第五章-k8s调度-pod调度控制篇

1. 创建一个pod,分配到指定标签node上

  • pod名: web
  • 镜像: nginx
  • node标签: disk=ssd

答案

1
2
# 给node打标签
kubectl label node n2 disk=ssd

kubectl run web --image=nginx -oyaml --dry-run >1.yaml创建yaml

1
2
3
4
5
6
7
8
9
10
11
12
13
apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web
  name: web
spec:
  containers:
  - image: nginx
    name: web
    imagePullPolicy: IfNotPresent
  nodeSelector:
    disk: ssd

kubectl apply -f 1.yaml创建pod

1
2
3
4
# 查看pod运行是否运行在n2节点
[m1 root ~]# kubectl get po -owide
NAME   READY   STATUS    RESTARTS   AGE     IP              NODE   NOMINATED NODE   READINESS GATES
web    1/1     Running   0          2m41s   100.104.217.1   n2     <none>           <none>

2.确保每个节点上运行一个pod

  • pod名称: nginx
  • 镜像: nginx

答案

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
apiVersion: apps/v1
kind: Deployment
metadata:
  name: redis-cache
spec:
  selector:
    matchLabels:
      app: store
  # 副本数3
  replicas: 3
  template:
    metadata:
      labels:
        app: store
    spec:
      affinity:
        # pod反亲和
        podAntiAffinity:
          # 硬需求/硬策略
          requiredDuringSchedulingIgnoredDuringExecution:
          - labelSelector:
              matchExpressions:
              # 禁止在同一node上部署多个标签为app=strore的副本
              - key: app
                operator: In
                values:
                - store
            # 用主机名来作为拓扑域
            topologyKey: "kubernetes.io/hostname"
      containers:
      - name: redis-server
        image: redis:3.2-alpine

3.查看集群中状态为ready的node数量,并将结果写到指定文件

答案:

1
2
3
4
5
6
7
8
9
# 查看nodes状态()
[m1 root ~]# kubectl describe nodes|grep -i taint
Taints:             <none>
Taints:             <none>
Taints:             <none>
Taints:             <none>

# 写入结果
echo 4 >/nodes_num.txt