简体中文（ZH-CN）English（EN）繁体中文（ZH-TW）日本語（JA）한국어（KO）Melayu（MS）Français（FR）Deutsch（DE）
Home
Article
Q&A
Course
Topic
Download
Game
Dictionary
Recent Updates

Home  >  Article  >  Operation and Maintenance  >  A detailed explanation of Linux netfilter and VRF and their summary

A detailed explanation of Linux netfilter and VRF and their summary

藏色散人
藏色散人forward
2021-09-01 17:06:433694browse

The experimental environment is as shown below:
A detailed explanation of Linux netfilter and VRF and their summary

The configuration is as follows: 

#!/bin/bash
sudo ip netns add ns1 
sudo ip link add ns1veth1 type veth peer name eth0 netns ns1
sudo ip netns add ns2
sudo ip link add ns2veth1 type veth peer name eth0 netns ns2
sudo ip link set ns1veth1 master vrftest
sudo ip link set ns2veth1 master vrftest
sudo ip link set ns2veth1 up
sudo ip link set ns1veth1 up
sudo ip addr add 1.1.1.254/24 dev ns1veth1 
sudo ip addr add 2.2.2.254/24 dev ns2veth1 
sudo ip netns exec ns2 ip addr add 2.2.2.1/24 dev eth0 
sudo ip netns exec ns1 ip addr add 1.1.1.1/24 dev eth0 
sudo ip netns exec ns1 ip link set eth0 up
sudo ip netns exec ns1 ip link set lo up
sudo ip netns exec ns1 ip route add default via 1.1.1.254 dev eth0
sudo ip netns exec ns2 ip link set eth0 up
sudo ip netns exec ns2 ip link set lo up
sudo ip netns exec ns2 ip route add default via 2.2.2.254 dev eth0
sudo iptables -t mangle -A PREROUTING   -s 1.1.1.1 -j LOG --log-prefix="vrf-test-prerouting" 
sudo iptables -t mangle -A FORWARD      -s 1.1.1.1 -j LOG --log-prefix="vrf-test-forward" 
sudo iptables -t mangle -A POSTROUTING  -s 1.1.1.1 -j LOG --log-prefix="vrf-test-postrouting" 
sudo iptables -t mangle -A PREROUTING   -d 1.1.1.1 -j LOG --log-prefix="vrf-test-prerouting" 
sudo iptables -t mangle -A FORWARD      -d 1.1.1.1 -j LOG --log-prefix="vrf-test-forward" 
sudo iptables -t mangle -A POSTROUTING  -d 1.1.1.1 -j LOG --log-prefix="vrf-test-postrouting" 
sudo iptables -t mangle -A INPUT        -d 1.1.1.1 -j LOG --log-prefix="vrf-test-localin" 
sudo iptables -t mangle -A INPUT        -s 1.1.1.1 -j LOG --log-prefix="vrf-test-localin" 
sudo iptables -t mangle -A OUTPUT       -s 1.1.1.1 -j LOG --log-prefix="vrf-test-localout" 
sudo iptables -t mangle -A OUTPUT       -d 1.1.1.1 -j LOG --log-prefix="vrf-test-localout"

Access this machine from external network

#ns1 ping gateway 1.1.1.254

admin@ubuntu:~$ sudo ip netns exec ns1 ping 1.1.1.254 -c 1       
PING 1.1.1.254 (1.1.1.254) 56(84) bytes of data.
64 bytes from 1.1.1.254: icmp_seq=1 ttl=64 time=0.064 ms

--- 1.1.1.254 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.064/0.064/0.064/0.000 ms
admin@ubuntu:~$

