In article #005 you learned how RARE/freeRouter is controlling a P4Emu/pcap dataplane. We also demonstrated that this setup could be integrated into real networks.
Requirement- Basic Linux/Unix knowledge
- Basic networking knowledge
| |
Overview
Though P4Emu/pcap can be used for SOHO and can handle nx1GE of traffic, this comes at a high CPU load cost and thus a higher power consumption.
Note |
---|
"Why write yet another software dataplane as freeRouter has already a working native software dataplane ?" The partial answer to the question raised in the previous article was: "decoupling control plane from the dataplane" We learned that P4Emu: - is able to understand the VERY same strict control message from freeRouter as it occurs with a P4 dataplane
- is able to switch packet emulating router.p4 using libpcap packet forwarding backend.
However, even though libpcap is a performant packet processing library, the kernel is still heavily sollicited and the higher the traffic rate is, the higher CPU workload becomes. |
Article objective
In this article we'll using freeRouter setup deployed in #005 and replace P4Emu/pcap's dataplane and replace it by P4Emu/dpdk dataplane.
Info |
---|
title | Source Wikipedia: https://en.wikipedia.org/wiki/Data_Plane_Development_Kit |
---|
|
The Data Plane Development Kit (DPDK) is an Open source software project managed by the Linux Foundation. It provides a set of data plane libraries and network interface controller polling-mode drivers for offloading TCP packet processing from the operating system kernel to processes running in user space. This offloading achieves higher computing efficiency and higher packet throughput than is possible using the interrupt-driven processing provided in the kernel. |
...
Note |
---|
It is important to note that though its name implies, P4Emu/dpdk is not emulating V1Model. P4Emu is emulating router.p4 packet processing logic and uses a packet forwarding library to effectively transmit packets at specific ingress port to the right egress port defined by freeRouter control plane message. However, in this precise case, packet processing is offloaded from the kernel to user space. The consequence is the ability with dpdk compatible NIC and driver, to reach tremendous traffic rate. DPDK is not available on all hardware, please refer to DPDK HCL. |
Diagram
[ #006 ] - Cookbook
Expand |
---|
title | Install your favorite operating system |
---|
|
In our example we will use the ubuntu focal as we need dpdk 19.11.1 (latest current version is 20.05.0)
and we add a bridge network interface to or laptop RJ45 connection.
|
...
Expand |
---|
title | Install FreeRouter as per #001 article: "Create freeRouter environment" |
---|
|
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | Install dpdk and dpdk-dev |
---|
| apt-get update
apt-get updradeupgrade
apt-get install dpdk dpdk-dev --no-install-recommends |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | flush enp0s3 so that it can be controlled by dpdk |
---|
| ip addr flush enp0s3 |
Info |
---|
title | Source Wikipedia:Add out of band management enp0s8 with Virtualbox |
---|
| You can add a second Host-only interface interface (enp0s8) in VirtualBox in order to connect the ubuntu focal VM guest as you might lose connection when you flushed enp0s3. |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | Setup up dpdk and veth pair for control plane dataplane discussion via pcapInt |
---|
| #!/bin/bash
echo 1 > /proc/sys/net/ipv6/conf/all/disable_ipv6
echo 64 > /proc/sys/vm/nr_hugepages
modprobe uio_pci_generic
dpdk-devbind.py -b uio_pci_generic 00:03.0
ip link add veth0a type veth peer name veth0b
ip link set veth0a up
ip link set veth0b up |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | Check that dpdk is controlling able to enp0s3 |
---|
| dpdk-devbind.py --status
Network devices using DPDK-compatible driver
============================================
0000:00:03.0 '82540EM Gigabit Ethernet Controller 100e' drv=uio_pci_generic unused=e1000,vfio-pci
Network devices using kernel driver
===================================
0000:00:08.0 '82540EM Gigabit Ethernet Controller 100e' if=enp0s8 drv=e1000 unused=vfio-pci,uio_pci_generic *Active*
No 'Baseband' devices detected
==============================
No 'Crypto' devices detected
============================
No 'Eventdev' devices detected
==============================
No 'Mempool' devices detected
=============================
No 'Compress' devices detected
==============================
No 'Misc (rawdev)' devices detected
=================================== |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
| mkdir -p ~/freeRouter/bin ~/freeRouter/lib ~/freeRouter/etc ~/freeRouter/log
cd ~/freeRouter/lib
wget http://freerouter.nop.hu/rtr.jar |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | Update & Upgrade system |
---|
| tree freeRouter
freeRouter
├── bin # binary files
├── etc # configuration files
├── lib # library files
└── log # log files
|
|
...
