阿里云可以建网站吗,广告平台代理,青岛房产网,wordpress建企业站教程QoS#xff08;Quality of Service#xff09;是服务质量的简称。对于网络业务来说#xff0c;服务质量包括哪些方面呢#xff1f;
从传统意义上来讲#xff0c;无非就是传输的带宽、传送的时延、数据的丢包率等#xff0c;而提高服务质量无非也就是保证传输的带宽…QoSQuality of Service是服务质量的简称。对于网络业务来说服务质量包括哪些方面呢
从传统意义上来讲无非就是传输的带宽、传送的时延、数据的丢包率等而提高服务质量无非也就是保证传输的带宽降低传送的时延降低数据的丢包率以及时延抖动等。
广义上讲服务质量涉及网络应用的方方面面只要是对网络应用有利的措施其实都是在提高服务质量。因此从这个意义上来说防火墙、策略路由、快速转发等也都是提高网络业务服务质量的措施之一。
服务质量相对网络业务而言在保证某类业务服务质量的同时可能就是在损害其它业务的服务质量。因为网络资源总是有限的只要存在抢夺网络资源的情况就会出现服务质量的要求。比如网络总带宽为100Mbps而BT下载占用了90Mbps其他业务就只能占用剩下的10Mbps。而如果限制BT下载占用的最大带宽为50Mbps也就提高了其他业务的服务质量使其他业务能够占用最少50Mbps的带宽但这是在损害BT业务的服务质量为前提的。
下面瑞哥将通过实验来给大家讲解Qos的配置。
文章目录一、实验目的二、实验拓扑三、实验场景四、实验步骤4.1 步骤一. 基础配置与 IP 编址R1R2R3R4R5S3S44.2 步骤二. 配置静态路由与 NQA4.3 步骤三. 配置优先级映射4.4 步骤四. 配置整形与监管4.5 步骤五. 配置基于队列的拥塞管理与拥塞避免4.6 步骤六. 配置基于流的拥塞管理与拥塞避免五、附加实验: 思考并验证六、最终设备配置R1R2R3R4S3S4一、实验目的
掌握使用NQA分析SLA的方法掌握进行优先级映射和流量监管的方法掌握配置流量整形的方法掌握实现基于队列和基于流分类的拥塞管理的方法掌握配置WRED实现拥塞避免的方法
二、实验拓扑 三、实验场景
你是公司的网络管理员。公司网络分成两部分其中R1与S1在公司总部R2与S2在公司分部之间通过专线实现互联。
随着网络的发展内网带宽逐渐增大而专线的带宽一直没有升级所以网络中出现了比较严重的重要业务反应较慢或无法正常使用的情况。
使用QoS的差分服务你可以调整相应的QoS特性保证重要的业务数据能更好的发送给目标。
实验中S3和S4使用NQA相互发送数据模拟大量数据流的发送。R3、R4与R5模拟客户端和服务器测试重要应用是否可以正常使用。
四、实验步骤
4.1 步骤一. 基础配置与 IP 编址
给所有路由器和交换机S3S4配置IP地址和掩码。
配置时需要将R1接口S1/0/0的波特率配置为72000模拟广域网链路因带宽不足而出现拥塞。
R1
Huaweisystem-view
Enter system view, return user view with CtrlZ.
[Huawei]sysname R1
[R1]interface Serial 1/0/0
[R1-Serial1/0/0]ip address 10.0.12.1 255.255.255.0
[R1-Serial1/0/0]baudrate 72000
[R1-Serial1/0/0]interface GigabitEthernet 0/0/1
[R1-GigabitEthernet0/0/1]ip address 10.0.145.1 255.255.255.0R2
Huaweisystem-view
Enter system view, return user view with CtrlZ.
[Huawei]sysname R2
[R2]interface Serial 1/0/0
[R2-Serial1/0/0]ip address 10.0.12.2 255.255.255.0
[R2-Serial1/0/0]interface GigabitEthernet 0/0/2
[R2-GigabitEthernet0/0/2]ip address 10.0.34.2 255.255.255.0R3
Huaweisystem-view
Enter system view, return user view with CtrlZ.