View log

Nov 20 20:34:10 ubuntu kernel: [180403.527204] vrf-test-preroutingIN=ns1veth1 OUT= MAC=b2:f8:2a:13:31:75:6e:17:d5:b2:55:14:08:00 SRC=1.1.1.1 DST=1.1.1.254 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=32492 DF PROTO=ICMP TYPE=8 CODE=0 ID=33955 SEQ=1 
Nov 20 20:34:10 ubuntu kernel: [180403.527213] vrf-test-preroutingIN=vrftest OUT= MAC=b2:f8:2a:13:31:75:6e:17:d5:b2:55:14:08:00 SRC=1.1.1.1 DST=1.1.1.254 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=32492 DF PROTO=ICMP TYPE=8 CODE=0 ID=33955 SEQ=1 
Nov 20 20:34:10 ubuntu kernel: [180403.527220] vrf-test-localinIN=vrftest OUT= MAC=b2:f8:2a:13:31:75:6e:17:d5:b2:55:14:08:00 SRC=1.1.1.1 DST=1.1.1.254 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=32492 DF PROTO=ICMP TYPE=8 CODE=0 ID=33955 SEQ=1 
Nov 20 20:34:10 ubuntu kernel: [180403.527231] vrf-test-localoutIN= OUT=vrftest SRC=1.1.1.254 DST=1.1.1.1 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=54845 PROTO=ICMP TYPE=0 CODE=0 ID=33955 SEQ=1 
Nov 20 20:34:10 ubuntu kernel: [180403.527233] vrf-test-postroutingIN= OUT=vrftest SRC=1.1.1.254 DST=1.1.1.1 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=54845 PROTO=ICMP TYPE=0 CODE=0 ID=33955 SEQ=1 
Nov 20 20:34:10 ubuntu kernel: [180403.527235] vrf-test-localoutIN= OUT=ns1veth1 SRC=1.1.1.254 DST=1.1.1.1 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=54845 PROTO=ICMP TYPE=0 CODE=0 ID=33955 SEQ=1 
Nov 20 20:34:10 ubuntu kernel: [180403.527242] vrf-test-postroutingIN= OUT=ns1veth1 SRC=1.1.1.254 DST=1.1.1.1 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=54845 PROTO=ICMP TYPE=0 CODE=0 ID=33955 SEQ=1

You can see it from the log Output:

The sequence of hook points that the request message passes through is as follows:

Serial number hook point Input interface Output interface
1 PREROUTING ns1veth1 None
2 PREROUTING vrftest None
3 INPUT vrftest None

The sequence of hook points that the response message passes through is as follows:

##2POSTROUTINGNonevrftest3OUTPUTNonens1veth14POSTROUTINGNonens1veth1
Serial number hook point Input interface Output interface
1 OUTPUT None vrftest
Forward message

ns1 ping ns2

admin@ubuntu:~$ sudo ip netns exec ns1 ping 2.2.2.1 -c 1
PING 2.2.2.1 (2.2.2.1) 56(84) bytes of data.
64 bytes from 2.2.2.1: icmp_seq=1 ttl=63 time=0.063 ms

--- 2.2.2.1 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.063/0.063/0.063/0.000 ms
admin@ubuntu:~$

