neutron之SDN简单测试

title: Neutron SDN 手动实现手册 date: 2017-04-13 23:37 tags: Network

本文旨在通过自己搭建类似neutron （openvswitch + gre）实现SDN 的环境，学习了解其工作原理，模拟核心原理，比如：同一租户自定义网络 instance 互通，手动为instance 分配 floating ip 等相关内容。

![image](E:公司资料github pagehexosource_posts文档图片手工SDN1.png)

###主机网卡配置

controller:
	     eth0:10.20.0.201   (management network)
	     eht1:172.16.0.201   (public/external network)
	     eht2:192.168.4.201  (private network，gre tunning)

compute01:
	     eth0:10.20.0.202   (management network)
	     eht1:(disabled)
	     eht2:192.168.4.202  (private network，gre tunning)

##模拟安装网络节点(Network1)

模拟Network 节点相关实现，比如L3、dhcp-agent实现，为了模拟多节点网络情况，这里Network同时也模拟一个计算节点，模拟M2 openvswitch 实现，上面运行instance1。

网络接口配置

vi /etc/sysconfig/network-scripts/ifcfg-eth0
	DEVICE=eth0
	TYPE=Ethernet
	ONBOOT=yes
	NM_CONTROLLED=yes
	BOOTPROTO=static
	IPADDR=10.20.0.201
	NETMASK=255.255.255.0

vi /etc/sysconfig/network-scripts/ifcfg-eth1
	DEVICE=eth1
	TYPE=Ethernet
	ONBOOT=yes
	NM_CONTROLLED=yes
	BOOTPROTO=static
	IPADDR=172.16.0.201
	NETMASK=255.255.255.0

vi /etc/sysconfig/network-scripts/ifcfg-eth2
	DEVICE=eth2
	TYPE=Ethernet
	ONBOOT=yes
	NM_CONTROLLED=yes
	BOOTPROTO=static
	IPADDR=192.168.4.201
	NETMASK=255.255.255.0

重启网络服务

service network restart

安装需要用到的包

yum install libvirt openvswitch python-virtinst xauth tigervnc qemu-* -y

移除默认的libvirt 网络，方便清晰分析网络情况

virsh net-destroy default
virsh net-autostart --disable default
virsh net-undefine default

设置允许ipforwarding

vi /etc/sysctl.conf 
	net.ipv4.ip_forward=1
	net.ipv4.conf.all.rp_filter=0
	net.ipv4.conf.default.rp_filter=0

立即生效

sysctl -p

启动openvswitch

service openvswitch start
chkconfig openvswitch on

创建一个linux bridge

brctl addbr qbr01
ip link set qbr01 up

创建一个instance，并连接到qbr01 Bridge，网络接口部分配置如下

<interface type='bridge'>
	      <source bridge='qbr01'/>
	      <target dev='tap01'/>
	      <model type='virtio'/>
	      <driver name='qemu'/>
	      <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/>
	</interface>

可以参考附件./gre/instance1.xml创建

cp ~/gre/ /var/tmp/
	cd /var/tmp/gre
	mv cirros-0.3.0-x86_64-disk.img instance1.img
	virsh define instance1.xml
	virsh start instance1
	virsh vncdisplay instance1
	vncviewer :0

启动console 以后,登录添加ip 地址 192.168.1.11

ip addr add 192.168.1.11/24 dev eth0
route add default gw 192.168.1.1

创建一个内部bridge br-int，模拟 OpenStack integrated bridge

ovs-vsctl add-br br-int
ovs-vsctl add-port br-int gre0 -- set interface gre0 type=gre options:remote_ip=192.168.4.202

创建一个veth peer，连接Linux Bridge 'qbr01' 和 OpenvSwich Bridge 'br-ini'

ip link add qvo01 type veth peer name qvb01
brctl addif qbr01 qvb01
ovs-vsctl add-port br-int qvo01
ovs-vsctl set port qvo01 tag=100
ip link set qvb01 up
ip link set qvo01 up