Expand |
---|
title | Install freeRouter net-tools |
---|
|
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | get freeRouter net-tools tarball |
---|
| wget freerouter.nop.hu/rtr.tarhttp://www.freertr.net/rtr-`uname -m`.tar -O rtr.tar
|
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | Install build tools |
---|
| tar xvf rtr.tar -C ~/freeRouter/bin/ |
Note |
---|
For those you would like to rebuild these binaries you can find the compilation shell script in freeRouter cloned git repository in: ~/freeRouter/src/native/c.sh |
|
...
Expand |
---|
title | Create configuration files for router: freerouter |
---|
|
FreeRouter uses 2 configuration files in order to run, let's write these configuration files for R1 in ~/freeRouter/etc Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | freeRouter hardware configuration file: dpdk-focal-1-hw.txt |
---|
| hwid hp
! cpu_port
int eth0 eth - 127.0.0.1 20001 127.0.0.1 20002
! freerouter control port for message
tcp2vrf 9080 v1 9080
! freerouter cli
tcp2vrf 2323 v1 23
! launch a process called "veth0" that actually link to veth0b
! cmd: ip link add veth0a type veth peer name veth0b
proc veth0 /root/freertr/bin/pcapInt.bin veth0a 20002 127.0.0.1 20001 127.0.0.1
proc p4emu /root/freertr/bin/p4dpdk.bin --vdev=net_af_packet0,iface=veth0b 127.0.0.1 9080 1 |
Info |
---|
title | Source WikipediaNote: |
---|
| Let's spend some times on this hardware configuration file, as you might have notice there are additional interesting lines worth to mention: It is possible within freeRouter startup to launch processes. We use here this feature to start control plane / dataplane communication via veth pair: veth0a and veth0b and also P4Emu/dpdk, p4dpdk.bin packet processing backend. - proc p4emu /root/freertr/bin/p4dpdk.bin --vdev=net_af_packet0,iface=veth0b 127.0.0.1 9080 1
In dpdk, by default dpdk interfaces have port_ids that are sequentially allocated and in the order of appearance in dpdk-devbind --status output usually sorted by pci_id. In the previous output interface enp0s3 has port_id #0 and in dpdk veth0b (CPU_PORT has alwasy the last port_id beside dpdk data port_id, so here it is 1. If for exaplem we dedicate enp0s3, enp0s8, enp0s9, enp0s10 in virtualbox the command would have been: proc p4emu /root/freertr/bin/p4dpdk.bin --vdev=net_af_packet0,iface=veth0b 127.0.0.1 9080 4 enp0s3 would be: #0 with pci_id: 00:03.0 enp0s8 would be: #1 with pci_id: 00:08.0 enp0s9 would be: #2 with pci_id: 00:09.0 enp0s10 would be: #3 with pci_id: 00:0a.0 |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | freeRouter software configuration file: p4dpdk-freerouter-sw.txt |
---|
| hostname dpdk-freerouter
buggy
!
!
vrf definition v1
rd 1:1
exit
!
interface ethernet0
description freerouter@P4_CPU_PORT[veth0a]
no shutdown
no log-link-change
exit
!
interface sdn1
description freerouter@P4_CPU_PORT[enp0s3]
mtu 1500
vrf forwarding v1
ipv4 address 192.168.0.131 255.255.255.0
ipv6 address 2a01:e0a:159:2850::666 ffff:ffff:ffff:ffff::
ipv6 enable
no shutdown
no log-link-change
exit
!
!
!
!
!
!
!
!
!
!
!
!
!
!
server telnet telnet
security protocol telnet
no exec authorization
no login authentication
vrf v1
exit
!
server p4lang p4
export-vrf v1 1
export-port sdn1 0 0
interconnect ethernet0
vrf v1
exit
!
!
end |
|
...
Expand |
---|
title | Launch freeRouter control plane alongside P4Emu/pcap dataplane |
---|
|
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | freeRouter launch with supplied p4dpdk-freerouter-hw.txt and p4dpdk-freerouter-sw.txt with a console prompt |
---|
| java -jar lib/rtr.jar routersc dpdk-focal-1-hw.txt dpdk-focal-1-sw.txt
info cfg.cfgInit.doInit:cfgInit.java:556 booting
info cfg.cfgInit.doInit:cfgInit.java:680 initializing hardware
info cfg.cfgInit.executeHWcommands:cfgInit.java:469 2:! cpu_port
info cfg.cfgInit.executeHWcommands:cfgInit.java:469 4:! freerouter control port for message
info cfg.cfgInit.executeHWcommands:cfgInit.java:469 6:! freerouter cli
info cfg.cfgInit.executeHWcommands:cfgInit.java:469 8:! launch a process called "veth0" that actually link to veth0b
info cfg.cfgInit.executeHWcommands:cfgInit.java:469 9:! cmd: ip link add veth0a type veth peer name veth0b
info cfg.cfgInit.doInit:cfgInit.java:687 applying defaults
info cfg.cfgInit.doInit:cfgInit.java:695 applying configuration
info cfg.cfgInit.doInit:cfgInit.java:721 done
welcome
line ready
dpdk-freertr-1# |
|
Verification
Expand |
---|
title | Check telnet access for freerouter@2323 |
---|
|
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | FreeRouter telnet access from Virtualbox VM guest via port 2323 |
---|
| root@focal-1:~# telnet 127.0.0.1 2323
Trying 127.0.0.1...
