docke网络之bridge、host、none

一、bridge网络

1.创建一个测试容器

[root@localhost ~]# docker run -d -it --name busybox_1 busybox /bin/sh -c "while true;do sleep 3600;done"
03b308c847edd23f21ba69afb825d92f7aaeb05b1ff4431dd47ccee439a0361a

2.查看当前机器docker有哪些网络

[root@localhost ~]# docker network ls
NETWORK ID          NAME                DRIVER              SCOPE
fa30a4d17b5b        bridge              bridge              local
a03aaca35833        host                host                local
d85c50eb947c        none                null                local

3.查看bridge详细信息(如果没有指定网络，默认使用bridge网络)

[root@localhost ~]# docker network inspect fa30a4d17b5b  #fa30a4d17b5b  为bridge的ID
.....
        "Containers": {                    #该字段表示名称为busybox_1的Container网络连接到的是bridge这个网络
            "03b308c847edd23f21ba69afb825d92f7aaeb05b1ff4431dd47ccee439a0361a": {  #容器的ID
                "Name": "busybox_1",        #容器的名称
                "EndpointID": "c850f22941894ef8655a80a96e4be4c5045699b70b4bc17201f80f07a27a3b4d",
                "MacAddress": "02:42:ac:11:00:02",
                "IPv4Address": "172.17.0.2/16",  #地址
                "IPv6Address": ""
            }
        },
......

4.查看宿主机及容器busybox_1这个容器的网络接口,其中宿主机的veth66a7ab0@if110与容器中的eth0@if111网络接口实际上是一对pari,而veth66a7ab0@if110又连接到docker0上

[root@localhost ~]# ip a   #查看宿主网络接口
: lo: <LOOPBACK,UP,LOWER_UP> mtu  qdisc noqueue state UNKNOWN group default qlen
    link/loopback ::::: brd :::::
    inet 127.0.0.1/ scope host lo
       valid_lft forever preferred_lft forever
: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu  qdisc pfifo_fast state UP group default qlen
    link/ether :::fd::4b brd ff:ff:ff:ff:ff:ff
    inet 172.16.150.135/ brd 172.16.150.255 scope global eth0
       valid_lft forever preferred_lft forever
: docker0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu  qdisc noqueue state UP group default
    link/ether :::c0::f9 brd ff:ff:ff:ff:ff:ff
    inet 172.17.0.1/ scope global docker0
       valid_lft forever preferred_lft forever
: veth66a7ab0@if110: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu  qdisc noqueue master docker0 state UP group default
    link/ether 7e:::8b::a2 brd ff:ff:ff:ff:ff:ff link-netnsid 

[root@localhost ~]# docker exec busybox_1  ip a  #查看busybox_1容器的网络接口
: lo: <LOOPBACK,UP,LOWER_UP> mtu  qdisc noqueue qlen
    link/loopback ::::: brd :::::
    inet 127.0.0.1/ scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::/ scope host
       valid_lft forever preferred_lft forever
: eth0@if111: <BROADCAST,MULTICAST,UP,LOWER_UP,M-DOWN> mtu  qdisc noqueue
    link/ether ::ac::: brd ff:ff:ff:ff:ff:ff
    inet 172.17.0.2/ scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80:::acff:fe11:/ scope link
       valid_lft forever preferred_lft forever

5.验证veth66a7ab0网络与docker0相连接

[root@localhost ~]# brctl show   #如果没有该命令,yum安装 bridge-utils软件包即可
bridge name    bridge id        STP enabled    interfaces
docker0        .024223c091f9    no        veth66a7ab0

6.创建第二测试容器

[root@localhost ~]# docker run -d -it --name busybox_2 busybox /bin/sh -c "while true;do sleep 3600;done"
b884db0bf4a862281b1dfb66457c7f565896fce1a40151619e80c2c5b1499216

