文章目录

Mininet教程（六）：多数据中心带宽实验
- 拓扑结构
- - 使能生成树协议(STP)
- 带宽测试
- - 自定义命令拓展
  - 流量随机模型下UDP带宽测试

Mininet教程（六）：多数据中心带宽实验

SDNLAB：Mininet入门与实战笔记

拓扑结构

在mininet/custom文件夹中新建py脚本（命名为fattree.py）：

from mininet.topo import Topo
from mininet.net import Mininet
from mininet.node import RemoteController,CPULimitedHost
from mininet.link import TCLink
from mininet.util import dumpNodeConnectionsclass MyTopo(Topo):def __init__(self):# Initialize topologyTopo.__init__(self)L1=2L2=L1*2L3=L2Core_Switches=[]Aggregation_Switches=[]Edge_Switches=[]# add swithesfor i in range(L1):sw=self.addSwitch('C{}'.format(i+1))Core_Switches.append(sw)for i in range(L2):sw=self.addSwitch('A{}'.format(i+1))Aggregation_Switches.append(sw)for i in range(L3):sw=self.addSwitch('E{}'.format(i+1))Edge_Switches.append(sw)# add linksfor sw1 in Core_Switches:for sw2 in Aggregation_Switches:self.addLink(sw1,sw2)for i in range(0,L2,2):for sw1 in Aggregation_Switches[i:i+2]:for sw2 in Edge_Switches[i:i+2]:self.addLink(sw1,sw2)# add hosts and linkscount=1for sw1 in Edge_Switches:for i in range(2):host=self.addHost('H{}'.format(count))self.addLink(sw1,host)count+=1topos={'mytopo':(lambda:MyTopo())}

使能生成树协议(STP)

在终端执行命令生成拓扑：

sudo mn --custom fattree.py --topo mytopo

进行主机互ping测试，发现不能ping通：

mininet> pingall
*** Ping: testing ping reachability
H1 -> X X X X X X X
H2 -> X X X X X X X
H3 -> X X X X X X X
H4 -> X X X X X X X
H5 -> X X X X X X X
H6 -> X X X X X X X
H7 -> X X X X X X X
H8 -> X X X X X X X
*** Results: 100% dropped (0/56 received)

如果使用iperf测试TCP带宽，会报错Could not connect to iperf on port 5001：

mininet> iperf H1 H2
*** Iperf: testing TCP bandwidth between H1 and H2
no route to 10.0.0.2:
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
10.0.0.0        0.0.0.0         255.0.0.0       U     0      0        0 H1-eth0
--------------------------------------------------------------------------------
Caught exception. Cleaning up...Exception: Could not connect to iperf on port 5001

这是因为该拓扑中包含了很多环路loop，不加抑制会有广播风暴的风险。查询mininet的接口手册，发现默认的ovs设置是：

def mininet.node.OVSSwitch.__init__  (       self,name,failMode = 'secure',datapath = 'kernel',inband = False,protocols = None,reconnectms = 1000,stp = False,batch = False,params
)#Parameters
#name           name for switch
#failMode       controller loss behavior (secure|standalone)
#datapath       userspace or kernel mode (kernel|user)
#inband         use in-band control (False)
#protocols      use specific OpenFlow version(s) (e.g. OpenFlow13) Unspecified (or old OVS version) uses OVS default
#reconnectms    max reconnect timeout in ms (0/None for default)
#stp            enable STP (False, requires failMode=standalone)
#batch          enable batch startup (False)

可以看到抑制环路的生成树协议（STP）是关闭的。使能STP还需要设置failMode属性为standalone。那么在什么地方设置呢？一种可行的方法是在运行mn的时候指定switch及属性，通过查询mn的帮助文件：

mn -h
Usage: mn [options]
(type mn -h for details)The mn utility creates Mininet network from the command line. It can create
parametrized topologies, invoke the Mininet CLI, and run tests.Options:-h, --help            show this help message and exit--switch=SWITCH       default|ivs|lxbr|ovs|ovsbr|ovsk|user[,param=value...]ovs=OVSSwitch default=OVSSwitch ovsk=OVSSwitchlxbr=LinuxBridge user=UserSwitch ivs=IVSSwitchovsbr=OVSBridge

在–switch里面，除了指定switch外，还可以附带param=value进行设置，所以在终端执行的命令里面额外加上switch设置：

sudo mn --custom=fattree.py --topo=mytopo --switch=ovsk,failMode='standalone',stp=True