[Huawei]sysname R3
[R3]interface GigabitEthernet 0/0/2
[R3-GigabitEthernet0/0/2]ip address 10.0.34.3 255.255.255.0R4
Huaweisystem-view
Enter system view, return user view with CtrlZ.
[Huawei]sysname R4
[R4]interface GigabitEthernet 0/0/1
[R4-GigabitEthernet0/0/1]ip address 10.0.145.4 255.255.255.0R5
Huaweisystem-view
Enter system view, return user view with CtrlZ.
[Huawei]sysname R5
[R5]interface GigabitEthernet 0/0/1
[R5-GigabitEthernet0/0/1]ip address 10.0.145.5 255.255.255.0S3
Huaweisystem-view
Enter system view, return user view with CtrlZ.
[Huawei]sysname S3
[S3]interface Vlanif 1
[S3-Vlanif1]ip address 10.0.145.3 255.255.255.0S4
Huaweisystem-view
Enter system view, return user view with CtrlZ.
[Huawei]sysname S4
[S4]interface Vlanif 1
[S4-Vlanif1]ip address 10.0.34.4 255.255.255.0配置完成后测试直连链路的连通性。
[R1]ping -c 1 10.0.12.2
PING 10.0.12.2: 56 data bytes, press CTRL_C to break
Reply from 10.0.12.2: bytes56 Sequence1 ttl255 time36 ms
--- 10.0.12.2 ping statistics ---
1 packet(s) transmitted
1 packet(s) received
0.00% packet loss
round-trip min/avg/max 36/36/36 ms[R1]ping -c 1 10.0.145.3
PING 10.0.145.3: 56 data bytes, press CTRL_C to break
Reply from 10.0.145.3: bytes56 Sequence1 ttl255 time35 ms
--- 10.0.145.3 ping statistics ---
1 packet(s) transmitted
1 packet(s) received
0.00% packet loss
round-trip min/avg/max 35/35/35 ms[R1]ping -c 1 10.0.145.4
PING 10.0.145.4: 56 data bytes, press CTRL_C to break
Reply from 10.0.145.4: bytes56 Sequence1 ttl255 time6 ms
--- 10.0.145.4 ping statistics ---
1 packet(s) transmitted
1 packet(s) received
0.00% packet loss
round-trip min/avg/max 6/6/6 ms[R1]ping -c 1 10.0.145.5
PING 10.0.145.5: 56 data bytes, press CTRL_C to break
Reply from 10.0.145.5: bytes56 Sequence1 ttl255 time6 ms
--- 10.0.145.5 ping statistics ---
1 packet(s) transmitted
1 packet(s) received
0.00% packet loss
round-trip min/avg/max 6/6/6 ms[R2]ping -c 1 10.0.34.3
PING 10.0.34.3: 56 data bytes, press CTRL_C to break
Reply from 10.0.34.3: bytes56 Sequence1 ttl255 time5 ms
--- 10.0.34.3 ping statistics ---
1 packet(s) transmitted
1 packet(s) received
0.00% packet loss
round-trip min/avg/max 5/5/5 ms[R2]ping -c 1 10.0.34.4
PING 10.0.34.4: 56 data bytes, press CTRL_C to break
Reply from 10.0.34.4: bytes56 Sequence1 ttl255 time36 ms
--- 10.0.34.4 ping statistics ---
1 packet(s) transmitted
1 packet(s) received
0.00% packet loss
round-trip min/avg/max 36/36/36 ms4.2 步骤二. 配置静态路由与 NQA
在所有路由器和交换机S3S4上配置静态路由。
[R1]ip route-static 10.0.34.0 255.255.255.0 10.0.12.2
[R2]ip route-static 10.0.145.0 255.255.255.0 10.0.12.1
[R3]ip route-static 0.0.0.0 0.0.0.0 10.0.34.2
[R4]ip route-static 0.0.0.0 0.0.0.0 10.0.145.1
