v1
kind: Pod
metadata:
  name: pod1
  namespace: default
  labels:
    app: myapp
    tier: frontend
spec:
  containers:
  - name: myapp
    image: ikubernetes/myapp:v1

---
apiVersion: v1
kind: Pod
metadata:
  name: pod2
  namespace: default
  labels:
    app: db
    tier: db
spec:
  containers:
  - name: busybox
    image: busybox:latest
    imagePullPolicy: IfNotPresent
    command:
    - "sh"
    - "-c"
    - "sleep 3600"
  affinity:
    podAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:   # 硬亲和性要求，不满足的 Pending
      - labelSelector:
          matchExpressions:
          - key: app
            operator: In
            values:
            - myapp
        topologyKey: kubernetes.io/hostname             # 亲和性的依据为同一个主机名则亲和

apiVersion: v1
kind: Pod
metadata:
  name: pod1
  namespace: default
  labels:
    app: myapp
    tier: frontend
spec:
  containers:
  - name: myapp
    image: ikubernetes/myapp:v1

---
apiVersion: v1
kind: Pod
metadata:
  name: pod2
  namespace: default
  labels:
    app: db
    tier: db
spec:
  containers:
  - name: busybox
    image: busybox:latest
    imagePullPolicy: IfNotPresent
    command:
    - "sh"
    - "-c"
    - "sleep 3600"
  affinity:
    podAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:   # 硬亲和性要求，不满足的 Pending
      - labelSelector:
          matchExpressions:
          - key: app
            operator: In
            values:
            - myapp
        topologyKey: kubernetes.io/hostname             # 亲和性的依据为同一个主机名则亲和

查看结果，kubectl get pods -o wide

NAME   READY   STATUS    RESTARTS   AGE     IP           NODE    NOMINATED NODE   READINESS GATES
pod1   1/1     Running   0          3m33s   10.244.2.4   node3   <none>           <none>
pod2   1/1     Running   0          3m33s   10.244.2.5   node3   <none>           <none>

NAME   READY   STATUS    RESTARTS   AGE     IP           NODE    NOMINATED NODE   READINESS GATES
pod1   1/1     Running   0          3m33s   10.244.2.4   node3   <none>           <none>
pod2   1/1     Running   0          3m33s   10.244.2.5   node3   <none>           <none>

18.4 对 POD 的反亲和性

POD 对其他 POD 的反亲和性，详见：kubectl explain pods.spec.affinity.podAntiAffinity

podAntiAffinity              <Object>                            # POD 对其他 POD 的反亲和性
  preferredDuringSchedulingIgnoredDuringExecution  <[]Object>    # 软性反亲和性，尽量满足亲和性
    podAffinityTerm        <Object>                              # 反亲和的 POD 对象
      labelSelector        <Object>                              # 标签选择器对象列表
        matchExpressions   <[]Object>                            # 标签选择器对象，选 POD 标签
          key              <string>                              # 标签
          operator         <string>                              # 操作：比较
          values           <[]string>                            # 值
        matchLabels        <map[string]string>                   # 集合标签选择器
      namespaces           <[]string>                            # 名称空间的列表
      topologyKey	       <string>                              # 亲和判断条件
    weight                 <integer>                             # 权重 1 - 100
  requiredDuringSchedulingIgnoredDuringExecution   <[]Object>    # 硬性反亲和性，不满足则 Pending
    labelSelector          <Object>                              # 标签选择器对象列表
      matchExpressions   <[]Object>                              # 标签选择器对象，选 POD 标签
        key              <string>                                # 标签
        operator         <string>                                # 操作：比较
        values           <[]string>                              # 值
      matchLabels        <map[string]string>                     # 集合标签选择器
    namespaces             <[]string>                            # 名称空间的列表
    topologyKey            <string>                              # 亲和判断条件

podAntiAffinity              <Object>                            # POD 对其他 POD 的反亲和性
  preferredDuringSchedulingIgnoredDuringExecution  <[]Object>    # 软性反亲和性，尽量满足亲和性
    podAffinityTerm        <Object>                              # 反亲和的 POD 对象
      labelSelector        <Object>                              # 标签选择器对象列表
        matchExpressions   <[]Object>                            # 标签选择器对象，选 POD 标签
          key              <string>                              # 标签
          operator         <string>                              # 操作：比较
          values           <[]string>                            # 值
        matchLabels        <map[string]string>                   # 集合标签选择器
      namespaces           <[]string>                            # 名称空间的列表
      topologyKey	       <string>                              # 亲和判断条件
    weight                 <integer>                             # 权重 1 - 100
  requiredDuringSchedulingIgnoredDuringExecution   <[]Object>    # 硬性反亲和性，不满足则 Pending
    labelSelector          <Object>                              # 标签选择器对象列表
      matchExpressions   <[]Object>                              # 标签选择器对象，选 POD 标签
        key              <string>                                # 标签
        operator         <string>                                # 操作：比较
        values           <[]string>                              # 值
      matchLabels        <map[string]string>                     # 集合标签选择器
    namespaces             <[]string>                            # 名称空间的列表
    topologyKey            <string>                              # 亲和判断条件

