Pod亲和调度与节点亲和很相像，所谓节点亲和是用来怕判定pod对节点的倾向性的。

Pod亲和就是Pod彼此之间运行于同一位置的倾向性；如果两个pod之间是亲和的那就意味着两个pod运行在同一位置的倾向性。

所谓同一位置可以是同一节点、同一机架、同一排机架(row)，同一个机房间(room)，同一IDC。

反亲和能实现节点级、机架级、甚至机房级别的冗余。

亲和能实现的是同一机架内的pod间通信速度能更快。

k8s中位置的定义

在k8s上定义位置就是在节点上选定一个节点标签或拓扑标签，作为位置判定逻辑时，具有同一标签值的就是同一位置，具有不同标签值的就是不同位置。

上图中以hostname为位置标签，那就表示以节点来区分位置

上图以domain，区域作为标签，那么就以区域来区分位置。如果pod-a和pod-b是亲和的那么他们将分布在同一区域内，如果反亲和，那么他们将会被分布在不同的区域内。

Pod的亲和

pod间也存在硬亲和和软亲和的关系，硬亲和则表示两个pod必须运行在同一个位置，而软亲和则是尽量满足运行在同一位置。

Pod亲和定义

1
2
3
4
5
6
7
8
9
10
11
spec:
  affinity:
    podAffinity:
    preferredDuringSchedulingIgnoredDuringExecution:
    - podAffinityTerm
      weight
    requiredDuringSchedulingIgnoredDuringExecution:  # 硬亲和
    - labelSelector:					# 标签选择器，选择参照系此处为pod标签
      namespaceSelector:      # 名称空间选择器，和namespace二选一
      namsepace:							# 表示参照哪一个名称空间内的pod标签
      topologyKey:            # 拓扑标签，表示将哪个标签作为位置定义机制的。
pod硬亲和示例

1.编写资源清单示例

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
root@k8s-master01:~/yaml/chapter11# vim pod-affinity-required-demo.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: redis
spec:
  replicas: 1
  selector:
    matchLabels:
      app: redis
      ctlr: redis
  template:
    metadata:
      labels:
        app: redis
        ctlr: redis
    spec:
      containers:
      - name: redis
        image: redis:6.0-alpine
---																# 以上pod为基准。
apiVersion: apps/v1
kind: Deployment
metadata:
  name: pod-affinity-required
spec:
  replicas: 5
  selector:
    matchLabels:
      app: demoapp
      ctlr: pod-affinity-required
  template:
    metadata:
      labels:
        app: demoapp
        ctlr: pod-affinity-required
    spec:
      containers:
      - name: demoapp
        image: ikubernetes/demoapp:v1.0
      affinity:
        podAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:      # 定义pod的硬亲和关系
          - labelSelector:
              matchExpressions:                               # 使用匹配表达式来匹配上面的基准pod
              - {key: app, operator: In, values: ["redis"]}
              - {key: ctlr, operator: In, values: ["redis"]}
            topologyKey: rack                                 # 设置节点标签

2.对个节点打标签

1
2
3
4
5
6
root@k8s-master01:~/yaml/chapter11# kubectl label node k8s-node01 rack=foo
node/k8s-node01 labeled
root@k8s-master01:~/yaml/chapter11# kubectl label node k8s-node02 rack=bar
node/k8s-node02 labeled
root@k8s-master01:~/yaml/chapter11# kubectl label node k8s-node03 rack=baz
node/k8s-node03 labeled

3.应用资源清单

1
2
3
root@k8s-master01:~/yaml/chapter11# kubectl apply -f pod-affinity-required-demo.yaml
deployment.apps/redis created
deployment.apps/pod-affinity-required created

4.查看pod运行状态

1
2
3
4
5
6
7
8
9
10
root@k8s-master01:~/yaml/chapter11# kubectl get pods -o wide
NAME                                       READY   STATUS    RESTARTS   AGE   IP               NODE         NOMINATED NODE   READINESS GATES
pod-affinity-required-776d6948f9-5r827     1/1     Running   0          9s    192.168.96.25    k8s-node02   <none>           <none>
pod-affinity-required-776d6948f9-mthcc     1/1     Running   0          14s   192.168.96.24    k8s-node02   <none>           <none>
pod-affinity-required-776d6948f9-q6fzq     1/1     Running   0          14s   192.168.96.22    k8s-node02   <none>           <none>
pod-affinity-required-776d6948f9-rhs6h     1/1     Running   0          8s    192.168.96.26    k8s-node02   <none>           <none>
pod-affinity-required-776d6948f9-s4pmd     1/1     Running   0          14s   192.168.96.23    k8s-node02   <none>           <none>
redis-55f46d9795-zsqp9                     1/1     Running   0          28m   192.168.96.11    k8s-node02   <none>           <none>

