前言
如果想隔离主机中进程的网络环境，可以使用network namespace（后面简称ns）来做。隔离方法很多种，本文介绍几种简单的可行方案。
实验环境准备
实验环境使用腾讯云上的两台cvm主机，两台机器在同一vpc同一子网下（underlay网络），这样两台机器天然内网互通。
cvm-1：
ubuntu@VM-6-43-ubuntu:~$ ip a
1: lo: mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
3: eth0: mtu 1500 qdisc fq_codel state UP group default qlen 1000
link/ether 52:54:00:9c:e7:91 brd ff:ff:ff:ff:ff:ff
inet 10.0.6.43/24 brd 10.0.6.255 scope global eth0
valid_lft forever preferred_lft forever
inet6 fe80::5054:ff:fe9c:e791/64 scope link
valid_lft forever preferred_lft forever
cvm-2：
ubuntu@VM-6-46-ubuntu:~$ ip a
1: lo: mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
3: eth0: mtu 1500 qdisc fq_codel state UP group default qlen 1000
link/ether 52:54:00:75:e7:ca brd ff:ff:ff:ff:ff:ff
inet 10.0.6.46/24 brd 10.0.6.255 scope global eth0
valid_lft forever preferred_lft forever
inet6 fe80::5054:ff:fe75:e7ca/64 scope link
valid_lft forever preferred_lft forever
我们之后的配置都在cvm-1上操作进行，使用cvm-2测试网络连通性。
方案详解
fafl.png
双网卡
方案概览：
2eth.png
设置方法：
#!/bin/bash
myns=blue
ip net add $myns
ip link set eth1 netns $myns
ip net exec $myns ip link set eth1 name eth0
ip net exec $myns ip addr add 10.0.6.14/24 dev eth0
ip net exec $myns ip link set eth0 up
ip net exec $myns ip route add default via 10.0.6.1 dev eth0
设置后机器上的网络情况为：
root@VM-6-43-ubuntu:/home/ubuntu# ip a
1: lo: mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
3: eth0: mtu 1500 qdisc fq_codel state UP group default qlen 1000
link/ether 52:54:00:9c:e7:91 brd ff:ff:ff:ff:ff:ff
inet 10.0.6.43/24 brd 10.0.6.255 scope global eth0
valid_lft forever preferred_lft forever
inet6 fe80::5054:ff:fe9c:e791/64 scope link
valid_lft forever preferred_lft forever
root@VM-6-43-ubuntu:/home/ubuntu# ip net exec blue ip a
1: lo: mtu 65536 qdisc noop state DOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
4: eth0: mtu 1500 qdisc fq_codel state UP group default qlen 1000
link/ether 20:90:6f:a4:6b:7e brd ff:ff:ff:ff:ff:ff
inet 10.0.6.14/24 scope global eth0
valid_lft forever preferred_lft forever
inet6 fe80::2290:6fff:fea4:6b7e/64 scope link
valid_lft forever preferred_lft forever
和另一台cvm测试网络连通性：
//隔离环境连通另外一台cvm
root@VM-6-43-ubuntu:/home/ubuntu# ip net exec blue ping 10.0.6.46 -c 3PING 10.0.6.46 (10.0.6.46) 56(84) bytes of data.
64 bytes from 10.0.6.46: icmp_seq=1 ttl=64 time=0.145 ms
64 bytes from 10.0.6.46: icmp_seq=2 ttl=64 time=0.168 ms
64 bytes from 10.0.6.46: icmp_seq=3 ttl=64 time=0.204 ms
--- 10.0.6.46 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2037ms
rtt min/avg/max/mdev = 0.145/0.172/0.204/0.026 ms
//另一台cvm连通隔离环境
ubuntu@VM-6-46-ubuntu:~$ ping 10.0.6.14 -c 3
PING 10.0.6.14 (10.0.6.14) 56(84) bytes of data.
