1、使用下面的命令操作使得master 可以作为node使用 承载pod

kubectl taint nodes --all node-role.kubernetes.io/master-

可能会出现下面的结果 因为taint（master标记的污点已经被去掉了）没有关系

taint "node-role.kubernetes.io/master" not found
taint "node-role.kubernetes.io/master" not found

2、可以用下面的

kubectl describe nodes master1 |grep Taint

命令确认一下taint已经没有了

[root@master1 ~]# kubectl describe nodes master1 |grep Taint
Taints:
[root@master1 ~]#

3、创建一个应用

kubectl create deployment --image=nginx --replicas=2 http

但是副本数量少的情况下有可能不会落在master机器上 这是k8s自己的策略 可以把

--replicas=20 改大一些 这样才会很明显的 看到pod分配到master和node节点了
注：如果想只起两个副本且必须分配到两个机器上 请参考nodeSelector以及亲和性和反亲和性的的用法，以后后续会写文章说明

[root@master1 ~]# kubectl get pods -o wide |grep http-54
http-54cc68bd9b-62gmc 1/1 Running 0 142m 10.130.209.203 node1
http-54cc68bd9b-7zkxj 1/1 Running 0 106m 10.130.209.208 node1
http-54cc68bd9b-95f74 1/1 Running 0 106m 10.129.40.90 master1
http-54cc68bd9b-9ntxh 1/1 Running 0 142m 10.130.209.201 node1
http-54cc68bd9b-b7q7l 1/1 Running 0 142m 10.130.209.202 node1
http-54cc68bd9b-bp5z4 1/1 Running 0 106m 10.130.209.211 node1
http-54cc68bd9b-fqtbh 1/1 Running 0 143m 10.130.209.200 node1
http-54cc68bd9b-g74dt 1/1 Running 0 106m 10.130.209.213 node1
http-54cc68bd9b-hj5w8 1/1 Running 0 106m 10.130.209.210 node1
http-54cc68bd9b-jtdg5 1/1 Running 0 106m 10.129.40.91 master1
http-54cc68bd9b-ktxhp 1/1 Running 0 142m 10.130.209.205 node1
http-54cc68bd9b-pxg7s 1/1 Running 0 106m 10.130.209.214 node1
http-54cc68bd9b-rbhmh 1/1 Running 0 106m 10.129.40.89 master1
http-54cc68bd9b-s7rx7 1/1 Running 0 143m 10.130.209.199 node1
http-54cc68bd9b-svsc6 1/1 Running 0 142m 10.130.209.206 node1
http-54cc68bd9b-tddbv 1/1 Running 0 106m 10.130.209.209 node1
http-54cc68bd9b-tjpdj 1/1 Running 0 106m 10.130.209.212 node1
http-54cc68bd9b-vrzv6 1/1 Running 0 106m 10.129.40.92 master1
http-54cc68bd9b-vw6wv 1/1 Running 0 142m 10.130.209.204 node1
http-54cc68bd9b-x86rz 1/1 Running 0 106m 10.130.209.207 node1
[root@master1 ~]#