[R5]ip route-static 0.0.0.0 0.0.0.0 10.0.145.1
[S3]ip route-static 0.0.0.0 0.0.0.0 10.0.145.1
[S4]ip route-static 0.0.0.0 0.0.0.0 10.0.34.2配置完成后测试网络连通性。
[S3]ping -c 1 10.0.34.4
PING 10.0.34.4: 56 data bytes, press CTRL_C to break
Reply from 10.0.34.4: bytes56 Sequence1 ttl252 time40 ms
--- 10.0.34.4 ping statistics ---
1 packet(s) transmitted
1 packet(s) received
0.00% packet loss
round-trip min/avg/max 40/40/40 ms[R4]ping -c 1 10.0.34.3
PING 10.0.145.4: 56 data bytes, press CTRL_C to break
Reply from 10.0.145.4: bytes56 Sequence1 ttl255 time3 ms
--- 10.0.145.4 ping statistics ---
1 packet(s) transmitted
1 packet(s) received
0.00% packet loss
round-trip min/avg/max 3/3/3 ms[R5]ping -c 1 10.0.34.3
PING 10.0.34.3: 56 data bytes, press CTRL_C to break
Reply from 10.0.34.3: bytes56 Sequence1 ttl253 time44 ms
--- 10.0.34.3 ping statistics ---
1 packet(s) transmitted
1 packet(s) received
0.00% packet loss
round-trip min/avg/max 44/44/44 msS3去往S4R4和R5去往R3可以连通证明网络通信正常。
公司总部和分部之间的链路为72K串行链路因而在实际情况中很容易造成拥塞。
实验中使用NQA在网络中产生流量。S4作为NQA服务器端S3作为NQA客户端。
定义UDPJitter两种NQA测试例分别用来模拟企业网中的数据流量和语音流量。
通过设置NQA测试例中的一些参数来实现两种流量中任何一种单独存在的情况下不会产生拥塞二者共存的情况下会产生拥塞来模拟实际环境。
在S4设上配置NQA服务器端监听的IP地址设为10.0.34.4UDP端口号设为6000。
[S4]nqa-server udpecho 10.0.34.4 6000在S3上配置UDP类型的NQA测试例模拟数据流量其中tos设为28包大小为5800字节包间隔设为1s周期设为3s超时设为1s并开启该测试。
[S3]nqa test-instance admin udp
[S3-nqa-admin-udp]test-type udp
[S3-nqa-admin-udp]destination-address ipv4 10.0.34.4
[S3-nqa-admin-udp]destination-port 6000
[S3-nqa-admin-udp]tos 28
[S3-nqa-admin-udp]datasize 5800
[S3-nqa-admin-udp]interval seconds 1
[S3-nqa-admin-udp]frequency 3
[S3-nqa-admin-udp]timeout 1
[S3-nqa-admin-udp]start now查看UDP测试结果。
[S3]display nqa results test-instance admin udp
1 . Test 2 result The test is finished
Send operation times: 3 Receive response times: 3
Completion:success RTD OverThresholds number: 0
Attempts number:1 Drop operation number:0
Disconnect operation number:0 Operation timeout number:0
System busy operation number:0 Connection fail number:0
Operation sequence errors number:0 RTT Stats errors number:0
Destination ip address:10.0.34.4
Min/Max/Average Completion Time: 930/950/943
Sum/Square-Sum Completion Time: 2830/2669900
Last Good Probe Time: 2010-10-10 18:10:02.4
Lost packet ratio: 0 %此时不丢包链路没有产生拥塞。关闭UDP测试。
[S3]nqa test-instance admin udp
[S3-nqa-admin-udp]stop在S3上配置Jitter类型的NQA测试例模拟语音流量其中tos设为46包大小为90字节包间隔设为20ms周期设为3s超时设为1s并开启该测试。
[S3]nqa test-instance admin jitter
[S3-nqa-admin-jitter]test-type jitter
[S3-nqa-admin-jitter]destination-address ipv4 10.0.34.4
[S3-nqa-admin-jitter]destination-port 6000
[S3-nqa-admin-jitter]tos 46
[S3-nqa-admin-jitter]datasize 90
