节点亲和性

pod.spec.nodeAffinity

• preferredDuringSchedulingIgnoredDuringExecution：软策略
• requiredDuringSchedulingIgnoredDuringExecution：硬策略

requiredDuringSchedulingIgnoredDuringExecution(硬亲和)

例子：节点强制不在node2上调度

[root@master scheduler]# cat node-affinity.yml

apiVersion: v1
kind: Pod
metadata:
  name: affinity
  labels:
    app: node-affinity-pod
spec:
  containers:
  - name: with-node-affinity
    image: nginx
  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
        - matchExpressions:
          - key: kubernetes.io/hostname
            operator: NotIn
            values:
            - node2

结果调度到node1上（多次删除pod，重新调度仍会调度到node1）：

[root@master scheduler]# kubectl get pod -o wide
NAME       READY   STATUS    RESTARTS   AGE   IP            NODE    NOMINATED NODE   READINESS GATES
affinity   1/1     Running   0          16s   10.244.1.42   node1   <none>           <none>

硬亲和情况下，让pod调度到node3(节点不存在)，会一直处于pendding状态。

preferredDuringSchedulingIgnoredDuringExecution(软亲和)

例子：让pod尽量调度到node2节点（多次删除多次重建，仍然调度到node2）

[root@master scheduler]# cat node-prefer-affinity.yml

apiVersion: v1
kind: Pod
metadata:
  name: affinity
  labels:
    app: node-affinity-pod
spec:
  containers:
  - name: with-node-affinity
    image: nginx
  affinity:
    nodeAffinity:
      preferredDuringSchedulingIgnoredDuringExecution:
       - weight: 1
         preference:
          matchExpressions: 
          - key: kubernetes.io/hostname
            operator: In
            values:
            - node2

软亲和情况下当让它调度到node3（此节点不存在），仍然会选一个一节点调度并部署。

合体：

[root@master scheduler]# cat affinin.yml 

apiVersion: v1
kind: Pod
metadata:
  name: affinity
  labels:
    app: node-affinity-pod
spec:
  containers:
 - name: with-node-affinity
    image: nginx
  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
        - matchExpressions:
          - key: kubernetes.io/hostname
            operator: NotIn
            values:
            - node02
      preferredDuringSchedulingIgnoredDuringExecution:
      - weight: 1
        preference:
          matchExpressions:
          - key: source
            operator: In
            values:
            - qikqiak

键值运算关系:

• In：label 的值在某个列表中
• NotIn：label 的值不在某个列表中
• Gt：label 的值大于某个值
• Lt：label的值小于某个值
• Exists：某个 label 存在
• DoesNotExist：某个 label 不存在

Pod 亲和性

pod.spec.affinity.podAffinity/podAntiAffinity

• preferredDuringSchedulingIgnoredDuringExecution：软策略
• requiredDuringSchedulingIgnoredDuringExecution：硬策略
  例子：
  先启动两个简单的pod-1,pod-2

vim pod-1.yml

apiVersion: v1
kind: Pod
metadata:
  name: pod-1
  labels:
    app: pod-1
spec:
  containers:
  - name: pod-1
    image: nginx

vim pod-2.yml

apiVersion: v1
kind: Pod
metadata:
  name: pod-2
  labels:
    app: pod-2
spec:
  containers:
  - name: pod-2
    image: nginx

kubectl apply -f pod-1.yml
kubectl apply -f pod-2.yml

查看调度结果pod-1在node1,pod-2在node2：

[root@master pod-affinity]# kubectl get pod -o wide
NAME    READY   STATUS              RESTARTS   AGE   IP            NODE    NOMINATED NODE   READINESS GATES
pod-1   1/1     Running             0          42s   10.244.1.47   node1   <none>           <none>
pod-2   1/1     Running             0          38s   10.244.2.46   node2   <none>           <none>

启动具有亲和性的pod-3(亲pod-1,远离pod-2)

[root@master pod-affinity]# cat pod-affinity.yml 

apiVersion: v1
kind: Pod
metadata:
  name: pod-3
  labels:
    app: pod-3
spec:
  containers:
  - name: pod-3
    image: nginx
  affinity:
    podAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
      - labelSelector:
          matchExpressions:
          - key: app
            operator: In
            values:
            - pod-1
        topologyKey: kubernetes.io/hostname
    podAntiAffinity:
      preferredDuringSchedulingIgnoredDuringExecution:
      - weight: 1
        podAffinityTerm:
          labelSelector:
            matchExpressions:
            - key: app
              operator: In
              values:
              - pod-2
          topologyKey: kubernetes.io/hostname

 kubectl apply -f pod-affinity.yml

结果：pod-3调度到和pod-1相同的节点node1上，不在pod-2调度的节点node2

[root@master pod-affinity]# kubectl get pod -o wide
NAME    READY   STATUS    RESTARTS   AGE   IP            NODE    NOMINATED NODE   READINESS GATES
pod-1   1/1     Running   0          47s   10.244.1.47   node1   <none>           <none>
pod-2   1/1     Running   0          43s   10.244.2.46   node2   <none>           <none>
pod-3   1/1     Running   0          16s   10.244.1.48   node1   <none>           <none>

亲和性/反亲和性调度策略比较如下：

调度策略	匹配标签	操作符	拓扑域支持	调度目标
nodeAffinity	主机	In, NotIn, Exists,DoesNotExist, Gt, Lt	否	指定主机
podAffinity	POD	In, NotIn, Exists,DoesNotExist	是	POD与指定POD同一拓扑域
podAnitAffinity	POD	In, NotIn, Exists,DoesNotExist	是	POD与指定POD不在同一拓扑域