在虚拟机上，搭建了一个k8s单机项目，创建了一个nginx-service

apiVersion: v1

kind: Service

metadata:

name: nginx-service

spec:

type: NodePort

ports:

- port: 80

nodePort: 30001

selector:

name: nginx-pod

服务已经启动：

kubectl describe svc nginx-service

Name:           nginx-service

Namespace:      default

Labels:         <none>

Annotations:        <none>

Selector:       name=nginx-pod

Type:           NodePort

IP:         10.254.96.42

Port:           <unset>  80/TCP

NodePort:       <unset>  30001/TCP

Endpoints:      192.168.4.130:80

Session Affinity:   None

Events:         <none>

虚拟机上执行如下三个都是没问题的。

虚拟机IP 10.211.55.7
curl 10.211.55.7:30001

curl 10.254.96.42

curl 192.168.4.130

但是在我的电脑（安装虚拟机的机器）上访问：10.211.55.7:30001 无法链接。
访问kubernetes api server是没问题的：10.211.55.7:8080

物理机网络情况：

en0: flags=8863<UP,BROADCAST,SMART,RUNNING,SIMPLEX,MULTICAST> mtu 1500

ether 60:f8:1d:d1:3c:2e

inet6 fe80::14cb:c20d:d275:e7ff%en0 prefixlen 64 secured scopeid 0x4

inet 192.168.22.151 netmask 0xffffff00 broadcast 192.168.22.255

nd6 options=201<PERFORMNUD,DAD>

media: autoselect

status: active

vnic0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500

options=3<RXCSUM,TXCSUM>

ether 00:1c:42:00:00:08

inet 10.211.55.2 netmask 0xffffff00 broadcast 10.211.55.255

media: autoselect

status: active

vnic1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500

options=3<RXCSUM,TXCSUM>

ether 00:1c:42:00:00:09

inet 10.37.129.2 netmask 0xffffff00 broadcast 10.37.129.255

media: autoselect

status: active

虚拟机的网络环境：

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000

link/ether 00:1c:42:79:26:7a brd ff:ff:ff:ff:ff:ff

inet 10.211.55.7/24 brd 10.211.55.255 scope global dynamic eth0

   valid_lft 1296sec preferred_lft 1296sec

inet6 fdb2:2c26:f4e4:0:5f7b:999b:5e7f:5284/64 scope global noprefixroute dynamic

   valid_lft 2591689sec preferred_lft 604489sec

inet6 fe80::3503:678d:3637:111a/64 scope link

   valid_lft forever preferred_lft forever

4: br-8373b59b5c3f: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN

link/ether 02:42:27:01:b7:3b brd ff:ff:ff:ff:ff:ff

inet 172.18.0.1/16 scope global br-8373b59b5c3f

   valid_lft forever preferred_lft forever

inet6 fe80::42:27ff:fe01:b73b/64 scope link

   valid_lft forever preferred_lft forever

119: flannel0: <POINTOPOINT,MULTICAST,NOARP,UP,LOWER_UP> mtu 1472 qdisc pfifo_fast state UNKNOWN qlen 500

link/none

inet 192.168.4.128/24 scope global flannel0

   valid_lft forever preferred_lft forever

inet6 fe80::6742:a52f:f02:f31e/64 scope link flags 800

   valid_lft forever preferred_lft forever

120: docker0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1472 qdisc noqueue state UP

link/ether 02:42:c4:ce:f0:fe brd ff:ff:ff:ff:ff:ff

inet 192.168.4.129/25 scope global docker0

   valid_lft forever preferred_lft forever

inet6 fe80::42:c4ff:fece:f0fe/64 scope link

   valid_lft forever preferred_lft forever

安装kubernetes配置的cluster-ip-range: 10.254.0.0/16
配置的etcd flanneld 网段为：192.168.4.0/24

请大牛帮看看，为什么在物理机上无法访问.