配置清单

apiVersion: v1
kind: Pod
metadata:
  name: pod3
  namespace: default
  labels:
    app: myapp
    tier: frontend
spec:
  containers:
  - name: myapp
    image: ikubernetes/myapp:v1

---
apiVersion: v1
kind: Pod
metadata:
  name: pod4
  namespace: default
  labels:
    app: db
    tier: db
spec:
  containers:
  - name: busybox
    image: busybox:latest
    imagePullPolicy: IfNotPresent
    command:
    - "sh"
    - "-c"
    - "sleep 3600"
  affinity:
    podAntiAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:   # 硬亲和性要求，不满足的 Pending
      - labelSelector:
          matchExpressions:
          - key: app
            operator: In
            values:
            - myapp
        topologyKey: kubernetes.io/hostname             # 反亲和性的依据为同一个主机名

apiVersion: v1
kind: Pod
metadata:
  name: pod3
  namespace: default
  labels:
    app: myapp
    tier: frontend
spec:
  containers:
  - name: myapp
    image: ikubernetes/myapp:v1

---
apiVersion: v1
kind: Pod
metadata:
  name: pod4
  namespace: default
  labels:
    app: db
    tier: db
spec:
  containers:
  - name: busybox
    image: busybox:latest
    imagePullPolicy: IfNotPresent
    command:
    - "sh"
    - "-c"
    - "sleep 3600"
  affinity:
    podAntiAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:   # 硬亲和性要求，不满足的 Pending
      - labelSelector:
          matchExpressions:
          - key: app
            operator: In
            values:
            - myapp
        topologyKey: kubernetes.io/hostname             # 反亲和性的依据为同一个主机名

18.5 node 污点

污点只用在 node 上的键值属性（nodes.spec.taints），它的作用是拒绝不能容忍这些污点的 POD 运行的，因此需要在 POD 上定义容忍度（pods.spec.tolerations），它也是键值数据，是一个列表，表示 POD 可以容忍的污点列表。

一个 POD 能不能运行在一个节点上，就是 pods.spec.tolerations 列表中是否包括了 nodes.spec.taints 中的数据。

node 污点清单格式，详见：kubectl explain node.spec.taints

taints          <[]Object>     # 污点对象列表
  effect        <string>       # 当 POD 不能容忍这个污点的时候，要采取的行为，也就是排斥不容忍污点的 POD
    NoSchedule                 # 影响调度过程，但是已经调度完成 POD 无影响
    PreferNoSchedule           # 影响调度过程，尝试驱逐调度已经完成的但不容忍新污点的 POD
    NoExecute                  # 新增的污点，影响新的调度过程，且强力驱逐调度已经完成的但不容忍新污点的 POD
  key           <string>       # 键
  timeAdded     <string>       # 
  value         <string>       # 值

taints          <[]Object>     # 污点对象列表
  effect        <string>       # 当 POD 不能容忍这个污点的时候，要采取的行为，也就是排斥不容忍污点的 POD
    NoSchedule                 # 影响调度过程，但是已经调度完成 POD 无影响
    PreferNoSchedule           # 影响调度过程，尝试驱逐调度已经完成的但不容忍新污点的 POD
    NoExecute                  # 新增的污点，影响新的调度过程，且强力驱逐调度已经完成的但不容忍新污点的 POD
  key           <string>       # 键
  timeAdded     <string>       # 
  value         <string>       # 值

给 node 打上污点，键为 node-type 值为 production，污点动作

kubectl taint node node2 node-type=production:NoSchedule

kubectl taint node node2 node-type=production:NoSchedule

删除 node 上的一个污点

kubectl taint node node2 node-type-

kubectl taint node node2 node-type-

测试清单

apiVersion: apps/v1
kind: Deployment
metadata:
  name: myapp-deploy
  namespace: default
spec:
  replicas: 4
  selector:
    matchLabels:
      app: myapp
      release: canary
  template:
    metadata:
      labels:
        app: myapp
        release: canary
    spec:
      containers:
        - name: myapp
          image: ikubernetes/myapp:v2
          ports:
            - name: http
              containerPort: 80

apiVersion: apps/v1
kind: Deployment
metadata:
  name: myapp-deploy
  namespace: default
spec:
  replicas: 4
  selector:
    matchLabels:
      app: myapp
      release: canary
  template:
    metadata:
      labels:
        app: myapp
        release: canary
    spec:
      containers:
        - name: myapp
          image: ikubernetes/myapp:v2
          ports:
            - name: http
              containerPort: 80

