quic协议中bbr拥塞控制性能试验

对quic协议中的bbr拥塞控制链路进行了测试，采用mininet，带宽链路为3Mbps,一共起了三条流，启动时刻分别为0s，30s，80s。
最终结果如图：
在这里插入图片描述
发送速率
增加一个接收端的带宽随时间变化的图。
接收速率
带宽利用率非常优秀，也算是基本保证了带宽的公平性。在带宽争抢阶段，测试代码[1]。
测试了不同的链路带宽，buffer的配置，bbr在有些情况下数据包丢包率高的吓人，有些情况下，效果很好。
我的测试拓补需要使用mininet，拓补文件为time.py。命令为

sudo su
python time.py

也可以在mininet的虚拟终端中测试，脚本为test_top.py:

#!/usr/bin/python
from mininet.topo import Topo
from mininet.net import Mininet
from mininet.cli import CLI
from mininet.link import TCLink
import time
import subprocess
import os,signal
import sys
#   0   1   2   3   4
#h1--r1--r2--r3--r4--h2
# 10.0.X.0
def rp_disable(host):
    ifaces = host.cmd('ls /proc/sys/net/ipv4/conf')
    ifacelist = ifaces.split()    # default is to split on whitespace
    for iface in ifacelist:
       if iface != 'lo': host.cmd('sysctl net.ipv4.conf.' + iface + '.rp_filter=0')

max_queue_size = 100
net = Mininet( cleanup=True )
h1 = net.addHost('h1',ip='10.0.0.1')
r1 = net.addHost('r1',ip='10.0.0.2')
r2 = net.addHost('r2',ip='10.0.1.2')
r3 = net.addHost('r3',ip='10.0.2.2')
r4 = net.addHost('r4',ip='10.0.3.2')
h2 = net.addHost('h2',ip='10.0.4.2')
c0 = net.addController( 'c0' )

net.addLink(h1,r1,intfName1='h1-eth0',intfName2='r1-eth0',cls=TCLink , bw=5, delay='20ms', max_queue_size=10*max_queue_size)
net.addLink(r1,r2,intfName1='r1-eth1',intfName2='r2-eth0',cls=TCLink , bw=5, delay='20ms', max_queue_size=10*max_queue_size)
net.addLink(r2,r3,intfName1='r2-eth1',intfName2='r3-eth0',cls=TCLink , bw=3, delay='10ms', max_queue_size=max_queue_size)
net.addLink(r3,r4,intfName1='r3-eth1',intfName2='r4-eth0',cls=TCLink , bw=5, delay='20ms', max_queue_size=10*max_queue_size)
net.addLink(r4,h2,intfName1='r4-eth1',intfName2='h2-eth0',cls=TCLink , bw=5, delay='20ms', max_queue_size=10*max_queue_size)
net.build()

h1.setIP('10.0.0.1', intf='h1-eth0')
h1.cmd("ifconfig h1-eth0 10.0.0.1 netmask 255.255.255.0")
h1.cmd("route add default gw 10.0.0.2 dev h1-eth0")

r1.cmd("ifconfig r1-eth0 10.0.0.2/24")
r1.cmd("ifconfig r1-eth1 10.0.1.1/24")
r1.cmd("ip route add to 10.0.4.0/24 via 10.0.1.2")
r2.cmd("ip route add to 10.0.0.0/24 via 10.0.0.1")
r1.cmd('sysctl net.ipv4.ip_forward=1')
rp_disable(r1)

r2.cmd("ifconfig r2-eth0 10.0.1.2/24")
r2.cmd("ifconfig r2-eth1 10.0.2.1/24")
r2.cmd("ip route add to 10.0.4.0/24 via 10.0.2.2")
r2.cmd("ip route add to 10.0.0.0/24 via 10.0.1.1")
r2.cmd('sysctl net.ipv4.ip_forward=1')
rp_disable(r2)

r3.cmd("ifconfig r3-eth0 10.0.2.2/24")
r3.cmd("ifconfig r3-eth1 10.0.3.1/24")
r3.cmd("ip route add to 10.0.4.0/24 via 10.0.3.2")
r3.cmd("ip route add to 10.0.0.0/24 via 10.0.2.1")
r3.cmd('sysctl net.ipv4.ip_forward=1')
rp_disable(r3)

r4.cmd("ifconfig r4-eth0 10.0.3.2/24")
r4.cmd("ifconfig r4-eth1 10.0.4.1/24")
r4.cmd("ip route add to 10.0.4.0/24 via 10.0.4.2")
r4.cmd("ip route add to 10.0.0.0/24 via 10.0.3.1")
r4.cmd('sysctl net.ipv4.ip_forward=1')
rp_disable(r4)

h2.setIP('10.0.4.2', intf='h2-eth0')
h2.cmd("ifconfig h2-eth0 10.0.4.2 netmask 255.255.255.0")
h2.cmd("route add default gw 10.0.4.1")

net.start()
time.sleep(1)
client_cmd="./t_sender --log %d --cp %d  --sp %d --duration %d --offset %d "
server_cmd="./t_receiver --log %d --port %d --duration %d --offset %d"
client1_cmd=client_cmd%(1,1234,4321,300000,0)
server1_cmd=server_cmd%(1,4321,310000,0)

CLI(net)
print "stop"

输入命令的步骤如图：
在这里插入图片描述

[1] quic-bbr bandwidth performance https://github.com/SoonyangZhang/stupid_simple/tree/master/quic-test
[2] TCP BBR失速控制的一个小trick一个小patch https://blog.csdn.net/dog250/article/details/80203520

quic协议中bbr拥塞控制性能试验

猜你喜欢