此时再进行连通性测试：

mininet> pingall
*** Ping: testing ping reachability
H1 -> H2 H3 H4 H5 H6 H7 H8
H2 -> H1 H3 H4 H5 H6 H7 H8
H3 -> H1 H2 H4 H5 H6 H7 H8
H4 -> H1 H2 H3 H5 H6 H7 H8
H5 -> H1 H2 H3 H4 H6 H7 H8
H6 -> H1 H2 H3 H4 H5 H7 H8
H7 -> H1 H2 H3 H4 H5 H6 H8
H8 -> H1 H2 H3 H4 H5 H6 H7
*** Results: 0% dropped (56/56 received)

带宽测试

同一边缘交换机内部主机间的带宽：

mininet> iperf H1 H2
*** Iperf: testing TCP bandwidth between H1 and H2
*** Results: ['52.2 Gbits/sec', '52.3 Gbits/sec']

同一汇聚交换机不同边缘交换机的主机之间带宽：

mininet> iperf H1 H3
*** Iperf: testing TCP bandwidth between H1 and H3
*** Results: ['45.3 Gbits/sec', '45.4 Gbits/sec']

同一核心交换机不同汇聚交换机的主机之间带宽：

mininet> iperf H1 H5
*** Iperf: testing TCP bandwidth between H1 and H5
*** Results: ['39.7 Gbits/sec', '39.8 Gbits/sec']

自定义命令拓展

接下来进行mininet自定义命令功能的拓展，生成一个流量随机模型，主机向在网络中的其他任意主机以等概率发送数据包。在mininet中进行自定义命令拓展分为四步：

修改mininet/net.py，实现功能代码；
修改mininet/cli.py，注册命令
修改mininet/bin/mn，加入到可执行文件中
install.sh -n，重新安装mininet核心文件

修改mininet/net.py文件，在Mininet类中添加方法iperf_single和iperfMulti。这里将它们放到iperf方法的后面，因为这些都是一类方法。

 def iperf_single( self, hosts=None, udpBw='10M', period=60, port=5001):""" Run iperf between two hosts using UDP and record communications.hosts: list of hosts; if None, uses opposite hosts returns: results two-element array of server and client speeds"""if not hosts:returnelse:assert len( hosts ) == 2client, server = hostsfilename = client.name[1:]+'.out'output( '*** Iperf: testing bandwidth between ' )output( "%s and %s\n" % ( client.name, server.name ) )iperfArgs = 'iperf -u 'bwArgs = '-b ' + udpBw + ' 'print( "***start server***" )server.cmd( iperfArgs + '-s -i 1' + ' > /home/App/mininet/log/' + filename + '&')print( "***start client***" )client.cmd( iperfArgs + '-t ' + str(period) + ' -c ' +server.IP() + ' ' + bwArgs +\'> /home/App/mininet/log/' + 'client' + filename + '&')def iperfMulti( self, bw, period=60 ):base_port = 5001server_list = []host_list = [h for h in self.hosts]_len = len(host_list)for i in xrange(0, _len):client = host_list[i]server = clientwhile( server == client ):server = random.choice(host_list)server_list.append(server)self.iperf_single(hosts=[client,server],udpBw=bw,period=period,port=base_port)sleep(.05)base_port += 1sleep(period)print("test has done")

其中，iperf_single方法主要实现了两个主机之间的UDP带宽测试，并将测试的结果存储在"/home/App/mininet/log/"目录下。iperfMulti则从Host1开始，随机选取UDP数据流的发送对象。持续测试period时长，默认是60秒。

修改mininet/cli.py，注册iperfMulti命令，对用户输入的指令进行解析。

 def do_iperfmulti( self, line ):"""Multi iperf UDP test between nodes"""args = line.split()if len(args) == 1:udpBw = args[0]self.mn.iperfMulti(udpBw)elif len(args) == 2:udpBw = args[0]period = args[1]err = Falseself.mn.iperfMulti(udpBw, float(period))else:error('invalid number of args: iperfmulti udpBw period\n' +\'udpBw examples: 1M 120\n' )