[S3-nqa-admin-jitter]interval milliseconds 20
[S3-nqa-admin-jitter]frequency 3
[S3-nqa-admin-jitter]timeout 1
[S3-nqa-admin-jitter]start now查看Jitter测试结果。
[S3]display nqa results test-instance admin jitter
NQA entry(admin, jitter) :testflag is active ,testtype is jitter
1 . Test 1 result The test is finished
SendProbe:60 ResponseProbe:60
Completion:success RTD OverThresholds number:0
Min/Max/Avg/Sum RTT:40/70/54/3260 RTT Square Sum:179800
NumOfRTT:60 Drop operation number:0
Operation sequence errors number:0 RTT Stats errors number:0
System busy operation number:0 Operation timeout number:0
Min Positive SD:10 Min Positive DS:10
Max Positive SD:10 Max Positive DS:10
Positive SD Number:5 Positive DS Number:11
Positive SD Sum:50 Positive DS Sum:110
Positive SD Square Sum:500 Positive DS Square Sum:1100
Min Negative SD:10 Min Negative DS:10
Max Negative SD:10 Max Negative DS:20
Negative SD Number:4 Negative DS Number:10
Negative SD Sum:40 Negative DS Sum:110
Negative SD Square Sum:400 Negative DS Square Sum:1300
Min Delay SD:20 Min Delay DS:19
Avg Delay SD:27 Avg Delay DS:26
Max Delay SD:35 Max Delay DS:34
Packet Loss SD:0 Packet Loss DS:0
Packet Loss Unknown:0 jitter out value:0.0937500
jitter in value:0.2291667 NumberOfOWD:60
OWD SD Sum:1630 OWD DS Sum:1570
TimeStamp unit: ms此时不丢包链路没有产生拥塞。关闭Jitter测试。
[S3]nqa test-instance admin jitter
[S3-nqa-admin-jitter]stop4.3 步骤三. 配置优先级映射
现在通过ping命令来模拟公司中一些不太重要的流量并且针对这部分流量将其DSCP优先级映射为BE不做QoS保证。
配置R1的接口G0/0/1与S1/0/0信任报文的DSCP优先级。
[R1]interface GigabitEthernet 0/0/1
[R1-GigabitEthernet0/0/1]trust dscp override
[R1-GigabitEthernet0/0/1]interface Serial 1/0/0
[R1-Serial1/0/0]trust dscp在接口G0/0/1上的trust命令中需要加上override参数使得接下来在R1上配置优先级映射后将DSCP值修改为映射后的值。
在R4上使用ping命令产生去往R3的流量并且将tos设为26。
[R4]ping –tos 26 10.0.34.3在R1上配置优先级映射关系将该流量的DSCP报文优先级26映射为0
[R1]qos map-table dscp-dscp
[R1-maptbl-dscp-dscp]input 26 output 0查看R1上的优先级映射信息。
[R1]display qos map-table dscp-dscp
Input DSCP DSCP
-------------------
0 0
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20
21 21
22 22
23 23
24 24
25 25
26 0
27 27
28 28
29 29
30 30此时可以观察到现在已将DSCP优先级为26的报文优先级映射成为了0而其余DSCP值都是默认映射值。
4.4 步骤四. 配置整形与监管
开启S3上的NQA的UDP与Jitter测试模拟公司总部与分部之间的72K链路产生拥塞。
[S3]nqa test-instance admin udp
[S3-nqa-admin-udp]start now
[S3-nqa-admin-udp]quit
[S3]nqa test-instance admin jitter
[S3-nqa-admin-jitter]start now在R4上使用ping命令实现模拟去往R3的流量设置包大小为700字节发10个包。
[R4]ping -s 700 -c 10 10.0.34.3
PING 10.0.34.3: 700 data bytes, press CTRL_C to break
Request time out
Request time out
Request time out
Request time out
Request time out
Request time out
Request time out
Request time out
Reply from 10.0.34.3: bytes700 Sequence9 ttl253 time1944 ms