7.再次查看bridge网络信息

[root@localhost ~]# docker network  inspect  bridge
......
      "Containers": {
            "03b308c847edd23f21ba69afb825d92f7aaeb05b1ff4431dd47ccee439a0361a": {
                "Name": "busybox_1",
                "EndpointID": "c850f22941894ef8655a80a96e4be4c5045699b70b4bc17201f80f07a27a3b4d",
                "MacAddress": "02:42:ac:11:00:02",
                "IPv4Address": "172.17.0.2/16",
                "IPv6Address": ""
            },
            "b884db0bf4a862281b1dfb66457c7f565896fce1a40151619e80c2c5b1499216": {
                "Name": "busybox_2",  #busybox_2 也连接到bridge
                "EndpointID": "a5e56917165daf2965bf7f24cf9ce58c88e4ff3c1118544c49ca5f25172af28d",
                "MacAddress": "02:42:ac:11:00:03",
                "IPv4Address": "172.17.0.3/16",
                "IPv6Address": ""
            }
        },
......

8.查看本地网络接口,发现多了113: vethc039e93@if112

[root@localhost ~]# ip a
: lo: <LOOPBACK,UP,LOWER_UP> mtu  qdisc noqueue state UNKNOWN group default qlen
    link/loopback ::::: brd :::::
    inet 127.0.0.1/ scope host lo
       valid_lft forever preferred_lft forever
: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu  qdisc pfifo_fast state UP group default qlen
    link/ether :::fd::4b brd ff:ff:ff:ff:ff:ff
    inet 172.16.150.135/ brd 172.16.150.255 scope global eth0
       valid_lft forever preferred_lft forever
: docker0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu  qdisc noqueue state UP group default
    link/ether :::c0::f9 brd ff:ff:ff:ff:ff:ff
    inet 172.17.0.1/ scope global docker0
       valid_lft forever preferred_lft forever
: veth66a7ab0@if110: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu  qdisc noqueue master docker0 state UP group default
    link/ether 7e:::8b::a2 brd ff:ff:ff:ff:ff:ff link-netnsid
: vethc039e93@if112: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu  qdisc noqueue master docker0 state UP group default
    link/ether ea:a2:a8:dc:: brd ff:ff:ff:ff:ff:ff link-netnsid 

9.查看bridge网络信息,发现docker0连接率两个接口了

[root@localhost ~]# brctl show
bridge name    bridge id        STP enabled    interfaces
docker0        .024223c091f9    no        veth66a7ab0
                                        vethc039e93

简单拓扑图：

总结：实质上docker容器之间通过与docker0接口连接.实现先互直接通信,感觉有点像交换机?

10.docker访问公网简单拓扑图

小结：实质上docker访问公网网络通过docker0 NAT转发实现

二、host网络

1.查看当前服务端口

[root@localhost ~]# netstat -tnlp  #除了sshd，没有其他服务端口
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address           Foreign Address         State       PID/Program name
tcp               0.0.0.0:              0.0.0.0:*               LISTEN      /sshd
tcp6              :::                   :::*                    LISTEN      /sshd    

2.创建一个测试容器，指定使用host网络

[root@localhost ~]# docker run  -d --privileged --name my_centos  --network host  docker.io/centos /usr/sbin/init  #建议使用centos镜像，后续测试需要
cfb8d105dcb44947ce794d890b67c905df9aa1ba67ef2675fd1a51177d47835d

3.查看host网络信息(注意容器的网络信息)

[root@localhost ~]# docker network inspect host
......
 "Containers": {
            "cfb8d105dcb44947ce794d890b67c905df9aa1ba67ef2675fd1a51177d47835d": {
                "Name": "my_centos",
                "EndpointID": "4250d74b28f8125688bd7d0f1475a7d107135c0e87367a9c35c197fd981b7cd4",
                "MacAddress": "",   #可以看到我们创建的容器此时并没有Mac和IP地址
                "IPv4Address": "",
                "IPv6Address": ""
            }

......