查看现在network1上的 br-int

ovs-vsctl show

##模拟安装计算节点(compute1)

##网络接口配置

vi /etc/sysconfig/network-scripts/ifcfg-eth0
	DEVICE=eth0
	TYPE=Ethernet
	ONBOOT=yes
	NM_CONTROLLED=yes
	BOOTPROTO=static
	IPADDR=10.20.0.202
	NETMASK=255.255.255.0

vi /etc/sysconfig/network-scripts/ifcfg-eth1
	DEVICE=eth1
	TYPE=Ethernet
	ONBOOT=yes
	NM_CONTROLLED=yes
	BOOTPROTO=static
	IPADDR=172.16.0.202
	NETMASK=255.255.255.0

vi /etc/sysconfig/network-scripts/ifcfg-eth2
	DEVICE=eth2
	TYPE=Ethernet
	ONBOOT=yes
	NM_CONTROLLED=yes
	BOOTPROTO=static
	IPADDR=192.168.4.202
	NETMASK=255.255.255.0

重启网络服务

service network restart

安装需要用到的包

yum install libvirt openvswitch python-virtinst xauth tigervnc  qemu-*

移除libvirt 默认的网络

virsh net-destroy default
virsh net-autostart --disableu default
virsh net-undefine default

设置允许ipforwarding

vi /etc/sysctl.conf 
    net.ipv4.ip_forward=1
    net.ipv4.conf.all.rp_filter=0
    net.ipv4.conf.default.rp_filter=0

立即生效

sysctl -p

启动openvswitch

service openvswitch start
chkconfig openvswitch on

创建一个linux bridge

brctl addbr qbr02
ip link set qbr02 up

创建一个vm，并连接到qbr02

<interface type='bridge'>
      <source bridge='qbr02'/>
      <target dev='tap02'/>
      <model type='virtio'/>
      <driver name='qemu'/>
      <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/>
</interface>

上gre目录到compute1 节点，可以参考附件./gre/instance2.xml创建

cp ~/gre/ /var/tmp/
cd /var/tmp/gre
mv cirros-0.3.0-x86_64-disk.img instance2.img
virsh define instance2.xml
virsh start instance2
virsh vncdisplay instance2
vncviewer :0

启动console 以后,登录添加ip得知 192.168.1.12

ip addr add 192.168.1.12/24 dev eth0
route add default gw 192.168.1.1

创建一个内部bridge br-int，模拟 OpenStack integrated bridge

ovs-vsctl add-br br-int
ovs-vsctl add-port br-int gre0 -- set interface gre0 type=gre options:remote_ip=192.168.4.201

创建一个veth peer，连接Linux Bridge 'qbr02' 和 OpenvSwich Bridge 'br-ini'

ip link add qvo02 type veth peer name qvb02
brctl addif qbr02 qvb02
ovs-vsctl add-port br-int qvo02
ovs-vsctl set port qvo02 tag=100
ip link set qvb02 up
ip link set qvo02 up

查看现在network1 上的 br-int

ovs-vsctl show

检查是否能连通instance1，在instance2的控制台

ping 192.168.1.11

##通过 Network Namespace 实现租户私有网络互访

添加一个namespace，dhcp01用于隔离租户网络。

ip netns add dhcp01

为私有网络192.168.1.0/24 ，在命名空间dhcp01 中创建dhcp 服务

ovs-vsctl add-port br-int tapdhcp01 -- set interface tapdhcp01 type=internal
ovs-vsctl set port tapdhcp01 tag=100
ip link set tapdhcp01 netns dhcp01
ip netns exec dhcp01 ip addr add 192.168.1.2/24 dev tapdhcp01
ip netns exec dhcp01 ip link set tapdhcp01 up

