clico网络问题

Question

现在的集群是 一个master 一个slave ，
master 的运行状态良好。
但是save加入calico/node 的时候 ，获取到的描述信息是

                 node.kubernetes.io/unschedulable:NoSchedule
Events:
  Type     Reason     Age                   From                        Message
  ----     ------     ----                  ----                        -------
  Normal   Scheduled  37m                   default-scheduler           Successfully assigned kube-system/calico-node-9f4mh to instance-esj3coh3
  Normal   Pulled     37m                   kubelet, instance-esj3coh3  Container image "calico/cni:v3.13.2" already present on machine
  Normal   Created    37m                   kubelet, instance-esj3coh3  Created container upgrade-ipam
  Normal   Started    37m                   kubelet, instance-esj3coh3  Started container upgrade-ipam
  Normal   Started    37m                   kubelet, instance-esj3coh3  Started container install-cni
  Normal   Pulled     37m                   kubelet, instance-esj3coh3  Container image "calico/cni:v3.13.2" already present on machine
  Normal   Created    37m                   kubelet, instance-esj3coh3  Created container install-cni
  Normal   Started    37m                   kubelet, instance-esj3coh3  Started container flexvol-driver
  Normal   Pulled     37m                   kubelet, instance-esj3coh3  Container image "calico/pod2daemon-flexvol:v3.13.2" already present on machine
  Normal   Created    37m                   kubelet, instance-esj3coh3  Created container flexvol-driver
  Normal   Pulled     37m                   kubelet, instance-esj3coh3  Container image "calico/node:v3.13.2" already present on machine
  Normal   Created    37m                   kubelet, instance-esj3coh3  Created container calico-node
  Normal   Started    37m                   kubelet, instance-esj3coh3  Started container calico-node
  Normal   Killing    36m                   kubelet, instance-esj3coh3  Container calico-node failed liveness probe, will be restarted
  Warning  Unhealthy  36m (x6 over 37m)     kubelet, instance-esj3coh3  Readiness probe failed: calico/node is not ready: felix is not ready: Get http://localhost:9099/readiness: dial tcp [::1]:9099: connect: connection refused
  Warning  Unhealthy  12m (x69 over 37m)    kubelet, instance-esj3coh3  Liveness probe failed: calico/node is not ready: Felix is not live: Get http://localhost:9099/liveness: dial tcp [::1]:9099: connect: connection refused
  Warning  BackOff    2m44s (x93 over 30m)  kubelet, instance-esj3coh3  Back-off restarting failed container

kubelet, instance-esj3coh3 Liveness probe failed: calico/node is not ready: Felix is not live: Get http://localhost:9099/liveness: dial tcp [::1]:9099: connect: connection
他获取的不因该是 master的IP吗怎么是localhost了。
我看了master上面 calico绑定的也是 127.0.0.1

Proto Recv-Q Send-Q Local Address           Foreign Address         State       PID/Program name    
tcp        0      0 127.0.0.1:9099          0.0.0.0:*               LISTEN      17421/calico-node   
tcp        0      0 192.168.0.3:2379        0.0.0.0:*               LISTEN      16224/etcd          
tcp        0      0 127.0.0.1:2379          0.0.0.0:*               LISTEN      16224/etcd          
tcp        0      0 192.168.0.3:2380        0.0.0.0:*               LISTEN      16224/etcd          
tcp        0      0 127.0.0.1:10257         0.0.0.0:*               LISTEN      16148/kube-controll 

怎么才能绑定外网IP呢。

具体的日志信息

                 node.kubernetes.io/unschedulable:NoSchedule
Events:
  Type     Reason     Age                   From                        Message
  ----     ------     ----                  ----                        -------
  Normal   Scheduled  37m                   default-scheduler           Successfully assigned kube-system/calico-node-9f4mh to instance-esj3coh3
  Normal   Pulled     37m                   kubelet, instance-esj3coh3  Container image "calico/cni:v3.13.2" already present on machine
  Normal   Created    37m                   kubelet, instance-esj3coh3  Created container upgrade-ipam
  Normal   Started    37m                   kubelet, instance-esj3coh3  Started container upgrade-ipam
  Normal   Started    37m                   kubelet, instance-esj3coh3  Started container install-cni
  Normal   Pulled     37m                   kubelet, instance-esj3coh3  Container image "calico/cni:v3.13.2" already present on machine
  Normal   Created    37m                   kubelet, instance-esj3coh3  Created container install-cni
  Normal   Started    37m                   kubelet, instance-esj3coh3  Started container flexvol-driver
  Normal   Pulled     37m                   kubelet, instance-esj3coh3  Container image "calico/pod2daemon-flexvol:v3.13.2" already present on machine
  Normal   Created    37m                   kubelet, instance-esj3coh3  Created container flexvol-driver
  Normal   Pulled     37m                   kubelet, instance-esj3coh3  Container image "calico/node:v3.13.2" already present on machine
  Normal   Created    37m                   kubelet, instance-esj3coh3  Created container calico-node
  Normal   Started    37m                   kubelet, instance-esj3coh3  Started container calico-node
  Normal   Killing    36m                   kubelet, instance-esj3coh3  Container calico-node failed liveness probe, will be restarted
  Warning  Unhealthy  36m (x6 over 37m)     kubelet, instance-esj3coh3  Readiness probe failed: calico/node is not ready: felix is not ready: Get http://localhost:9099/readiness: dial tcp [::1]:9099: connect: connection refused
  Warning  Unhealthy  12m (x69 over 37m)    kubelet, instance-esj3coh3  Liveness probe failed: calico/node is not ready: Felix is not live: Get http://localhost:9099/liveness: dial tcp [::1]:9099: connect: connection refused
  Warning  BackOff    2m44s (x93 over 30m)  kubelet, instance-esj3coh3  Back-off restarting failed container
[root@JD addons]# kubectl  --namespace kube-system get pods -o wide 
NAME                                       READY   STATUS    RESTARTS   AGE   IP               NODE                NOMINATED NODE   READINESS GATES
calico-kube-controllers-6898678fbd-wwtql   1/1     Running   0          39m   192.168.29.194   jd                  <none>           <none>
calico-node-9f4mh                          0/1     Running   14         38m   172.16.0.4       instance-esj3coh3   <none>           <none>
calico-node-qwwpk                          1/1     Running   0          39m   192.168.0.3      jd                  <none>           <none>
coredns-8567978547-qf2tk                   1/1     Running   0          39m   192.168.29.193   jd                  <none>           <none>
coredns-8567978547-rpnzg                   1/1     Running   0          39m   192.168.29.195   jd                  <none>           <none>
etcd-jd                                    1/1     Running   0          38m   192.168.0.3      jd                  <none>           <none>
kube-apiserver-jd                          1/1     Running   0          38m   192.168.0.3      jd                  <none>           <none>
kube-controller-manager-jd                 1/1     Running   0          38m   192.168.0.3      jd                  <none>           <none>
kube-proxy-p7lwr                           1/1     Running   0          38m   172.16.0.4       instance-esj3coh3   <none>           <none>
kube-proxy-tkl9t                           1/1     Running   0          39m   192.168.0.3      jd                  <none>           <none>
kube-scheduler-jd                          1/1     Running   0          38m   192.168.0.3      jd                  <none>           <none>
[root@JD addons]# netstat -tpnul

Answer 1

tonybai 2020-04-14 15:02:38

亲，calico不在本课程范围内。但通过你给出的日志来看，你看一下安装calico时，calico的yaml文件中关于相关pod 的readiess和liveness的探针定义是怎么写的。或到calico项目的issue中看看是否有人遇到与你相似的问题了。祝学习愉快！