64 bytes from 10.0.6.14: icmp_seq=1 ttl=64 time=0.172 ms
64 bytes from 10.0.6.14: icmp_seq=2 ttl=64 time=0.178 ms
64 bytes from 10.0.6.14: icmp_seq=3 ttl=64 time=0.174 ms
--- 10.0.6.14 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2040ms
rtt min/avg/max/mdev = 0.172/0.174/0.178/0.015 ms
单网卡
双IP
方案概览：
1eth2ip.png
本方案创建一个虚拟网卡，使用ipvlan的L2模式，设置方法如下：
#!/bin/bash
myns=blue
tmpName=ipv1
ip net add $myns
ip link add link eth0 name $tmpName type ipvlan mode l2
ip link set $tmpName netns $myns
ip net exec $myns ip link set $tmpName name eth0
ip net exec $myns ip addr add 10.0.6.22/24 dev eth0
ip net exec $myns ip link set eth0 up
ip net exec $myns ip route add default via 10.0.6.1 dev eth0
设置后隔离环境网络情况为：
root@VM-6-43-ubuntu:/home/ubuntu# ip net exec blue ip a
1: lo: mtu 65536 qdisc noop state DOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
4: eth0@if3: mtu 1500 qdisc noqueue state UNKNOWN group default qlen 1000
link/ether 52:54:00:9c:e7:91 brd ff:ff:ff:ff:ff:ff
inet 10.0.6.22/24 scope global eth0
valid_lft forever preferred_lft forever
inet6 fe80::5254:0:19c:e791/64 scope link
valid_lft forever preferred_lft forever
和另一台cvm测试网络连通性：
//隔离环境连通另外一台cvm
root@VM-6-43-ubuntu:/home/ubuntu# ip net exec blue ping 10.0.6.46 -c 3PING 10.0.6.46 (10.0.6.46) 56(84) bytes of data.
64 bytes from 10.0.6.46: icmp_seq=1 ttl=64 time=0.297 ms
64 bytes from 10.0.6.46: icmp_seq=2 ttl=64 time=0.204 ms
64 bytes from 10.0.6.46: icmp_seq=3 ttl=64 time=0.183 ms
--- 10.0.6.46 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2033ms
rtt min/avg/max/mdev = 0.183/0.228/0.297/0.049 ms
//另一台cvm连通隔离环境
ubuntu@VM-6-46-ubuntu:~$ ping 10.0.6.22 -c 3
PING 10.0.6.22 (10.0.6.22) 56(84) bytes of data.
64 bytes from 10.0.6.22: icmp_seq=1 ttl=64 time=0.164 ms
64 bytes from 10.0.6.22: icmp_seq=2 ttl=64 time=0.169 ms
64 bytes from 10.0.6.22: icmp_seq=3 ttl=64 time=0.175 ms
--- 10.0.6.22 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2039ms
rtt min/avg/max/mdev = 0.164/0.169/0.175/0.011 ms
单IP -- 隔离网络使用fake ip
方案概览：
1eth1.png
设置方法如下：
#!/bin/bash
myns=blue
ip net add $myns
ip link add veth0 type veth peer name veth_host
ip link set veth0 netns $myns
ip net exec $myns ip link set veth0 name eth0
ip net exec $myns ip addr add 169.250.1.2/24 dev eth0
ip net exec $myns ip link set eth0 up
ip net exec $myns ip route add default via 169.250.1.1 dev eth0
ip link set veth_host up
ip addr add 169.250.1.1/24 dev veth_host
iptables -t nat -A POSTROUTING -s 169.250.1.2 -o eth0 -j MASQUERADE
sysctl -w net.ipv4.ip_forward=1
设置后主机环境网络情况为：
root@VM-6-43-ubuntu:/home/ubuntu# ip a
1: lo: mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
3: eth0: mtu 1500 qdisc fq_codel state UP group default qlen 1000