检查网络是否连通，在namespace 访问instance1 和 instance2

ip netns exec dhcp01 ping 192.168.1.12
ip netns exec dhcp01 ping 192.168.1.11

##通过 Network Namespace 和Iptables 实现L3 router

ovs-vsctl add-br br-ex

重新配置eth1 和 br-ex

vi /etc/sysconfig/network-scripts/ifcfg-eth1

	DEVICE=eth1
	ONBOOT=yes
	BOOTPROTO=none
	PROMISC=yes
	MTU=1546
	###################################
	DEVICE=ens160
    TYPE=OVSPort
    DEVICETYPE=ovs
    OVS_BRIDGE=br-ex
    ONBOOT=yes

	####################################

vi /etc/sysconfig/network-scripts/ifcfg-br-ex

	DEVICE=br-ex
	TYPE=Bridge
	ONBOOT=yes
	BOOTPROTO=none
	IPADDR0=172.16.0.201
	PREFIX0=24
	#######################################
	DEVICE=br-ex
    ONBOOT=yes
    DEVICETYPE=ovs
    TYPE=OVSBridge
    BOOTPROTO=static
    IPADDR=192.168.2.134
    NETMASK=255.255.255.0
    GATEWAY=192.168.2.1
    DNS1=218.2.2.2

	######################################

重启启动网络服务

ovs-vsctl add-port br-ex eth1 && service network restart

检查网络，配置后是否连通

ping 172.16.0.201

添加一个namespace，router01 用于路由和floating ip 分配

ip netns add router01

在br-int添加一个接口，作为私有网络192.168.1.0/24的网关

ovs-vsctl add-port br-int qr01 -- set interface qr01 type=internal
ovs-vsctl set port qr01 tag=100
ip link set qr01 netns router01
ip netns exec router01 ip addr add 192.168.1.1/24 dev qr01
ip netns exec router01 ip link set qr01 up
ip netns exec router01 ip link set lo up

在br-ex中添加一个接口，用于私网192.168.1.0/24设置下一跳地址

ovs-vsctl add-port br-ex qg01 -- set interface qg01  type=internal
ip link set qg01  netns router01
ip netns exec router01 ip addr add 172.16.0.100/24 dev qg01 
ip netns exec router01 ip link set qg01 up
ip netns exec router01 ip link set lo up

模拟分配floating ip 访问instance1

为instance1 192.168.1.11 分配floating ip，172.16.0.101

ip netns exec router01 ip addr add 172.16.0.101/32 dev qg01 
ip netns exec router01  iptables -t nat -A OUTPUT -d 172.16.0.101/32  -j DNAT --to-destination 192.168.1.11
ip netns exec router01  iptables -t nat -A PREROUTING -d 172.16.0.101/32 -j DNAT --to-destination 192.168.1.11
ip netns exec router01  iptables -t nat -A POSTROUTING -s 192.168.1.11/32 -j SNAT --to-source 172.16.0.101
ip netns exec router01  iptables -t nat -A POSTROUTING -s 192.168.1.0/24 -j SNAT --to-source 172.16.0.100

测试floating ip

ping 172.16.0.101

如果需要清除nat chain

iptables -t nat -F

ip netns exec router01 iptables -t nat -A OUTPUT -d 192.168.2.102/32 -j DNAT --to-destination 192.168.10.11
ip netns exec router01 iptables -t nat -A PREROUTING -d 192.168.2.102/32 -j DNAT --to-destination 192.168.10.11
ip netns exec router01 iptables -t nat -A POSTROUTING -s 192.168.10.11/32 -j SNAT --to-source 192.168.2.102

ip netns exec router01 ip addr add 192.168.2.103/32 dev qg01
ip netns exec router01 iptables -t nat -A OUTPUT -d 192.168.2.103/32 -j DNAT --to-destination 192.168.10.11
ip netns exec router01 iptables -t nat -A PREROUTING -d 192.168.2.103/32 -j DNAT --to-destination 192.168.10.11
ip netns exec router01 iptables -t nat -A POSTROUTING -s 192.168.10.11/32 -j SNAT --to-source 192.168.2.103

ip netns exec router01 route add default gw 192.168.2.1
ip netns exec router01 route -n