Request time out
--- 10.0.34.3 ping statistics ---
10 packet(s) transmitted
1 packet(s) received
90.00% packet loss
round-trip min/avg/max 1944/1944/1944 ms此时公司总部与分部之间的链路发生了严重拥塞丢包现象严重。即使有通过的数据包延迟也非常大。此时R4无法与R3建立正常通信。
下面将介绍分别通过使用流量监管和流量整形的方法来消除链路上的拥塞使得公司总部的客户端R4与分部的客户端R3能够建立正常通信。
首先通过流量监管来消除拥塞。在S1上针对拥塞流量在入接口G0/0/13上配置流量监管CIR设为64kbit/s。
[S1]interface GigabitEthernet 0/0/13
[S1-GigabitEthernet0/0/13]qos lr inbound cir 64查看S1上流量监管的配置信息。
[S1]display qos lr inbound interface GigabitEthernet 0/0/13
GigabitEthernet0/0/13 lr inbound:
cir: 64 Kbps, cbs: 8000 Byte现在再回到R4上使用ping命令实现模拟去往R3的流量设置包大小为700字节发10个包。
[R4]ping -s 700 -c 10 10.0.34.3
PING 10.0.34.3: 700 data bytes, press CTRL_C to break
Reply from 10.0.34.3: bytes700 Sequence1 ttl253 time1412 ms
Reply from 10.0.34.3: bytes700 Sequence2 ttl253 time255 ms
Reply from 10.0.34.3: bytes700 Sequence3 ttl253 time736 ms
Reply from 10.0.34.3: bytes700 Sequence4 ttl253 time1746 ms
Reply from 10.0.34.3: bytes700 Sequence5 ttl253 time246 ms
Reply from 10.0.34.3: bytes700 Sequence6 ttl253 time746 ms
Reply from 10.0.34.3: bytes700 Sequence7 ttl253 time1736 ms
Reply from 10.0.34.3: bytes700 Sequence8 ttl253 time258 ms
Reply from 10.0.34.3: bytes700 Sequence9 ttl253 time766 ms
Reply from 10.0.34.3: bytes700 Sequence10 ttl253 time1736 ms
--- 10.0.34.3 ping statistics ---
10 packet(s) transmitted
10 packet(s) received
0.00% packet loss
round-trip min/avg/max 246/963/1746 ms此时流量监管产生效果不丢包R4和R3之间能够建立起正常通信。删除S1上流量监管配置。
[S1]interface GigabitEthernet 0/0/13
[S1-GigabitEthernet0/0/13]undo qos lr inbound现在通过流量整形的方式来达到消除拥塞的目的。在S3上针对拥塞流量在出接口E0/0/13上配置流量整形CIR设为64kbit/s。
[S3]interface GigabitEthernet0/0/13
[S3-GigabitEthernet0/0/13]qos lr outbound cir 64在R4上使用ping命令实现模拟去往R3的流量设置包大小为700字节发10个包。
[R4]ping -s 700 -c 10 10.0.34.3PING 10.0.34.3: 700 data bytes, press CTRL_C to break
Reply from 10.0.34.3: bytes700 Sequence1 ttl253 time240 ms
Reply from 10.0.34.3: bytes700 Sequence2 ttl253 time284 ms
Reply from 10.0.34.3: bytes700 Sequence3 ttl253 time334 ms
Reply from 10.0.34.3: bytes700 Sequence4 ttl253 time224 ms
Reply from 10.0.34.3: bytes700 Sequence5 ttl253 time344 ms
Reply from 10.0.34.3: bytes700 Sequence6 ttl253 time275 ms
Reply from 10.0.34.3: bytes700 Sequence7 ttl253 time534 ms
Reply from 10.0.34.3: bytes700 Sequence8 ttl253 time184 ms
Reply from 10.0.34.3: bytes700 Sequence9 ttl253 time204 ms
Reply from 10.0.34.3: bytes700 Sequence10 ttl253 time314 ms
--- 10.0.34.3 ping statistics ---
10 packet(s) transmitted
10 packet(s) received
0.00% packet loss
round-trip min/avg/max 184/293/534 ms此时流量监管产生效果不丢包R4和R3之间能够建立起正常通信。删除S3上的流量整形配置。
[S3]interface GigabitEthernet0/0/13
