Question

在Ubuntu 16.04机器上设置Kubernetes集群时遇到问题，我已经正确设置了：

主节点（机器IP：10.10.1.122）
工人节点（机器IP：10.10.1.38）

包含以下信息的集群：

    NAME                                        READY   STATUS    RESTARTS   AGE
pod/coredns-86c58d9df4-78lnp                1/1     Running   0          80m
pod/coredns-86c58d9df4-lw7vl                1/1     Running   0          80m
pod/etcd-di-linux-host                      1/1     Running   0          111m
pod/kube-apiserver-di-linux-host            1/1     Running   0          110m
pod/kube-controller-manager-di-linux-host   1/1     Running   0          111m
pod/kube-flannel-ds-amd64-6wvkh             1/1     Running   0          109m
pod/kube-flannel-ds-amd64-p7ftb             1/1     Running   0          110m
pod/kube-proxy-rbfvz                        1/1     Running   0          109m
pod/kube-proxy-zwr7b                        1/1     Running   0          111m
pod/kube-scheduler-di-linux-host            1/1     Running   0          111m
pod/kubernetes-dashboard-79ff88449c-9f8qw   1/1     Running   0          89m

NAME                           TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)         AGE
service/kube-dns               ClusterIP   10.96.0.10      <none>        53/UDP,53/TCP   111m
service/kubernetes-dashboard   ClusterIP   10.98.188.215   <none>        443/TCP         89m

NAME                                     DESIRED   CURRENT   READY   UP-TO-DATE   AVAILABLE   NODE SELECTOR                     AGE
daemonset.apps/kube-flannel-ds-amd64     2         2         2       2            2           beta.kubernetes.io/arch=amd64     110m
daemonset.apps/kube-flannel-ds-arm       0         0         0       0            0           beta.kubernetes.io/arch=arm       110m
daemonset.apps/kube-flannel-ds-arm64     0         0         0       0            0           beta.kubernetes.io/arch=arm64     110m
daemonset.apps/kube-flannel-ds-ppc64le   0         0         0       0            0           beta.kubernetes.io/arch=ppc64le   110m
daemonset.apps/kube-flannel-ds-s390x     0         0         0       0            0           beta.kubernetes.io/arch=s390x     110m
daemonset.apps/kube-proxy                2         2         2       2            2           <none>                            111m

NAME                                   READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/coredns                2/2     2            2           111m
deployment.apps/kubernetes-dashboard   1/1     1            1           89m

NAME                                              DESIRED   CURRENT   READY   AGE
replicaset.apps/coredns-86c58d9df4                2         2         2       111m
replicaset.apps/kubernetes-dashboard-79ff88449c   1         1         1       89m

我的集群信息

Kubernetes master is running at https://10.10.1.122:6443
KubeDNS is running at https://10.10.1.122:6443/api/v1/namespaces/kube-system/services/kube-dns:dns/proxy

我的豆荚

NAME                     READY   STATUS    RESTARTS   AGE
guids-68898f7dc9-c65nv   1/1     Running   0          102m
Name:               guids-68898f7dc9-c65nv
Namespace:          default
Priority:           0
PriorityClassName:  <none>
Node:               gmf.com/10.10.1.38
Start Time:         Sun, 16 Dec 2018 15:43:41 +0200
Labels:             pod-template-hash=68898f7dc9
                    run=guids
Annotations:        <none>
Status:             Running
IP:                 10.244.1.15
Controlled By:      ReplicaSet/guids-68898f7dc9
Containers:
  guids:
    Container ID:   docker://125ceccad4e572b514538292aaeaa55e22050c5e9129f834de8e01dfd452c6c1
    Image:          alexellis2/guid-service:latest
    Image ID:       docker-pullable://alexellis2/guid-service@sha256:17207f799760ccdccd0fa1e7f37838af5df915a33f9f27e97951a6eeee8c3a6f
    Port:           9000/TCP
    Host Port:      0/TCP
    State:          Running
      Started:      Sun, 16 Dec 2018 15:43:46 +0200
    Ready:          True
    Restart Count:  0
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-hnwtc (ro)
Conditions:
  Type              Status
  Initialized       True 
  Ready             True 
  ContainersReady   True 
  PodScheduled      True 
Volumes:
  default-token-hnwtc:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-hnwtc
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:          <none>

我在这里面临的问题是，无论何时我想卷曲服务或主节点上的pod IP都不会碰到，同时卷曲来自工作节点的服务/ pod很好，我是新手到Kubernetes，但我找不到如何诊断此问题的线索，我们将不胜感激。

当我尝试卷曲暴露的服务时，即使我从主服务器那里得到了这个结果：

NAME         TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)    AGE
guids        ClusterIP   10.97.160.160   <none>        9000/TCP   92m
kubernetes   ClusterIP   10.96.0.1       <none>        443/TCP    139m
ubuntu@master:/$ curl http://10.97.160.160:9000/guid
curl: (7) Failed to connect to 10.97.160.160 port 9000: Connection timed out

Answer 1

可从其他节点访问的Pod IP和可从kubernetes集群内的Pod访问的clusterIP。

服务的集群IP不是容器的IP地址，它是根据服务中定义的规则映射到容器IP地址的虚拟地址，并且由kubernetes集群中的kube-proxy守护程序管理

集群内部的通信特别需要集群IP，以访问服务背后的Pod，而无需关心Pod的副本数或节点Pod的存在以及clusterIP是静态的，这与Pod的IP不同。

如果可以阅读有关official documentation中服务IP的工作方式

我希望您遵循一些调试步骤：

您可以使用以下方法从pod内检查服务名称是否解析为ClusterIP：

kubectl exec -it <pod_name> bash
nslookup  <svc_name>.<namespace>.svc.cluster.local

以上命令将为您提供服务的clusterIP。

检查工作节点是否可以从主节点ping通，如果不是，则说明覆盖网络存在问题，如果是法兰绒，则为问题。

Kubernetes网络问题

1 个答案: