使用ovs构建docker网络,实现跨主机通信与网络隔离

TaoTaoFu 2019-06-25

初始化环境

升级内核:
      rpm --import https://www.elrepo.org/RPM-GPG-KEY-elrepo.org
      rpm -Uvh http://www.elrepo.org/elrepo-release-7.0-2.el7.elrepo.noarch.rpm
      yum --enablerepo=elrepo-kernel install  kernel-ml-devel kernel-ml
      awk -F\' '$1=="menuentry " {print $2}' /etc/grub2.cfg
      grub2-set-default 0
      reboot
      uname -a

安装docker:
     yum install -y yum-utils
     yum-config-manager     --add-repo     https://download.docker.com/linux/centos/docker-ce.repo
     yum-config-manager --disable docker-ce-edge
     yum makecache fast
     yum install docker-ce
     service docker start


安装open vswitch:
     yum -y install wget openssl-devel gcc make python-devel openssl-devel kernel-devel graphviz kernel-debug-devel autoconf automake rpm-build redhat-rpm-config libtool python-twisted-core python-zope-interface PyQt4 desktop-file-utils libcap-ng-devel groff checkpolicy selinux-policy-devel
     adduser ovs
     su - ovs
     yum localinstall /home/ovs/rpmbuild/RPMS/x86_64/openvswitch-2.5.0-1.el7.centos.x86_64.rpm -y
     systemctl start openvswitch.service
     systemctl is-active openvswitch
     ovs-vsctl -V
     systemctl enable openvswitch

安装pipework:
     yum install git
     git clone https://github.com/jpetazzo/pipework
     cp pipework/pipework /bin

一些工具安装:
yum install bridge-utils # 如果brctl不能用

yum install net-tools # 如果route命令不能用

单节点上使用ovs vlan划分网络

启动四个容器:

docker run -itd --name con1 ubuntu:14.04 /bin/bash
docker run -itd --name con2 ubuntu:14.04 /bin/bash
docker run -itd --name con3 ubuntu:14.04 /bin/bash
docker run -itd --name con4 ubuntu:14.04 /bin/bash

创建ovs网桥并绑定端口

pipework ovs0 con1 192.168.0.1/24 @100
pipework ovs0 con2 192.168.0.2/24 @100

pipework ovs0 con3 192.168.0.3/24 @200
pipework ovs0 con4 192.168.0.4/24 @200

这样con1 和 con2是通的,con3和con4是通的,这个比较简单。pipework干的具体的事是:

ovs-vsctl add-port ovs0 [容器的虚拟网卡设备] tag=100

ovs划分vlan处理的原理也非常简单,包进入到switch时打上tag,发出去时去掉tag,发出去的端口与包的tag不匹配时不处理,这便实现了二层隔离。

access端口与trunk端口的区别是,trunk端口可接受多个tag。

跨主机vlan

准备两个主机,在host1上:

docker run -itd --name con1 ubuntu:14.04 /bin/bash
docker run -itd --name con2 ubuntu:14.04 /bin/bash
pipework ovs0 con1 192.168.0.1/24 @100
pipework ovs0 con2 192.168.0.2/24 @200

如果是单张网卡的话,把eth0桥接到switch上时会造成网络中断,所以以下几步不要通过ssh操作:
如果非得ssh去操作的话把以下命令放在一条命令中执行(用&&连接各个命令)

ovs-vsctl add-port ovs0 eth0
ifconfig ovs0 10.1.86.201 netmask 255.255.255.0   # 这里地址和掩码与eth0的配置一致
ifconfig ovs0 up
ifconfig eth0 0.0.0.0
route add default gw 10.1.86.1  # 执行之前看看eth0的gw是什么,保持一致,这样eth0就桥接到ovs0上去了。

查看switch端口:

[root@dev-86-204 ~]# ovs-vsctl show
c5ddf9e8-daac-4ed2-80f5-16e6365425fa
    Bridge "ovs0"
        Port "veth1pl41885"
            tag: 100
            Interface "veth1pl41885"
        Port "ovs0"
            Interface "ovs0"
                type: internal
        Port "eth0"
            Interface "eth0"
        Port "veth1pl41805"
            tag: 200
            Interface "veth1pl41805"
    ovs_version: "2.5.1"

在host2上:

docker run -itd --name con3 ubuntu:14.04 /bin/bash
docker run -itd --name con4 ubuntu:14.04 /bin/bash
pipework ovs0 con3 192.168.0.3/24 @100
pipework ovs0 con4 192.168.0.4/24 @200

同样要桥接eth0到ovs0上,同host1的操作,然后con1与con3可通,con2与con4可通.

GRE实现overlay网络

linux内核需要3.11以上,本尊在3.10内核上实践失败,在虚拟机中升级内核时虚拟机启动不了,CPU飙到100%,以后再试。

发现把容器直接挂ovs网桥上是可以通的

gre与下面的vxlan非常类似,只需要在添加往外连的端口时改成下面命令即可:

ovs-vsctl add-port ovs0 gre0 -- set interface gre0 type=gre options:remote_ip=172.31.244.185

Vxlan实现跨主机通信

host1:10.1.86.203 
   ovs0
    |
    |-veth1 <-------> eth1 192.168.0.3  con3
    |
    |-vxlan1-------------+
    |                    |
                         |
    host2:10.1.86.204    |
    ovs0                 |
     |                   |
     |-vxlan1------------+
     |
     |-veth1 <--------> eth1 192.168.0.4 con4
     |

可以看到con3和con4在搭建vxlan之前是无法通信的。

在host1上:

[root@dev-86-203 ~]# docker run --name con3 -itd ubuntu:14.04 /bin/bash
[root@dev-86-203 ~]# ovs-vsctl add-br ovs0
[root@dev-86-203 ~]# pipework ovs0 con3 192.168.0.3/24   # 给容器分配地址并挂到ovs0上

[root@dev-86-203 ~]# ovs-vsctl add-port ovs0 vxlan1 -- set interface vxlan1 type=vxlan options:remote_ip=10.1.86.204 options:key=flow # 创建vxlan

[root@dev-86-203 ~]# ovs-vsctl show
5e371797-db70-451c-a0f2-d70c6d00cd05
    Bridge "ovs0"
        Port "veth1pl3342"
            Interface "veth1pl3342"
        Port "ovs0"
            Interface "ovs0"
                type: internal
        Port "vxlan1"
            Interface "vxlan1"
                type: vxlan
                options: {key=flow, remote_ip="10.1.86.204"}
    ovs_version: "2.5.1"

host2上同理:

[root@dev-86-204 ~]# docker run --name con4 -itd ubuntu:14.04 /bin/bash
[root@dev-86-204 ~]# ovs-vsctl add-br ovs0
[root@dev-86-204 ~]# pipework ovs0 con4 192.168.0.4/24   # 给容器分配地址并挂到ovs0上

[root@dev-86-204 ~]# ovs-vsctl add-port ovs0 vxlan1 -- set interface vxlan1 type=vxlan options:remote_ip=10.1.86.203 options:key=flow # 创建vxlan

[root@dev-86-204 ~]# ovs-vsctl show
c5ddf9e8-daac-4ed2-80f5-16e6365425fa
    Bridge "ovs0"
        Port "ovs0"
            Interface "ovs0"
                type: internal
        Port "veth1pl52846"
            Interface "veth1pl52846"
        Port "vxlan1"
            Interface "vxlan1"
                type: vxlan
                options: {key=flow, remote_ip="10.1.86.203"}
    ovs_version: "2.5.1"

验证:

[root@dev-86-204 ~]# docker exec con4 ping 192.168.0.3  # con4容器中ping con3的地址,可通
PING 192.168.0.3 (192.168.0.3) 56(84) bytes of data.
64 bytes from 192.168.0.3: icmp_seq=1 ttl=64 time=0.251 ms
64 bytes from 192.168.0.3: icmp_seq=2 ttl=64 time=0.170 ms

更多问题欢迎联系我: https://github.com/fanux

相关推荐