[S3-GigabitEthernet0/0/13]undo qos lr outbound现在再回到R4上使用ping命令实现模拟去往R3的流量设置包大小为700字节发10个包。
[R4]ping -s 700 -c 10 10.0.34.3
PING 10.0.34.3: 700 data bytes, press CTRL_C to break
Reply from 10.0.34.3: bytes700 Sequence1 ttl253 time1918 ms
Request time out
Reply from 10.0.34.3: bytes700 Sequence3 ttl253 time1762 ms
Request time out
Request time out
Request time out
Request time out
Request time out
Request time out
Request time out
--- 10.0.34.3 ping statistics ---
10 packet(s) transmitted
2 packet(s) received
80.00% packet loss
round-trip min/avg/max 1762/1840/1918 ms删除配置之后丢包严重并且通过的数据包延迟也非常大。R4与R3之间无法建立起正常通信。
4.5 步骤五. 配置基于队列的拥塞管理与拥塞避免
为了解决公司总部与分部之间产生的网络拥塞现在通过配置基于队列的拥塞管理和拥塞避免的方式解决。
在R1上创建WRED丢弃模板data使其基于DSCP优先级进行丢弃将阀值上限设为90下限设为50丢弃概率设为30。
[R1]drop-profile data
[R1-drop-profile-data]wred dscp
[R1-drop-profile-data]dscp af32 low-limit 50 high-limit 90 discard-percentage 30在R1上创建队列模板queue-profile1将数据流量放入WFQ队列并和丢弃模板data绑定将需要高优先级低延迟保证的语音流量放入PQ队列。
[R1]qos queue-profile queue-profile1
[R1-qos-queue-profile-queue-profile1]schedule wfq 3 pq 5
[R1-qos-queue-profile-queue-profile1]queue 3 drop-profile data在R1的S1/0/0上应用队列模板。
[R1]interface Serial 1/0/0
[R1-Serial1/0/0]qos queue-profile queue-profile1查看配置的队列模板信息。
[R1]display qos queue-profile queue-profile1
Queue-profile: queue-profile1
Queue Schedule Weight Length(Bytes/Packets) Gts(CIR/CBS)
-----------------------------------------------------------------
3 WFQ 10 0/0 -/-
5 PQ - 0/0 -/-此时数据流量与语音流量分别使用了WFQ与PQ队列。查看配置的丢弃模板信息。
[R1]display drop-profile data
Drop-profile[1]: data
DSCP Low-limit High-limit Discard-percentage
-----------------------------------------------------------------
default 30 100 10
1 30 100 10
2 30 100 10
3 30 100 10
4 30 100 10
5 30 100 10
6 30 100 10
7 30 100 10
cs1 30 100 10
9 30 100 10
af11 30 100 10
11 30 100 10
af12 30 100 10
13 30 100 10
af13 30 100 10
15 30 100 10
cs2 30 100 10
17 30 100 10
af21 30 100 10
19 30 100 10
af22 30 100 10
21 30 100 10
af23 30 100 10
23 30 100 10
cs3 30 100 10
25 30 100 10
af31 30 100 10
27 30 100 10
af32 50 90 30
29 30 100 10
af33 30 100 10
31 30 100 10
cs4 30 100 10
33 30 100 10
af41 30 100 10可以观察到配置上限下限阀值与丢弃概率产生的效果其余没有配置的丢弃模板设置对应的都是默认值。
4.6 步骤六. 配置基于流的拥塞管理与拥塞避免
为了解决公司总部与分部之间产生的网络拥塞现在通过配置基于流的拥塞管理和拥塞避免的方式解决。
现在将公司总部的客户端R4与分部的客户端R3之间的流量定义为重要流量通过对其做QoS保证使得R4与R3能够建立正常的通信。
删除步骤五中R1接口S1/0/0上调用的队列模板。
[R1]interface GigabitEthernet 0/0/1
[R1-GigabitEthernet0/0/1]undo qos queue-profile在R4上使用ping命令测试去往R3的连通性设置源地址为10.0.145.4包大小为700字节发10个包。
[R4]ping -a 10.0.145.4 -s 700 -c 10 10.0.34.3
PING 10.0.34.3: 700 data bytes, press CTRL_C to break
Reply from 10.0.34.3: bytes700 Sequence1 ttl253 time1279 ms
Request time out