Check the log

Nov 20 20:28:31 ubuntu kernel: [180065.076713] vrf-test-preroutingIN=ns1veth1 OUT= MAC=b2:f8:2a:13:31:75:6e:17:d5:b2:55:14:08:00 SRC=1.1.1.1 DST=2.2.2.1 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=38312 DF PROTO=ICMP TYPE=8 CODE=0 ID=33948 SEQ=1 
Nov 20 20:28:31 ubuntu kernel: [180065.076722] vrf-test-preroutingIN=vrftest OUT= MAC=b2:f8:2a:13:31:75:6e:17:d5:b2:55:14:08:00 SRC=1.1.1.1 DST=2.2.2.1 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=38312 DF PROTO=ICMP TYPE=8 CODE=0 ID=33948 SEQ=1 
Nov 20 20:28:31 ubuntu kernel: [180065.076730] vrf-test-forwardIN=vrftest OUT=ns2veth1 MAC=b2:f8:2a:13:31:75:6e:17:d5:b2:55:14:08:00 SRC=1.1.1.1 DST=2.2.2.1 LEN=84 TOS=0x00 PREC=0x00 TTL=63 ID=38312 DF PROTO=ICMP TYPE=8 CODE=0 ID=33948 SEQ=1 
Nov 20 20:28:31 ubuntu kernel: [180065.076732] vrf-test-postroutingIN= OUT=ns2veth1 SRC=1.1.1.1 DST=2.2.2.1 LEN=84 TOS=0x00 PREC=0x00 TTL=63 ID=38312 DF PROTO=ICMP TYPE=8 CODE=0 ID=33948 SEQ=1 
Nov 20 20:28:31 ubuntu kernel: [180065.076746] vrf-test-preroutingIN=ns2veth1 OUT= MAC=02:25:0e:fe:52:35:ba:19:4d:37:ac:8b:08:00 SRC=2.2.2.1 DST=1.1.1.1 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=47601 PROTO=ICMP TYPE=0 CODE=0 ID=33948 SEQ=1 
Nov 20 20:28:31 ubuntu kernel: [180065.076749] vrf-test-preroutingIN=vrftest OUT= MAC=02:25:0e:fe:52:35:ba:19:4d:37:ac:8b:08:00 SRC=2.2.2.1 DST=1.1.1.1 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=47601 PROTO=ICMP TYPE=0 CODE=0 ID=33948 SEQ=1 
Nov 20 20:28:31 ubuntu kernel: [180065.076752] vrf-test-forwardIN=vrftest OUT=ns1veth1 MAC=02:25:0e:fe:52:35:ba:19:4d:37:ac:8b:08:00 SRC=2.2.2.1 DST=1.1.1.1 LEN=84 TOS=0x00 PREC=0x00 TTL=63 ID=47601 PROTO=ICMP TYPE=0 CODE=0 ID=33948 SEQ=1 
Nov 20 20:28:31 ubuntu kernel: [180065.076753] vrf-test-postroutingIN= OUT=ns1veth1 SRC=2.2.2.1 DST=1.1.1.1 LEN=84 TOS=0x00 PREC=0x00 TTL=63 ID=47601 PROTO=ICMP TYPE=0 CODE=0 ID=33948 SEQ=1
It can be seen from the log:

The sequence of hook points passed by the request message is as follows:

Serial numberhook pointInput interfaceOutput interface1PREROUTINGns1veth1None2PREROUTING vrftestNone3FORWARDvrftestns2veth14POSTROUTINGNonens2veth1
Response message The sequence of hook points passed is as follows:

Serial numberhook pointInput interfaceOutput interface1PREROUTINGns2veth12PREROUTINGvrftest##34This machine accesses the external network


FORWARD vrftest ns2veth1
POSTROUTING None ns2veth1

vrftest ping 1.1.1.1

admin@ubuntu:~$ sudo ping 1.1.1.1 -I vrftest -c 1
ping: Warning: source address might be selected on device other than vrftest.
PING 1.1.1.1 (1.1.1.1) from 1.1.1.254 vrftest: 56(84) bytes of data.
64 bytes from 1.1.1.1: icmp_seq=1 ttl=64 time=0.039 ms

--- 1.1.1.1 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.039/0.039/0.039/0.000 ms
admin@ubuntu:~$

