Node节点

Node节点

还有其他叫法Woker节点，minion节点

 

Node节点上主要安装两个软件：

    Kubelet：

    负责监听借调上pod状态，同时负责上报节点和节点上面的pod状态，负责与Master节点通信，并管理节点上的pod

   

    Kube-proxy：

    负责pod直接的通信和负载均衡，将指定流量分发到后端正确的机器上

    查看Kube-proxy工作模式查看：curl 127.0.0.1:10249/proxyMode

    

    IPVS：

    监听Master节点增加和删除service以及endpoint的信息，调用Netlink接口创建相应的IPVS规则。

    通过IPVS规则，将流量转发到响应的pod上

    查看ipvs转发表：ipvsadm -ln

 

    Iptables：

    监听Master节点增加和删除service以及endpoint的消息，对应每一个service，它都会创建一个iptables规则

    将service的clusterIP代理到后端对应的pod

 

 

    Calico：

    符合CNI标准的网络插件，给每个pod生成唯一的IP地址，并且把每个节点当做一个路由器

    跨节点pod通信的时候都是可以互通的

[root@master1 ~]# kubectl get po -n kube-system -owide
NAME                                       READY   STATUS    RESTARTS   AGE     IP               NODE          NOMINATED NODE   READINESS GATES
calico-kube-controllers-7b5bcff94c-v8lz8   1/1     Running   0          3h37m   172.16.159.131   k8s-master1   <none>           <none>
calico-node-55txl                          1/1     Running   0          124m    192.168.1.230    node3         <none>           <none>
calico-node-65j2t                          1/1     Running   0          124m    192.168.1.100    node1         <none>           <none>
calico-node-87w4z                          1/1     Running   0          3h37m   192.168.1.243    k8s-master1   <none>           <none>
calico-node-rlfj8                          1/1     Running   0          124m    192.168.1.175    node2         <none>           <none>
calico-node-sd77l                          1/1     Running   0          3h9m    192.168.1.128    master3       <none>           <none>
calico-node-zlv45                          1/1     Running   0          3h10m   192.168.1.189    master2       <none>           <none>
coredns-546565776c-68cfn                   1/1     Running   0          4h19m   172.16.159.129   k8s-master1   <none>           <none>
coredns-546565776c-hwqbw                   1/1     Running   0          4h19m   172.16.159.130   k8s-master1   <none>           <none>
etcd-k8s-master1                           1/1     Running   0          4h20m   192.168.1.243    k8s-master1   <none>           <none>
etcd-master2                               1/1     Running   0          3h10m   192.168.1.189    master2       <none>           <none>
etcd-master3                               1/1     Running   0          3h10m   192.168.1.128    master3       <none>           <none>
kube-apiserver-k8s-master1                 1/1     Running   0          4h20m   192.168.1.243    k8s-master1   <none>           <none>
kube-apiserver-master2                     1/1     Running   0          3h10m   192.168.1.189    master2       <none>           <none>
kube-apiserver-master3                     1/1     Running   0          3h8m    192.168.1.128    master3       <none>           <none>
kube-controller-manager-k8s-master1        1/1     Running   1          4h20m   192.168.1.243    k8s-master1   <none>           <none>
kube-controller-manager-master2            1/1     Running   0          3h10m   192.168.1.189    master2       <none>           <none>
kube-controller-manager-master3            1/1     Running   0          3h9m    192.168.1.128    master3       <none>           <none>
kube-proxy-22672                           1/1     Running   0          124m    192.168.1.230    node3         <none>           <none>
kube-proxy-6sw7r                           1/1     Running   0          4h19m   192.168.1.243    k8s-master1   <none>           <none>
kube-proxy-crdrf                           1/1     Running   0          3h9m    192.168.1.128    master3       <none>           <none>
kube-proxy-kvx97                           1/1     Running   0          124m    192.168.1.175    node2         <none>           <none>
kube-proxy-l2ntr                           1/1     Running   0          3h10m   192.168.1.189    master2       <none>           <none>
kube-proxy-wxqsk                           1/1     Running   0          124m    192.168.1.100    node1         <none>           <none>
kube-scheduler-k8s-master1                 1/1     Running   1          4h20m   192.168.1.243    k8s-master1   <none>           <none>
kube-scheduler-master2                     1/1     Running   0          3h10m   192.168.1.189    master2       <none>           <none>
kube-scheduler-master3                     1/1     Running   0          3h8m    192.168.1.128    master3       <none>           <none>

 

[root@master1 ~]# route -n
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
0.0.0.0         192.168.1.1     0.0.0.0         UG    0      0        0 enp0s3
169.254.0.0     0.0.0.0         255.255.0.0     U     1002   0        0 enp0s3
172.16.104.0    192.168.1.175   255.255.255.192 UG    0      0        0 tunl0
172.16.135.0    192.168.1.230   255.255.255.192 UG    0      0        0 tunl0
172.16.136.0    192.168.1.128   255.255.255.192 UG    0      0        0 tunl0
172.16.159.128  0.0.0.0         255.255.255.192 U     0      0        0 *
172.16.159.129  0.0.0.0         255.255.255.255 UH    0      0        0 cali2f7160f9e3b
172.16.159.130  0.0.0.0         255.255.255.255 UH    0      0        0 calid636110dee6
172.16.159.131  0.0.0.0         255.255.255.255 UH    0      0        0 cali7447d7d4a6b
172.16.166.128  192.168.1.100   255.255.255.192 UG    0      0        0 tunl0
172.16.180.0    192.168.1.189   255.255.255.192 UG    0      0        0 tunl0
172.17.0.0      0.0.0.0         255.255.0.0     U     0      0        0 docker0
192.168.1.0     0.0.0.0         255.255.255.0   U     0      0        0 enp0s3

 

    CoreDNS：

    用于kubernetes集群内部的service解析，可以让pod把service名称解释程IP地址

    通过service的IP地址去连接对应的应用上