link/ether 52:54:00:9c:e7:91 brd ff:ff:ff:ff:ff:ff
inet 10.0.6.43/24 brd 10.0.6.255 scope global eth0
valid_lft forever preferred_lft forever
inet6 fe80::5054:ff:fe9c:e791/64 scope link
valid_lft forever preferred_lft forever
4: veth_host@if5: mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 9a:c1:6d:de:de:b7 brd ff:ff:ff:ff:ff:ff link-netnsid 0
inet 169.250.1.1/24 scope global veth_host
valid_lft forever preferred_lft forever
inet6 fe80::98c1:6dff:fede:deb7/64 scope link
valid_lft forever preferred_lft forever
设置后隔离环境网络情况为：
root@VM-6-43-ubuntu:/home/ubuntu# ip net exec blue ip a
1: lo: mtu 65536 qdisc noop state DOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
5: eth0@if4: mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 16:8d:05:d9:34:44 brd ff:ff:ff:ff:ff:ff link-netnsid 0
inet 169.250.1.2/24 scope global eth0
valid_lft forever preferred_lft forever
inet6 fe80::148d:5ff:fed9:3444/64 scope link
valid_lft forever preferred_lft forever
和另一台cvm测试网络连通性：
//隔离环境连通另外一台cvm
root@VM-6-43-ubuntu:/home/ubuntu# ip net exec blue ping 10.0.6.46 -c 3PING 10.0.6.46 (10.0.6.46) 56(84) bytes of data.
64 bytes from 10.0.6.46: icmp_seq=1 ttl=63 time=0.323 ms
64 bytes from 10.0.6.46: icmp_seq=2 ttl=63 time=0.201 ms
64 bytes from 10.0.6.46: icmp_seq=3 ttl=63 time=0.238 ms
--- 10.0.6.46 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2036ms
rtt min/avg/max/mdev = 0.201/0.254/0.323/0.051 ms
本方案从外部cvm访问隔离网络中服务需要配置DNAT。
单IP -- 主机网络使用fake ip
方案概览：
1eth2.png
设置方法如下：
#!/bin/bash
myns=blue
ip net add $myns
ip link add veth0 type veth peer name veth_host
ip link set veth0 netns $myns
ip link set eth0 netns $myns
ip net exec $myns ip addr add 169.250.1.2/24 dev veth0
ip net exec $myns ip link set veth0 up
ip net exec $myns ip addr add 10.0.6.43/24 dev eth0
ip net exec $myns ip link set eth0 up
ip net exec $myns ip route add default via 10.0.6.1 dev eth0
ip link set veth_host up
ip addr add 169.250.1.1/24 dev veth_host
ip route add default via 169.250.1.2 dev veth_host
ip net exec $myns iptables -t nat -A POSTROUTING -s 169.250.1.1 -o eth0 -j MASQUERADE
ip net exec $myns sysctl -w net.ipv4.ip_forward=1
这里设置后之前通过ssh登录到主机环境会断开无法登录，需要使用cvm提供的vnc方式登录。
设置后主机环境网络情况为：
zj.png
设置后隔离环境网络情况为：
gl.png
和另一台cvm测试网络连通性：
隔离环境连通另外一台cvm：
ping.png
//另一台cvm连通隔离环境
ubuntu@VM-6-46-ubuntu:~$ ping 10.0.6.43 -c 3
PING 10.0.6.43 (10.0.6.43) 56(84) bytes of data.
64 bytes from 10.0.6.43: icmp_seq=1 ttl=64 time=0.188 ms
64 bytes from 10.0.6.43: icmp_seq=2 ttl=64 time=0.190 ms
64 bytes from 10.0.6.43: icmp_seq=3 ttl=64 time=0.183 ms
--- 10.0.6.43 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2049ms
rtt min/avg/max/mdev = 0.183/0.187/0.190/0.003 ms
本方案从外部cvm访问主机网络中服务需要配置DNAT。
写在后面
网络隔离通常用于容器服务，上面是自己基于underlay网络做的一些实践，如有错误，还请指出，感谢阅读！