因为在OpenStack的Neutron中比较常用，所以参考别人的博客试了下OVS的一些隧道封装功能（GRE，VXLAN）。
实验:实现两个host的Network namespace之间的通信，NS可以理解为host里的VM（VM通过NS隔离，参考资料里涉及到了VM，KVM还不熟悉，这里就略掉了），如下图所示：
配置host 1
在host1中创建一个network namespace
red
，其中veth0添加到red里，配置IP并且up起来。
$ sudo ip netns add red
$ sudo ip link add veth0 type veth peer name veth1
$ sudo ip link set veth0 netns red
$ sudo ip netns exec red ip addr add 10.1.1.1/24 dev veth0
$ sudo ip netns exec red ip a
7: lo: mtu 16436 qdisc noop state DOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
9: veth0: mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether ca:4f:8c:a1:3b:24 brd ff:ff:ff:ff:ff:ff
inet 10.1.1.1/24 scope global veth0
inet6 fe80::c84f:8cff:fea1:3b24/64 scope link
valid_lft forever preferred_lft forever
OVS创建一个bridge br-int，并且把前面创建的veth对里的veth1作为端口添加到这个bridge上并up起来，然后给端口打上vlan tag 10。
$ sudo ovs-vsctl add-br br-int
$ sudo ovs-vsctl add-port br-int veth1
$ sudo ip link set veth1 up
$ sudo ovs-vsctl set port veth1 tag=10
配置host2
host2的配置和host1几乎完全相同，唯一不同的是veth0的IP地址时10.1.1.2/24
$ sudo ip netns exec red ip a
9: lo: mtu 16436 qdisc noop state DOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
11: veth0: mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 12:09:83:aa:97:57 brd ff:ff:ff:ff:ff:ff
inet 10.1.1.2/24 scope global veth0
inet6 fe80::1009:83ff:feaa:9757/64 scope link
valid_lft forever preferred_lft forever
配置host1和host2之间的GRE隧道
在host1上配置
$ sudo ovs-vsctl add-port br-int gre0 -- set interface gre0 type=gre options:remote_ip=192.168.4.202
其中192.168.4.202是host2上的eth2
在host2上配置
$ sudo ovs-vsctl add-port br-int gre0 -- set interface gre0 type=gre options:remote_ip=192.168.4.201
其中192.168.4.201是host1的eth2地址。
可以通过OVS查看配置
$ sudo ovs-vsctl show
d33f919d-82b7-4541-a4c2-39ca45286a83
Bridge br-int
Port "veth1"
tag: 10
Interface "veth1"
Port br-int
Interface br-int
type: internal
Port "gre0"
Interface "gre0"
type: gre
options: {remote_ip="192.168.4.201"}
ovs_version: "1.11.0"
验证两个host上的red network namespace是否连通
在host1上ping
$ ip netns exec red ping -c 4 10.1.1.2
PING 10.1.1.2 (10.1.1.2) 56(84) bytes of data.
64 bytes from 10.1.1.2: icmp_seq=1 ttl=64 time=0.977 ms
64 bytes from 10.1.1.2: icmp_seq=2 ttl=64 time=0.915 ms
64 bytes from 10.1.1.2: icmp_seq=3 ttl=64 time=1.00 ms
64 bytes from 10.1.1.2: icmp_seq=4 ttl=64 time=0.658 ms
--- 10.1.1.2 ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 3005ms
rtt min/avg/max/mdev = 0.658/0.889/1.008/0.140 ms
同时在host2的eth2上抓包可以看到
$ sudo tcpdump -i eth2
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on eth2, link-type EN10MB (Ethernet), capture size 65535 bytes
16:23:59.597256 IP 192.168.4.201 > 192.168.4.202: GREv0, length 106: IP 10.1.1.1 > 10.1.1.2: ICMP echo request, id 13321, seq 1, length 64
16:23:59.597499 IP 192.168.4.202 > 192.168.4.201: GREv0, length 106: IP 10.1.1.2 > 10.1.1.1: ICMP echo reply, id 13321, seq 1, length 64
16:24:00.598288 IP 192.168.4.201 > 192.168.4.202: GREv0, length 106: IP 10.1.1.1 > 10.1.1.2: ICMP echo request, id 13321, seq 2, length 64