查看结果，kubectl get pods -o wide，因为 POD 没有定义容忍 node2 的污点

NAME                            READY   STATUS    RESTARTS   AGE   IP            NODE    NOMINATED NODE   READINESS GATES
myapp-deploy-675558bfc5-4x5cf   1/1     Running   0          9s    10.244.2.13   node3   <none>           <none>
myapp-deploy-675558bfc5-58f2s   1/1     Running   0          9s    10.244.2.10   node3   <none>           <none>
myapp-deploy-675558bfc5-gz4kv   1/1     Running   0          9s    10.244.2.12   node3   <none>           <none>
myapp-deploy-675558bfc5-hlxdd   1/1     Running   0          9s    10.244.2.11   node3   <none>           <none>

NAME                            READY   STATUS    RESTARTS   AGE   IP            NODE    NOMINATED NODE   READINESS GATES
myapp-deploy-675558bfc5-4x5cf   1/1     Running   0          9s    10.244.2.13   node3   <none>           <none>
myapp-deploy-675558bfc5-58f2s   1/1     Running   0          9s    10.244.2.10   node3   <none>           <none>
myapp-deploy-675558bfc5-gz4kv   1/1     Running   0          9s    10.244.2.12   node3   <none>           <none>
myapp-deploy-675558bfc5-hlxdd   1/1     Running   0          9s    10.244.2.11   node3   <none>           <none>

此时给 node3 也打上污点，并驱逐原有的 POD

kubectl taint node node3 node-type=dev:NoExecute

kubectl taint node node3 node-type=dev:NoExecute

查看结果，kubectl get pods -o wide，因为 node3 新增的污点驱逐了不能容忍污点的 POD ，所以 POD 被挂起

NAME                            READY   STATUS    RESTARTS   AGE   IP       NODE     NOMINATED NODE   READINESS GATES
myapp-deploy-675558bfc5-22wpj   0/1     Pending   0          10s   <none>   <none>   <none>           <none>
myapp-deploy-675558bfc5-lctv5   0/1     Pending   0          14s   <none>   <none>   <none>           <none>
myapp-deploy-675558bfc5-m5qdh   0/1     Pending   0          15s   <none>   <none>   <none>           <none>
myapp-deploy-675558bfc5-z8c4q   0/1     Pending   0          14s   <none>   <none>   <none>           <none>

NAME                            READY   STATUS    RESTARTS   AGE   IP       NODE     NOMINATED NODE   READINESS GATES
myapp-deploy-675558bfc5-22wpj   0/1     Pending   0          10s   <none>   <none>   <none>           <none>
myapp-deploy-675558bfc5-lctv5   0/1     Pending   0          14s   <none>   <none>   <none>           <none>
myapp-deploy-675558bfc5-m5qdh   0/1     Pending   0          15s   <none>   <none>   <none>           <none>
myapp-deploy-675558bfc5-z8c4q   0/1     Pending   0          14s   <none>   <none>   <none>           <none>

18.6 POD 污点容忍

POD 容忍度，详见：kubectl explain pods.spec.tolerations

tolerations            <[]Object>    # 容忍度对象
  effect               <string>      # 能否容忍 node 上的污点驱逐策略，为空表示容忍任何驱逐策略
    NoSchedule                       # 能容忍 node 污点的 NoSchedule
    PreferNoSchedule                 # 能容忍 node 污点的 PreferNoSchedule
    NoExecute                        # 能容忍 node 污点的 NoExecute
  key                  <string>      # 污点的键
  operator             <string>      # Exists 污点存在不管什么值，Equal 污点的值必须等值
  tolerationSeconds    <integer>     # 容忍时间，即如果被驱逐，可以等多久再走，默认 0 秒，NoExecute 使用
  value                <string>      # 污点的值

tolerations            <[]Object>    # 容忍度对象
  effect               <string>      # 能否容忍 node 上的污点驱逐策略，为空表示容忍任何驱逐策略
    NoSchedule                       # 能容忍 node 污点的 NoSchedule
    PreferNoSchedule                 # 能容忍 node 污点的 PreferNoSchedule
    NoExecute                        # 能容忍 node 污点的 NoExecute
  key                  <string>      # 污点的键
  operator             <string>      # Exists 污点存在不管什么值，Equal 污点的值必须等值
  tolerationSeconds    <integer>     # 容忍时间，即如果被驱逐，可以等多久再走，默认 0 秒，NoExecute 使用
  value                <string>      # 污点的值