Connected to 127.0.0.1.
Escape character is '^]'.
welcome
line ready
dpdk-freerouter# |
|
...
Expand |
---|
title | Connectivity IPv4/IPv6 check for freeRouter |
---|
|
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | Ping IPv4 from freerouter -> LAN router gateway |
---|
| dpdk-freerouter#ping 192.168.0.254 /vrf v1
pinging 192.168.0.254, src=null, cnt=5, len=64, tim=1000, ttl=255, tos=0, sweep=false
!!!!!
result=100%, recv/sent/lost=5/5/0, rtt min/avg/max/total=1/1/1/6 |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | Ping IPv4 from freerouter -> LAN server |
---|
| dpdk-freerouter#ping 192.168.0.62 /vrf v1
pinging 192.168.0.62, src=null, cnt=5, len=64, tim=1000, ttl=255, tos=0, sweep=false
.!!!!
result=80%, recv/sent/lost=4/5/1, rtt min/avg/max/total=1/1/2/1005 |
Note |
---|
Please observe the 1st ICMP packet loss that triggered ARP learning for respectively 192.168.0.254 and 192.168.0.62. |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | IPv4 arp check |
---|
| dpdk-freerouter#sh ipv4 arp sdn1
mac address time static
e03f.496d.1899 192.168.0.62 00:00:24 false <----- Host server
0024.d4a0.0cd3 192.168.0.254 00:00:24 false <----- LAN gateway |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | Ping IPv6 from freerouter -> LAN router |
---|
| dpdk-freerouter#ping 2a01:e0a:159:2850::1 /vrf v1
pinging 2a01:e0a:159:2850::1, src=null, cnt=5, len=64, tim=1000, ttl=255, tos=0, sweep=false
.!!!!
result=80%, recv/sent/lost=4/5/1, rtt min/avg/max/total=1/1/2/1005 |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | Ping IPv6 from freerouter -> Host server and SSH connection test |
---|
| dpdk-freerouter#ping 2a01:e0a:159:2850:e23f:49ff:fe6d:1899 /vrf v1
pinging 2a01:e0a:159:2850:e23f:49ff:fe6d:1899, src=null, cnt=5, len=64, tim=1000, ttl=255, tos=0, sweep=false
.!!!!
result=80%, recv/sent/lost=4/5/1, rtt min/avg/max/total=1/1/2/1006 |
Note |
---|
Please observe the 1st ICMP packet loss that triggered IPv6 neighbor discovery for respectively 2a01:e0a:159:2850::1 and 2a01:e0a:159:2850:e23f:49ff:fe6d:1899 |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | IPv6 neighbor discovery check |
---|
| dpdk-freerouter#show ipv6 neighbors sdn1
mac address time static router
0024.d4a0.0cd3 2a01:e0a:159:2850::1 00:00:39 false false <----- LAN gateway
e03f.496d.1899 2a01:e0a:159:2850:e23f:49ff:fe6d:1899 00:00:39 false false <----- Host server
0024.d4a0.0cd3 fe80::224:d4ff:fea0:cd3 00:00:39 false false <----- Link local LAN gateway EUI64 IPv6 address
e03f.496d.1899 fe80::e23f:49ff:fe6d:1899 00:00:39 false false <----- Link local host server IPv6 address |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | Initiate IPv4 ssh from freerouter -> LAN router gateway |
---|
| dpdk-freerouter#ssh 192.168.0.62 /vrf v1 /user my-nas
- connecting to 192.168.0.62 22
password: *******
- securing connection
Last login: Tue Jul 7 17:40:55 2020 from 2a01:e0a:159:2850::666
FreeBSD 11.3-RELEASE-p9 (FreeNAS.amd64) #0 r325575+588899735f7(HEAD): Mon Jun 1 15:04:31 EDT 2020
FreeNAS (c) 2009-2020, The FreeNAS Development Team
All rights reserved.
FreeNAS is released under the modified BSD license.
For more information, documentation, help or support, go here:
http://freenas.org
Welcome to FreeNAS
MY-NAS% |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | Initiate IPv6 ssh from freerouter -> LAN router gateway |
---|
| dpdk-freerouter#ssh 2a01:e0a:159:2850:e23f:49ff:fe6d:1899 /vrf v1 /user my-nas
- connecting to 2a01:e0a:159:2850:e23f:49ff:fe6d:1899 22
password: *******
- securing connection
Last login: Wed Jul 8 11:28:32 2020 from 192.168.0.131
FreeBSD 11.3-RELEASE-p9 (FreeNAS.amd64) #0 r325575+588899735f7(HEAD): Mon Jun 1 15:04:31 EDT 2020
FreeNAS (c) 2009-2020, The FreeNAS Development Team
All rights reserved.