# 可以看出所有的pod都运行在和reids相同的节点上。
pod软亲和示例

1.编辑资源清单

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
apiVersion: apps/v1
kind: Deployment
metadata:
  name: redis-preferred
spec:
  replicas: 1
  selector:
    matchLabels:
      app: redis
      ctlr: redis-preferred
  template:
    metadata:
      labels:
        app: redis
        ctlr: redis-preferred
    spec:
      containers:
      - name: redis
        image: redis:6.0-alpine
        resources:
          requests:
            cpu: 500m
            memory: 512Mi
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: pod-affinity-preferred
spec:
  replicas: 4
  selector:
    matchLabels:
      app: demoapp
      ctlr: pod-affinity-preferred
  template:
    metadata:
      labels:
        app: demoapp
        ctlr: pod-affinity-preferred
    spec:
      containers:
      - name: demoapp
        image: ikubernetes/demoapp:v1.0
        resources:
          requests:
            cpu: 1500m
            memory: 1Gi
      affinity:
        podAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:   # 软亲和
          - weight: 100																		# 权重100
            podAffinityTerm:          # pod亲和条目
              labelSelector:          # 标签选择器
                matchExpressions:
                - {key: app, operator: In, values: ["redis"]}
                - {key: ctlr, operator: In, values: ["redis-preferred"]}
              topologyKey: kubernetes.io/hostname
          - weight: 50                        # 权重50
            podAffinityTerm:
              labelSelector:
                matchExpressions:
                - {key: app, operator: In, values: ["redis"]}
                - {key: ctlr, operator: In, values: ["redis-preferred"]}
              topologyKey: rack

# 此处软亲和有2个选择
# 1.运行在和基准pod同一个节点之上
# 2.若无法运行在和基准pod同一节点之上时，运行在有rack标签的节点之上

2.应用资源清单

1
2
3
root@k8s-master01:~/yaml/chapter11# kubectl apply -f pod-affinity-preferred-demo.yaml
deployment.apps/redis-preferred created
deployment.apps/pod-affinity-preferred created

3.查看pod调度结果

1
2
3
4
5
6
7
8
9
10
root@k8s-master01:~/yaml/chapter11# kubectl get pods -o wide
NAME                                      READY   STATUS    RESTARTS   AGE    IP               NODE         NOMINATED NODE   READINESS GATES
pod-affinity-preferred-5dcfdf44c5-9jlkl   1/1     Running   0          101s   192.168.30.13    k8s-node03   <none>           <none>
pod-affinity-preferred-5dcfdf44c5-jwqf2   1/1     Running   0          101s   192.168.131.15   k8s-node01   <none>           <none>
pod-affinity-preferred-5dcfdf44c5-k42xh   1/1     Running   0          101s   192.168.96.27    k8s-node02   <none>           <none>
pod-affinity-preferred-5dcfdf44c5-lcbwz   1/1     Running   0          101s   192.168.96.28    k8s-node02   <none>           <none>
redis-preferred-56c75b8669-98q6j          1/1     Running   0          17s    192.168.96.29    k8s-node02   <none>           <none>

# 可以看出redis运行在node02之上，大多数的pod也被运行在了node02之上。
# 当node02上节点资源不足时，pod被调度到了拥有rack标签的其他节点之上了。
pod反亲和示例

1.编辑资源清单

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
root@k8s-master01:~/yaml/chapter11# vim pod-antiaffinity-required-demo.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: pod-antiaffinity-required
spec:
  replicas: 4
  selector:
    matchLabels:
      app: demoapp
      ctlr: pod-antiaffinity-required
  template:
    metadata:
      labels:
        app: demoapp
        ctlr: pod-antiaffinity-required
    spec:
      containers:
      - name: demoapp
        image: ikubernetes/demoapp:v1.0
      affinity:
        podAntiAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:    # 强制反亲和
          - labelSelector:
              matchExpressions:
              - {key: app, operator: In, values: ["demoapp"]}      # 此处两个label表示pod自身
              - {key: ctrl, operator: In, values: ["pod-antiaffinity-required"]}
            topologyKey: kubernetes.io/hostname   # 以节点名称为区域划分

2.应用资源清单

1
2
root@k8s-master01:~/yaml/chapter11# kubectl apply -f pod-antiaffinity-required-demo.yaml
deployment.apps/pod-antiaffinity-required created

3.查看pod调度

1
2
3
4
5
6
7
8
root@k8s-master01:~/yaml/chapter11# kubectl get pods -o wide
NAME                                         READY   STATUS    RESTARTS   AGE    IP               NODE         NOMINATED NODE   READINESS GATES
pod-antiaffinity-required-645c87f57d-8jwsl   1/1     Running   0          17s    192.168.30.19    k8s-node03   <none>           <none>
pod-antiaffinity-required-645c87f57d-ncmct   1/1     Running   0          17s    192.168.131.21   k8s-node01   <none>           <none>
pod-antiaffinity-required-645c87f57d-njnjg   0/1     Pending   0          17s    <none>           <none>       <none>           <none>
pod-antiaffinity-required-645c87f57d-q4jxq   1/1     Running   0          17s    192.168.96.32    k8s-node02   <none>           <none>

# 当亲有3个节点，第四个pod由于无法满足条件被挂起。