View log

Nov 20 21:21:11 ubuntu kernel: [183224.956734] vrf-test-localoutIN= OUT=vrftest SRC=1.1.1.254 DST=1.1.1.1 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=35042 DF PROTO=ICMP TYPE=8 CODE=0 ID=34186 SEQ=1 
Nov 20 21:21:11 ubuntu kernel: [183224.956740] vrf-test-postroutingIN= OUT=vrftest SRC=1.1.1.254 DST=1.1.1.1 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=35042 DF PROTO=ICMP TYPE=8 CODE=0 ID=34186 SEQ=1 
Nov 20 21:21:11 ubuntu kernel: [183224.956745] vrf-test-localoutIN= OUT=ns1veth1 SRC=1.1.1.254 DST=1.1.1.1 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=35042 DF PROTO=ICMP TYPE=8 CODE=0 ID=34186 SEQ=1 
Nov 20 21:21:11 ubuntu kernel: [183224.956746] vrf-test-postroutingIN= OUT=ns1veth1 SRC=1.1.1.254 DST=1.1.1.1 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=35042 DF PROTO=ICMP TYPE=8 CODE=0 ID=34186 SEQ=1 
Nov 20 21:21:11 ubuntu kernel: [183224.956762] vrf-test-preroutingIN=ns1veth1 OUT= MAC=b2:f8:2a:13:31:75:6e:17:d5:b2:55:14:08:00 SRC=1.1.1.1 DST=1.1.1.254 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=323 PROTO=ICMP TYPE=0 CODE=0 ID=34186 SEQ=1 
Nov 20 21:21:11 ubuntu kernel: [183224.956765] vrf-test-preroutingIN=vrftest OUT= MAC=b2:f8:2a:13:31:75:6e:17:d5:b2:55:14:08:00 SRC=1.1.1.1 DST=1.1.1.254 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=323 PROTO=ICMP TYPE=0 CODE=0 ID=34186 SEQ=1 
Nov 20 21:21:11 ubuntu kernel: [183224.956769] vrf-test-localinIN=vrftest OUT= MAC=b2:f8:2a:13:31:75:6e:17:d5:b2:55:14:08:00 SRC=1.1.1.1 DST=1.1.1.254 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=323 PROTO=ICMP TYPE=0 CODE=0 ID=34186 SEQ=1
It can be seen from the log:

The sequence of hook points that the request message passes through is as follows:

Serial number12 34

应答报文经过的hook点顺序如下：

hook point Input interface Output interface
OUTPUT None vrftest
POSTROUTING None vrftest
OUTPUT None ns1veth1
POSTROUTING None ns1veth1
序号 hook点 输入接口 输出接口
1 PREROUTING ns1veth1 无
2 PREROUTING vrftest 无
3 INPUT vrftest 无

本机访问本机，即环回

vrftest ping vrf接口地址9.9.9.9/24

添加如下netfilter规则：

admin@ubuntu:~/vrftcpdump$ cat netfilter1.sh 
sudo iptables -t mangle -A PREROUTING   -s 9.9.9.9 -j LOG --log-prefix="vrf-test-prerouting" 
sudo iptables -t mangle -A FORWARD      -s 9.9.9.9 -j LOG --log-prefix="vrf-test-forward" 
sudo iptables -t mangle -A POSTROUTING  -s 9.9.9.9 -j LOG --log-prefix="vrf-test-postrouting"
sudo iptables -t mangle -A INPUT        -s 9.9.9.9 -j LOG --log-prefix="vrf-test-localin" 
sudo iptables -t mangle -A OUTPUT       -s 9.9.9.9 -j LOG --log-prefix="vrf-test-localout" 
admin@ubuntu:~/vrftcpdump$

配置vrftest接口地址为9.9.9.9/24

admin@ubuntu:~/vrftcpdump$ sudo ip addr add 9.9.9.9/24 dev vrftest           
admin@ubuntu:~/vrftcpdump$ sudo ping 9.9.9.9 -I vrftest -c 1             
PING 9.9.9.9 (9.9.9.9) from 9.9.9.9 vrftest: 56(84) bytes of data.
64 bytes from 9.9.9.9: icmp_seq=1 ttl=64 time=0.050 ms

--- 9.9.9.9 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.050/0.050/0.050/0.000 ms
admin@ubuntu:~/vrftcpdump$