Reply from 10.0.34.3: bytes700 Sequence3 ttl253 time1587 ms
Reply from 10.0.34.3: bytes700 Sequence4 ttl253 time1827 ms
Request time out
Reply from 10.0.34.3: bytes700 Sequence6 ttl253 time1717 ms
Request time out
Request time out
Request time out
Request time out
--- 10.0.34.3 ping statistics ---
10 packet(s) transmitted
4 packet(s) received
60.00% packet loss
round-trip min/avg/max 1279/1602/1827 ms此时公司总部与分部之间的链路发生严重拥塞丢包现象严重R4无法与R3建立正常通信。
在R1上创建ACL3001匹配从10.0.145.4去往10.0.34.3的流量。
[R1]acl number 3001
[R1-acl-adv-3001]rule 0 per ip source 10.0.145.4 0.0.0.0 destination 10.0.34.3 0.0.0.0创建流分类class-ef匹配ACL3001创建流行为behavior-ef配置队列调度方式为EF带宽为10Kbps。
[R1]traffic classifier class-ef
[R1-classifier-class-ef]if-match acl 3001
[R1-classifier-class-ef]quit
[R1]traffic behavior behavior-ef
[R1-behavior-behavior-ef]queue ef bandwidth 10创建流分类class-af32匹配DSCP值为AF32的数据流量创建流行为behavior-af32配置队列调度方式为AF带宽为30Kbps与丢弃模板data绑定。
[R1]traffic classifier class-af32
[R1-classifier-class-af32]if-match dscp af32
[R1-classifier-class-af32]quit
[R1]traffic behavior behavior-af32
[R1-behavior-behavior-af32]queue af bandwidth 30
[R1-behavior-behavior-af32]drop-profile data创建流策略policy-1关联流分类class-ef和流动作behavior-ef流分类class-af32和流动作behavior-af32并在R1的接口S1/0/0上应用。
[R1]traffic policy policy-1
[R1-trafficpolicy-policy-1]classifier class-ef behavior behavior-ef
[R1-trafficpolicy-policy-1]classifier class-af32 behavior behavior-af32
[R1-trafficpolicy-policy-1]quit
[R1]interface Serial 1/0/0
[R1-Serial1/0/0]traffic-policy policy-1 outbound在R4上使用ping命令测试去往R3的连通性设置每个包大小为700源地址为10.0.145.4个数为10。
[R4]ping -a 10.0.145.4 -s 700 -c 10 10.0.34.3
PING 10.0.34.3: 700 data bytes, press CTRL_C to break
Reply from 10.0.34.3: bytes700 Sequence1 ttl253 time694 ms
Reply from 10.0.34.3: bytes700 Sequence2 ttl253 time391 ms
Reply from 10.0.34.3: bytes700 Sequence3 ttl253 time361 ms
Reply from 10.0.34.3: bytes700 Sequence4 ttl253 time671 ms
Reply from 10.0.34.3: bytes700 Sequence5 ttl253 time211 ms
Reply from 10.0.34.3: bytes700 Sequence6 ttl253 time611 ms
Reply from 10.0.34.3: bytes700 Sequence7 ttl253 time688 ms
Reply from 10.0.34.3: bytes700 Sequence8 ttl253 time391 ms
Reply from 10.0.34.3: bytes700 Sequence9 ttl253 time301 ms
Reply from 10.0.34.3: bytes700 Sequence10 ttl253 time651 ms
--- 10.0.34.3 ping statistics ---
10 packet(s) transmitted
10 packet(s) received
0.00% packet loss
round-trip min/avg/max 211/497/694 ms将R4去往R3的流量设置为EF队列后现在R4可以与R3建立正常通信。
五、附加实验: 思考并验证
QoS是使用差分服务来实现对不同业务服务质量保证的保证了带宽和延迟。试想一下不使用QoS通过增加带宽的方式是否可以彻底解决服务质量问题
实验完成后回想理论课程中关于QoS的逻辑处理过程。将路由器实现QoS的过程总结一下。
六、最终设备配置
R1
R1display current-configuration
[V200R007C00SPC600]
#
sysname R1
#
acl number 3001