修改mininet/bin/mn，将iperfmulti加入TEST和ALTSPELLING列表中：

# TESTS dict can contain functions and/or Mininet() method names
# XXX: it would be nice if we could specify a default test, but
# this may be tricky
TESTS = { name: Truefor name in ( 'pingall', 'pingpair', 'iperf', 'iperfudp', 'iperfmulti' ) }# Map to alternate spellings of Mininet() methods
ALTSPELLING = { 'pingall': 'pingAll', 'pingpair': 'pingPair','iperfudp': 'iperfUdp', 'iperfmulti': 'iperfMulti' }

进入mininet/util目录下，执行:

sudo ./install.sh -n

重新编译mininet

流量随机模型下UDP带宽测试

再次启动mn并执行iperfmulti，这里指定了带宽0.025M，测试时长10秒：

mininet> iperfmulti 0.025M 10
*** Iperf: testing bandwidth between H1 and H5
***start server***
***start client***
*** Iperf: testing bandwidth between H2 and H8
***start server***
***start client***
*** Iperf: testing bandwidth between H3 and H1
***start server***
***start client***
*** Iperf: testing bandwidth between H4 and H2
***start server***
***start client***
*** Iperf: testing bandwidth between H5 and H6
***start server***
***start client***
*** Iperf: testing bandwidth between H6 and H3
***start server***
***start client***
*** Iperf: testing bandwidth between H7 and H4
***start server***
***start client***
*** Iperf: testing bandwidth between H8 and H6
***start server***
***start client***
test has done

测试完成后，可以查看log中的记录文件：

ls
1.out  3.out  5.out  7.out  client1.out  client3.out  client5.out  client7.out
2.out  4.out  6.out  8.out  client2.out  client4.out  client6.out  client8.out

查看1.out：

------------------------------------------------------------
Server listening on UDP port 5001
Receiving 1470 byte datagrams
UDP buffer size: 2.00 MByte (default)
------------------------------------------------------------
[  3] local 10.0.0.5 port 5001 connected with 10.0.0.1 port 58102
[ ID] Interval       Transfer     Bandwidth        Jitter   Lost/Total Datagrams
[  3]  0.0- 1.0 sec  4.31 KBytes  35.3 Kbits/sec   0.027 ms    0/    3 (0%)
[  3]  1.0- 2.0 sec  2.87 KBytes  23.5 Kbits/sec   0.026 ms    0/    2 (0%)
[  3]  2.0- 3.0 sec  2.87 KBytes  23.5 Kbits/sec   0.026 ms    0/    2 (0%)
[  3]  3.0- 4.0 sec  2.87 KBytes  23.5 Kbits/sec   0.025 ms    0/    2 (0%)
[  3]  4.0- 5.0 sec  2.87 KBytes  23.5 Kbits/sec   0.024 ms    0/    2 (0%)
[  3]  5.0- 6.0 sec  2.87 KBytes  23.5 Kbits/sec   0.024 ms    0/    2 (0%)
[  3]  6.0- 7.0 sec  2.87 KBytes  23.5 Kbits/sec   0.021 ms    0/    2 (0%)
[  3]  7.0- 8.0 sec  4.31 KBytes  35.3 Kbits/sec   0.021 ms    0/    3 (0%)
[  3]  8.0- 9.0 sec  2.87 KBytes  23.5 Kbits/sec   0.021 ms    0/    2 (0%)
[  3]  9.0-10.0 sec  2.87 KBytes  23.5 Kbits/sec   0.022 ms    0/    2 (0%)
[  3]  0.0-10.8 sec  33.0 KBytes  25.0 Kbits/sec   0.021 ms    0/   23 (0%)

这里记录了Server 10.0.0.5 (Host5) 侦听端口5001的记录，即与10.0.0.1 (Host1)的测试记录。一共测试了10秒，每隔1秒记录一次，记录内容有转发包的大小，带宽，延时抖动和丢包率等。

查看client1.out:

------------------------------------------------------------
Client connecting to 10.0.0.5, UDP port 5001
Sending 1470 byte datagrams, IPG target: 470400.00 us (kalman adjust)
UDP buffer size: 2.00 MByte (default)
------------------------------------------------------------
[  3] local 10.0.0.1 port 58102 connected with 10.0.0.5 port 5001
[ ID] Interval       Transfer     Bandwidth
[  3]  0.0-10.8 sec  33.0 KBytes  25.0 Kbits/sec
[  3] Sent 23 datagrams
[  3] Server Report:
[  3]  0.0-10.8 sec  33.0 KBytes  25.0 Kbits/sec   0.020 ms    0/   23 (0%)

这里记录了Client 10.0.0.1 (Host1)的测试记录。