4.进入容器内部，查看网络信息

[root@localhost ~]# docker exec -it my_centos /bin/bash
[root@localhost /]# ip a    #默认没有相关查看命令
bash: ip: command not found
[root@localhost /]# ifconfig
bash: ifconfig: command not found
[root@localhost /]# yum install net-tools -y  #我们yum安装(是不是很奇怪，没有网络地址怎么安装)
[root@localhost /]# ifconfig    #内容太多不粘了，但是显示的应该是当前宿主机的网络信息
[root@localhost /]# netstat -tnlp
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address           Foreign Address         State       PID/Program name
tcp               0.0.0.0:              0.0.0.0:*               LISTEN      -
tcp6              :::                   :::*                    LISTEN      -   

5.在容器总安装httpd服务器，并启动

[root@localhost /]# yum install httpd -y
[root@localhost /]# systemctl restart httpd
[root@localhost /]# netstat -tnlp  #发现多了80端口
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address           Foreign Address         State       PID/Program name
tcp               0.0.0.0:              0.0.0.0:*               LISTEN      /httpd
tcp               0.0.0.0:              0.0.0.0:*               LISTEN      -
tcp6              :::                   :::*                    LISTEN      -   

6.退出容器，在宿主机上查看当前网络连接

[root@localhost ~]# netstat -tnlp #宿主机上80端口也被开放了
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address           Foreign Address         State       PID/Program name
tcp               0.0.0.0:              0.0.0.0:*               LISTEN      /httpd
tcp               0.0.0.0:              0.0.0.0:*               LISTEN      /sshd
tcp6              :::                   :::*                    LISTEN      /sshd   

7.对于host模式的思考:

host模式下是怎么占领端口的？

host模式端口占用模式是你的容器占用你主机上当前所监听的端口(官网描述为publish)，比如我们都知道Nginx占用80端口，那么当我们用host模式启动的时候，主机上的80端口会被Nginx占用，这个时候其他的容器就不能指定我们的8080端口了，但是可以指定其他端口，所以说一台主机上可以运行多个host模式的容器，只要彼此监听的端口不一样就行。

host模式下使用-p或者-P会出现WARNING: Published ports are discarded when using host network mode

当你是host模式的时候，主机会自动把他上面的端口分配给容器，这个时候使用-p或者-P是无用的。但是还是可以在Dockerfile中声明EXPOSE端口

host模式设计的原因

host模式设计出来就是为了性能，访问主机的端口就能访问到我们的容器，使容器直接暴露在公网下，但是这却对docker的隔离性造成了破坏，使得安全性大大降低。这种模式有利有弊，对于每个人来说看法都不一样，具体取舍看个人。

三、none网络

1.创建一个测试容器

[root@localhost ~]#  docker run -d --name test1 --network none busybox  /bin/sh -c "while true;do sleep 36000;done"
ca1771ebfe436137156568cd570c116d12bd85e782dbec365c9f62a70209d028

2.查看none网络信息

[root@localhost ~]# docker network inspect none
......
        "Containers": {
            "ca1771ebfe436137156568cd570c116d12bd85e782dbec365c9f62a70209d028": {
                "Name": "test1",  #容器无法看到Mac和IP地址
                "EndpointID": "ddcff44cdedb78f59108c6978345a256baa8bb09965461b2ffac58d5334fdba6",
                "MacAddress": "",
                "IPv4Address": "",
                "IPv6Address": ""
            }
......

3.进入容器查看网络信息

[root@localhost ~]# docker exec -it test1 /bin/sh
/ # ip a           #只有回环接口
: lo: <LOOPBACK,UP,LOWER_UP> mtu  qdisc noqueue qlen
    link/loopback ::::: brd :::::
    inet 127.0.0.1/ scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::/ scope host
       valid_lft forever preferred_lft forever

 小结：none 网络的容器只有一个自己的回环接口，没有任何 IP 地址分配，所有这个网络中的每个容器都是孤立的。但是所有网络配置我们都可以自己配置，如IP、网卡等