rule 0 permit ip source 10.0.145.4 0 destination 10.0.34.3 0
#
drop-profile data
wred dscp
dscp af32 low-limit 50 high-limit 90 discard-percentage 30
#
qos queue-profile queue-profile1
queue 3 drop-profile data
schedule wfq 3 pq 5
#
qos map-table dscp-dscp
input 26 output 0
#
traffic classifier class-ef operator orif-match acl 3001
traffic classifier class-af32 operator or
if-match dscp af32
#
traffic behavior behavior-ef
queue ef bandwidth 10 cbs 250
traffic behavior behavior-af32
queue af bandwidth 30
drop-profile data
traffic behavior behavir-af32
queue af bandwidth 30
#
traffic policy policy-1
classifier class-ef behavior behavior-ef
classifier class-af32 behavior behavior-af32
#
interface Serial1/0/0
link-protocol ppp
ip address 10.0.12.1 255.255.255.0
trust dscp
traffic-policy policy-1 outbound
baudrate 72000
#
interface GigabitEthernet0/0/1
ip address 10.0.145.1 255.255.255.0
trust dscp override
#
ip route-static 10.0.34.0 255.255.255.0 10.0.12.2
#
returnR2
R2display current-configuration
[V200R007C00SPC600]
#
sysname R2
#
interface Serial1/0/0
link-protocol ppp
ip address 10.0.12.2 255.255.255.0
#
interface GigabitEthernet0/0/2
ip address 10.0.34.2 255.255.255.0
#ip route-static 10.0.145.0 255.255.255.0 10.0.12.1
#
returnR3
R3display current-configuration
[V200R007C00SPC600]
#
sysname R3
#
interface GigabitEthernet0/0/2
ip address 10.0.34.3 255.255.255.0
#
ip route-static 0.0.0.0 0.0.0.0 10.0.34.2
#
returnR4
R4display current-configuration
[V200R007C00SPC600]
#
sysname R4
#
interface GigabitEthernet0/0/1
ip address 10.0.145.4 255.255.255.0
#
ip route-static 0.0.0.0 0.0.0.0 10.0.145.1
#
return
R5display current-configuration
[V200R007C00SPC600]
#
sysname R5
#
interface GigabitEthernet0/0/1
ip address 10.0.145.5 255.255.255.0
#
ip route-static 0.0.0.0 0.0.0.0 10.0.145.1
#
returnS3
S3display current-configuration
#!Software Version V200R008C00SPC500
sysname S3
#
interface Vlanif1
ip address 10.0.145.3 255.255.255.0
#
ip route-static 0.0.0.0 0.0.0.0 10.0.145.1
#
nqa test-instance admin udp
test-type udp
destination-address ipv4 10.0.34.4
destination-port 6000
tos 28
frequency 3
interval seconds 1
timeout 1
datasize 5800
start now
nqa test-instance admin jitter
test-type jitter
destination-address ipv4 10.0.34.4
destination-port 6000
tos 46
frequency 3
interval milliseconds 20
timeout 1
datasize 90
start now
#
returnS4
S4display current-configuration
#
!Software Version V200R008C00SPC500
sysname S4
#
interface Vlanif1
ip address 10.0.34.4 255.255.255.0
#
nqa-server udpecho 10.0.34.4 6000
#
ip route-static 0.0.0.0 0.0.0.0 10.0.34.2
#
return至此本文结束大家可以根据此文将QOS基础打夯实
如果本文对你有帮助麻烦点赞、在看有啥问题和建议欢迎在评论区讨论