16:24:00.598461 IP 192.168.4.202 > 192.168.4.201: GREv0, length 106: IP 10.1.1.2 > 10.1.1.1: ICMP echo reply, id 13321, seq 2, length 64
16:24:01.600200 IP 192.168.4.201 > 192.168.4.202: GREv0, length 106: IP 10.1.1.1 > 10.1.1.2: ICMP echo request, id 13321, seq 3, length 64
16:24:01.600441 IP 192.168.4.202 > 192.168.4.201: GREv0, length 106: IP 10.1.1.2 > 10.1.1.1: ICMP echo reply, id 13321, seq 3, length 64
16:24:02.133695 IP 192.168.4.1.17500 > 192.168.4.255.17500: UDP, length 104
16:24:02.601673 IP 192.168.4.201 > 192.168.4.202: GREv0, length 106: IP 10.1.1.1 > 10.1.1.2: ICMP echo request, id 13321, seq 4, length 64
16:24:02.601833 IP 192.168.4.202 > 192.168.4.201: GREv0, length 106: IP 10.1.1.2 > 10.1.1.1: ICMP echo reply, id 13321, seq 4, length 64
可以看到有ICMP的包通过GRE的封装发送过来。
配置host1和host2之间VXLAN封装
首先在host1和host2上把ovs上的gre0 port 删掉
$ sudo ovs-vsctl del-port gre0
然后在host1和host2上分别配置vxlan，和GRE的区别就是type字段改成vxlan
$ sudo ovs-vsctl add-port br-int vxlan0 -- set interface vxlan0 type=vxlan options:remote_ip=192.168.4.201
$ sudo ovs-vsctl show
d33f919d-82b7-4541-a4c2-39ca45286a83
Bridge br-int
Port "veth1"
tag: 10
Interface "veth1"
Port br-int
Interface br-int
type: internal
Port "vxlan0"
Interface "vxlan0"
type: vxlan
options: {remote_ip="192.168.4.201"}
ovs_version: "1.11.0"
验证
tcpdump -i eth2
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on eth2, link-type EN10MB (Ethernet), capture size 65535 bytes
16:38:40.018375 ARP, Request who-has 192.168.4.202 tell 192.168.4.201, length 46
16:38:40.018422 ARP, Reply 192.168.4.202 is-at 08:00:27:fd:39:c8 (oui Unknown), length 28
16:38:40.018724 IP 192.168.4.201.60827 > 192.168.4.202.4789: UDP, length 54
16:38:40.019324 IP 192.168.4.202.56167 > 192.168.4.201.4789: UDP, length 54
16:38:40.020405 IP 192.168.4.201.57614 > 192.168.4.202.4789: UDP, length 110
16:38:40.021146 IP 192.168.4.202.39657 > 192.168.4.201.4789: UDP, length 110
16:38:41.018502 IP 192.168.4.201.57614 > 192.168.4.202.4789: UDP, length 110
16:38:41.018674 IP 192.168.4.202.39657 > 192.168.4.201.4789: UDP, length 110
16:38:42.019908 IP 192.168.4.201.57614 > 192.168.4.202.4789: UDP, length 110
16:38:42.020198 IP 192.168.4.202.39657 > 192.168.4.201.4789: UDP, length 110
16:38:43.021400 IP 192.168.4.201.57614 > 192.168.4.202.4789: UDP, length 110
16:38:43.021597 IP 192.168.4.202.39657 > 192.168.4.201.4789: UDP, length 110
16:38:45.018441 ARP, Request who-has 192.168.4.201 tell 192.168.4.202, length 28
16:38:45.019091 ARP, Reply 192.168.4.201 is-at 08:00:27:f4:b2:82 (oui Unknown), length 46
16:38:45.020773 IP 192.168.4.202.56167 > 192.168.4.201.4789: UDP, length 54
16:38:45.021771 IP 192.168.4.201.55994 > 192.168.4.202.4789: UDP, length 54
端口4789就是vxlan的端口。
参考：
http://blog.scottlowe.org/2013/09/09/namespaces-vlans-open-vswitch-and-gre-tunnels/
http://blog.scottlowe.org/2013/05/07/using-gre-tunnels-with-open-vswitch/