FreeNAS is released under the modified BSD license.
For more information, documentation, help or support, go here:
http://freenas.org
Welcome to FreeNAS
MY-NAS% |
Code Block |
---|
language | bash |
---|
theme | Midnight |
---|
title | freeRouter p4dpdk hardware statistics |
---|
| dpdk-freerouter#sh int snd1 hw
hwcounters - hardware counters
hwdrhistory - hardware historic drop byte counters
hwdrphistory - hardware historic drop packet counters
hwhistory - hardware historic byte counters
hwnumhist - hardware numeric historic byte counters
hwnumphist - hardware numeric historic packet counters
hwphistory - hardware historic packet counters
hwrates - hardware traffic rates
hwrealtime - hardware realtime counters
hwrxhistory - hardware historic rx byte counters
hwrxphistory - hardware historic rx packet counters
hwtxhistory - hardware historic tx byte counters
hwtxphistory - hardware historic tx packet counters
dpdk-freerouter#show interfaces sdn1 hwrates
packet byte
time tx rx drop tx rx drop
1sec 5 20 0 1498 4668 0
1min 39 104 0 48056 56745 0
1hour 31 174 0 10162 137481 0
dpdk-freerouter#show interfaces sdn1 hwhistory
217k|
195k| #
173k| #
151k| # # #
130k| # # # # # # #
108k| # # # # # # # # #
86k| # # # # # ### # # # ## ##
65k| # # # # # # # # ##### ## # # ##### # ##
43k|## # #### # #### ### # ########## ###### # ##### # #####
21k|## ###### # ##### ##### ########## ########################
0|###########################################################
bps|0---------10--------20--------30--------40--------50-------- seconds
43m|
39m| *
34m| *
30m| * *
26m| * *
21m| * *
17m| * *
13m| * * *
8684k| * * * * * *
4342k| * ** * * * ** * ** *
0|###########################################################
bps|0---------10--------20--------30--------40--------50-------- minutes
70m|
63m| * *
56m| * *
49m| * *
42m| * *
35m| * *
28m| * *
21m|** *
14m|** *
7017k|****
0|##*#
bps|0---------10--------20--------30--------40--------50-------- hours
|
|
Conclusion
In this article you:
- had a demonstration of how to integrate freeRouter into a local area network (Similar to article #002)
- However instead of using P4Emu/dpdk we used a P4Emu/dpdk dataplane
- communication between freeRouter control plane and P4Emu/dpdk is ensured by pcapInt via veth pair [ veth0a - veth0b ]
- In this example the freeRouter with P4Emu/dpdk has only 1 dataplane interface that is bound to enp0s3 VM interface exposed to the local network as a bridged interface
Tip |
---|
title | [ #005 #006 ] RARE/FreeRouter-101 - key take-away |
---|
|
- FreeRouter is using UNIX socket in order to forward packet dedicated to control plane + dataplane communication.
This essential paradigm is used to ensure communication between freeRouter and P4Emu/dpdk dataplane. It is ensured by pcapInt binary from freeRouter net-tools that will bind freeRouter socket (veth0a@locathost:22001) to a virtual network interface (veth0b@localhost:22002) connected to CPU_PORT 1. - freeRouter is the control plane for P4Emu/dpdk dataplane
freeRouter is doing all the control plane route computation and write/modify/remove message entry P4 entries are created/modified/removed accordingly from P4Emu/dpdk tables. Although the name is P4Emu, it does not emulate BMv2 V1Model.p4, but rather router.p4 dpkg port_id allocation follow pci_id port naming convention starting from id 0. p4dpdk.bin is invoked with the parameter: (number_of_dpdk_port - 1) + 1 <--- CPU_PORT - In this setup the combination of freeRouter/P4Emu/dpdk delivers a solution for small campus network having 10GE links (100GE links to be validated)
dpkg removed the kernel intervention calls for each packet processed. In that configuration packet processing is now off loaded to user space. Reducing kernel intervention to ~ 0%. Congratulation you have a hardware NIC assisted forwarding is system ! Note |
---|
In subsequent article we will see how this setup behaves with a DELL 640 server powered by Intel(R) Xeon(R) Gold 6138 CPU x 2 and equipped with a Mellanox ConnectX-5 EX Dual Port 100GbE QSFP28 PCIe Adapter Low Profile card. We will also see how to connect this server to a P4 switch, BF2556X-1T. So stay tuned ! |
|
...