无法从OpenStack上具有OpenShift源的其他节点主机到达节点主机上的pod IP地址

Question

我原来的问题可以在这里找到，但是可以扩展到更广泛的社

https://github.com/openshift/origin/issues/16447

我选择了一个4节点集群，如下所示，

[masters]
ci-master-01.example.com  
openshift_public_hostname="ci-master-01.example.com" ansible_host="10.145.78.110"

[etcd]
ci-etcd-node-01.example.com 
ansible_host="10.145.78.113"

[nodes]
ci-master-01.example.com 
openshift_schedulable=False  ansible_host="10.145.78.110" 

ci-infra-node-01.example.com 
openshift_schedulable=False  openshift_node_labels="{'region': 'infra', 
'zone': 'default'}" ansible_host="10.145.78.112"
ci-primary-node-01.example.com  
openshift_node_labels="{'region': 'primary', 'zone': 'default'}" 
ansible_host="10.145.78.111"

# Service Network CIDR
openshift_portal_net=172.30.0.0/16

# Pod Network CIDR
osm_cluster_network_cidr=10.128.0.0/14

我用这个库存来安装openshift origin v3.6。安装成功但我在运行应用程序之前尝试进行非常简单的网络测试。

oc version
oc v3.6.0+c4dd4cf
kubernetes v1.6.1+5115d708d7
features: Basic-Auth GSSAPI Kerberos SPNEGO

Server https://ci-master-01.example.com:443
openshift v3.6.0+c4dd4cf
kubernetes v1.6.1+5115d708d7

实际：

在师父身上

oc login

oc get endpoints
NAME               ENDPOINTS                                                
docker-registry    10.128.0.3:5000                                           
kubernetes         10.145.78.110:443,10.145.78.110:8053,10.145.78.110:8053   
registry-console   10.128.0.5:9090                                           
router             <none>

ping 10.128.0.3 --> Not reachable
ping 10.128.0.5 --> Not reachable

这两个PODS正在NODE主机上运行，​​我正试图从主站接收它们。

此外，我无法使用其名称或IP地址访问服务端点，

oc get svc
NAME               CLUSTER-IP       EXTERNAL-IP   PORT(S)                   AGE
docker-registry    172.30.115.147   <none>        5000/TCP                  6h
kubernetes         172.30.0.1       <none>        443/TCP,53/UDP,53/TCP     6h
registry-console   172.30.189.195   <none>        9000/TCP                  6h
router             172.30.179.178   <none>        80/TCP,443/TCP,1936/TCP   6h

curl -v 172.30.115.147:5000
* About to connect() to 172.30.115.147 port 5000 (#0)
*   Trying 172.30.115.147...
* No route to host
* Failed connect to 172.30.115.147:5000; No route to host
* Closing connection 0
curl: (7) Failed connect to 172.30.115.147:5000; No route to host


curl -v docker-registry:5000
curl: (6) Could not resolve host: docker-registry; Unknown error

预期：

应该能够达到这些地址。

附加信息：

我已经遵循了所有可用的故障排除技术，但没有在哪里。

https://access.redhat.com/documentation/en-us/openshift_enterprise/3.1/html/cluster_administration/admin-guide-sdn-troubleshooting#debugging-local-networking

Answer 1

我碰巧有类似的问题，最后是AWS Security Group的变化。在允许必要的端口之后，它就像魅力一样。