查看log

Nov 20 22:13:41 ubuntu kernel: [186374.589186] vrf-test-localoutIN= OUT=vrftest SRC=9.9.9.9 DST=9.9.9.9 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=39425 DF PROTO=ICMP TYPE=8 CODE=0 ID=34815 SEQ=1 
Nov 20 22:13:41 ubuntu kernel: [186374.589192] vrf-test-postroutingIN= OUT=vrftest SRC=9.9.9.9 DST=9.9.9.9 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=39425 DF PROTO=ICMP TYPE=8 CODE=0 ID=34815 SEQ=1 
Nov 20 22:13:41 ubuntu kernel: [186374.589202] vrf-test-preroutingIN=vrftest OUT= MAC=ca:f9:f0:37:4c:6c:ca:f9:f0:37:4c:6c:08:00 SRC=9.9.9.9 DST=9.9.9.9 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=39425 DF PROTO=ICMP TYPE=8 CODE=0 ID=34815 SEQ=1 
Nov 20 22:13:41 ubuntu kernel: [186374.589204] vrf-test-localinIN=vrftest OUT= MAC=ca:f9:f0:37:4c:6c:ca:f9:f0:37:4c:6c:08:00 SRC=9.9.9.9 DST=9.9.9.9 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=39425 DF PROTO=ICMP TYPE=8 CODE=0 ID=34815 SEQ=1 
Nov 20 22:13:41 ubuntu kernel: [186374.589210] vrf-test-localoutIN= OUT=vrftest SRC=9.9.9.9 DST=9.9.9.9 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=39426 PROTO=ICMP TYPE=0 CODE=0 ID=34815 SEQ=1 
Nov 20 22:13:41 ubuntu kernel: [186374.589211] vrf-test-postroutingIN= OUT=vrftest SRC=9.9.9.9 DST=9.9.9.9 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=39426 PROTO=ICMP TYPE=0 CODE=0 ID=34815 SEQ=1 
Nov 20 22:13:41 ubuntu kernel: [186374.589215] vrf-test-preroutingIN=vrftest OUT= MAC=ca:f9:f0:37:4c:6c:ca:f9:f0:37:4c:6c:08:00 SRC=9.9.9.9 DST=9.9.9.9 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=39426 PROTO=ICMP TYPE=0 CODE=0 ID=34815 SEQ=1 
Nov 20 22:13:41 ubuntu kernel: [186374.589217] vrf-test-localinIN=vrftest OUT= MAC=ca:f9:f0:37:4c:6c:ca:f9:f0:37:4c:6c:08:00 SRC=9.9.9.9 DST=9.9.9.9 LEN=84 TOS=0x00 PREC=0x00 TTL=64 ID=39426 PROTO=ICMP TYPE=0 CODE=0 ID=34815 SEQ=1

从log可以看出：

请求报文经过的hook点顺序如下：

序号 hook点 输入接口 输出接口
1 OUTPUT 无 vrftest
2 POSTROUTING 无 vrftest
3 PREROUTING vrftest 无
4 INPUT vrftest 无

应答报文经过的hook点顺序如下：

序号 hook点 输入接口 输出接口
1 OUTPUT 无 vrftest
2 POSTROUTING 无 vrftest
3 PREROUTING vrftest 无
4 INPUT vrftest 无

环回请求和应答是一样的。

总结

  • linux内核支持VRF后，报文会两次进入PREROUTING节点，第一次是原始接口，第二次是原始接口依附的vrf主接口。
  • linux内核支持VRF后，本机输出的报文，先以vrf接口作为输出接口经过OUTPUT和POSTROUTING节点，然后再以真实出去的接口经过OUPUT和POSTROUTING节点。
  • 本机报文进行环回时，与没有VRF的环境是一样的。

推荐学习：《linux视频教程》

The above is the detailed content of A detailed explanation of Linux netfilter and VRF and their summary. For more information, please follow other related articles on the PHP Chinese website!

接口 linux
Statement：
This article is reproduced at:segmentfault.com. If there is any infringement, please contact admin@php.cn delete
Previous article：Learn about case sensitivity issues in Linux and MacOS terminals in simple terms (with code)Next article：Learn about case sensitivity issues in Linux and MacOS terminals in simple terms (with code)

Related articles

See more