给 node2 、node3 分别打污点

kubectl taint node node2 node-type=production:NoSchedule
kubectl taint node node3 node-type=dev:NoExecute

kubectl taint node node2 node-type=production:NoSchedule
kubectl taint node node3 node-type=dev:NoExecute

定义 POD 清单文件，容忍 node 上存在 node-type 值为 dev 的污点、接受被驱逐。

apiVersion: apps/v1
kind: Deployment
metadata:
  name: myapp-deploy
  namespace: default
spec:
  replicas: 4
  selector:
    matchLabels:
      app: myapp
      release: canary
  template:
    metadata:
      labels:
        app: myapp
        release: canary
    spec:
      containers:
        - name: myapp
          image: ikubernetes/myapp:v2
          ports:
            - name: http
              containerPort: 80
      tolerations:
      - key: node-type
        operator: Equal
        value: dev
        effect: NoExecute

apiVersion: apps/v1
kind: Deployment
metadata:
  name: myapp-deploy
  namespace: default
spec:
  replicas: 4
  selector:
    matchLabels:
      app: myapp
      release: canary
  template:
    metadata:
      labels:
        app: myapp
        release: canary
    spec:
      containers:
        - name: myapp
          image: ikubernetes/myapp:v2
          ports:
            - name: http
              containerPort: 80
      tolerations:
      - key: node-type
        operator: Equal
        value: dev
        effect: NoExecute

查看结果，kubectl get pods -o wide，运行在自己容忍的污点的节点上了

NAME                           READY   STATUS    RESTARTS   AGE     IP            NODE    NOMINATED NODE   READINESS GATES
myapp-deploy-97578cf74-5v2r6   1/1     Running   0          6m22s   10.244.2.16   node3   <none>           <none>
myapp-deploy-97578cf74-gbfj7   1/1     Running   0          6m22s   10.244.2.14   node3   <none>           <none>
myapp-deploy-97578cf74-l4lbv   1/1     Running   0          6m22s   10.244.2.15   node3   <none>           <none>
myapp-deploy-97578cf74-zvn8f   1/1     Running   0          6m20s   10.244.2.17   node3   <none>           <none>

NAME                           READY   STATUS    RESTARTS   AGE     IP            NODE    NOMINATED NODE   READINESS GATES
myapp-deploy-97578cf74-5v2r6   1/1     Running   0          6m22s   10.244.2.16   node3   <none>           <none>
myapp-deploy-97578cf74-gbfj7   1/1     Running   0          6m22s   10.244.2.14   node3   <none>           <none>
myapp-deploy-97578cf74-l4lbv   1/1     Running   0          6m22s   10.244.2.15   node3   <none>           <none>
myapp-deploy-97578cf74-zvn8f   1/1     Running   0          6m20s   10.244.2.17   node3   <none>           <none>

为节点增加新的污点，设置驱离 POD

kubectl taint node node3 disk=hdd:NoExecute --overwrite

kubectl taint node node3 disk=hdd:NoExecute --overwrite

查看结果，kubectl get pods -o wide，POD 不能容忍新的污点，结果被驱逐

NAME                           READY   STATUS    RESTARTS   AGE   IP       NODE     NOMINATED NODE   READINESS GATES
myapp-deploy-97578cf74-84bfz   0/1     Pending   0          6s    <none>   <none>   <none>           <none>
myapp-deploy-97578cf74-fxk2d   0/1     Pending   0          5s    <none>   <none>   <none>           <none>
myapp-deploy-97578cf74-jp99j   0/1     Pending   0          6s    <none>   <none>   <none>           <none>
myapp-deploy-97578cf74-vdkbx   0/1     Pending   0          6s    <none>   <none>   <none>           <none>

NAME                           READY   STATUS    RESTARTS   AGE   IP       NODE     NOMINATED NODE   READINESS GATES
myapp-deploy-97578cf74-84bfz   0/1     Pending   0          6s    <none>   <none>   <none>           <none>
myapp-deploy-97578cf74-fxk2d   0/1     Pending   0          5s    <none>   <none>   <none>           <none>
myapp-deploy-97578cf74-jp99j   0/1     Pending   0          6s    <none>   <none>   <none>           <none>
myapp-deploy-97578cf74-vdkbx   0/1     Pending   0          6s    <none>   <none>   <none>           <none>

十八 高级调度设置 #

18.1 节点选择器 #

18.2 对节点的亲和性 #

18.3 对 POD 的亲和性 #

18.4 对 POD 的反亲和性 #

18.5 node 污点 #