RE: I am Trex user, I got 2 questions need help.

Hanoch Haim (hhaim)

unread,

Jun 12, 2017, 10:10:15 AM6/12/17

to Song Lifeng(宋丽风), trex...@lists.fd.io, trex...@googlegroups.com

BTW, our active forum is this mailer trex...@googlegroups.com

Thanks,

Hanoh

From: Hanoch Haim (hhaim)
Sent: Monday, June 12, 2017 5:07 PM
To: 'Song Lifeng(宋丽风)'; trex...@lists.fd.io
Subject: RE: I am Trex user, I got 2 questions need help.

Hi Song,

See inline

Hanoh

From: trex-dev...@lists.fd.io [mailto:trex-dev...@lists.fd.io] On Behalf Of Song Lifeng(???)
Sent: Monday, June 12, 2017 12:56 PM
To: trex...@lists.fd.io
Subject: [trex-dev] I am Trex user, I got 2 questions need help.

Hi Trex:

I am Trex user, this tool is great, but I got 2 questions need help.

1. ipackets more then opackets when I test throughput follow RFC2544, there are some other packets were receive, Does it support Packets filter or How to solve this problem?

[hh] not sure what is your setup configuration and the rates you try to achive. “Port” statistic count all the packets usualy when doing NDR drop of 0.01% is valid so counting a few packets more or less is OK.

In case you want to filter this you can use flow-stats see here how to use and the limitation

https://trex-tgn.cisco.com/trex/doc/trex_stateless.html#_tutorial_per_stream_statistics

2. My DUT support PPPoE Client, I need test PPPoE Throughput follow RFC2544, but I can’t resolve this problem, Does it support PPPoE Server or How does test PPPoE performance use Trex

[hh] using “services” mode you can simulate PPPoE “clients” and then generate traffic that match the clients IP. See a blog on this one with DHCP example

https://communities.cisco.com/community/developer/trex/blog/2017/05/04/using-trex-services-to-enhance-your-tests

we will be happhy to get contribution of this PPPoE service

I’m lookong forward to your early reply. Thank You Very much.

Best regards

From Song

本电子邮件（包括任何的附件）为本公司保密文件。本文件仅仅可为以上指定的收件人或公司使用，如果阁下非电子邮件所指定之收件人，那么阁下对该邮件部分或全部的泄漏、阅览、复印、变更、散布或对邮件内容的使用都是被严格禁止的。如果阁下接收了该错误传送的电子邮件，敬请阁下通过回复该邮件的方式立即通知寄件人，同时删除你所接收到的文本。 This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.

Song Lifeng(宋丽风)

unread,

Jun 14, 2017, 10:43:16 PM6/14/17

to Hanoch Haim (hhaim), trex...@lists.fd.io, trex...@googlegroups.com

Hi Hanoh,

Thank for your feedback.

For my second question, I need describe it clealy.

My TRex be installed Dell’s Server and OS is Ubuntu16.04.

Test topology is: TRex A port ----connected by ETH cable----DUT------ connected by ETH cable-------TRex B port

DUT is Router, e.g. Netgear R6250 . DUT support PPPoE Client. So I needed Trex B port support PPPoE server. But I have no idea and I need help.

Thank You Very much.

Best regards

发件人: Hanoch Haim (hhaim) [mailto:hh...@cisco.com]
发送时间: 2017年6月12日星期一 22:10
收件人: Song Lifeng(宋丽风); trex...@lists.fd.io
抄送: trex...@googlegroups.com
主题: RE: I am Trex user, I got 2 questions need help.

Hanoch Haim (hhaim)

unread,

Jun 15, 2017, 2:27:09 AM6/15/17

to Song Lifeng(宋丽风), trex...@lists.fd.io, trex...@googlegroups.com

In that case you can implement PPPoE server as a TRex service, initiate all the clients from DUT and then run RFC2544

Have a look into the link I’ve sent

Thanks,

Hanoh

Song Lifeng(宋丽风)

unread,

Jun 15, 2017, 2:34:42 AM6/15/17

to Hanoch Haim (hhaim), trex...@lists.fd.io, trex...@googlegroups.com

Hi Hanoh,

Thank you for your support. I will be try again.

Thanks

Song

Best regards

发件人: Hanoch Haim (hhaim) [mailto:hh...@cisco.com]

发送时间: 2017年6月15日星期四 14:27

Hanoch Haim (hhaim)

unread,

Jun 15, 2017, 2:46:31 AM6/15/17

to Song Lifeng(宋丽风), trex...@lists.fd.io, trex...@googlegroups.com

If I understand you correctly, you need to support server that will handle *one* client (DUT)

Is it correct?

Song Lifeng(宋丽风)

unread,

Jun 15, 2017, 2:56:47 AM6/15/17

to Hanoch Haim (hhaim), trex...@lists.fd.io, trex...@googlegroups.com

Hi Hanoh,

Yes, you are right.

Song

Best regards

发件人: Hanoch Haim (hhaim) [mailto:hh...@cisco.com]

发送时间: 2017年6月15日星期四 14:46

Song Lifeng(宋丽风)

unread,

Jun 16, 2017, 6:18:40 AM6/16/17

to Song Lifeng(宋丽风), Hanoch Haim (hhaim), trex...@lists.fd.io, trex...@googlegroups.com

Hi Hanoh,

The ‘Behind the hood’ section said “TRex services are implemented using SimPy “ on the link you sent to me .

so my question is, does it have to use Simpy when I want to create pppoe server service instances following the diagram ?

Thanks,

Song

Hanoch Haim (hhaim)

unread,

Jun 16, 2017, 6:29:32 AM6/16/17

to Song Lifeng(宋丽风), trex...@lists.fd.io, trex...@googlegroups.com

SimPy is running under the hood. You don’t need to know what it

Just follow the example of DHCP, you need to create state-machine that implement PPoE server for one client.

Something like that

Wait_for_ppoe_request

Answer_ppoe_response

Another example is ARP service,,in this example it is client side (REQ/RES)

https://github.com/cisco-system-traffic-generator/trex-core/blob/master/scripts/automation/trex_control_plane/stl/trex_stl_lib/services/trex_stl_service_arp.py

Thanks,

Hanoh

From: Song Lifeng(宋丽风) [mailto:songl...@twsz.com]
Sent: Friday, June 16, 2017 1:19 PM
To: Song Lifeng(宋丽风); Hanoch Haim (hhaim); trex...@lists.fd.io
Cc: trex...@googlegroups.com
Subject: 答复: I am Trex user, I got 2 questions need help.

Hi Hanoh,

Song Lifeng(宋丽风)

unread,

Jun 16, 2017, 6:45:04 AM6/16/17

to Hanoch Haim (hhaim), trex...@lists.fd.io, trex...@googlegroups.com

Get it. Thank you very much

Thanks,

Song

发件人: Hanoch Haim (hhaim) [mailto:hh...@cisco.com]

发送时间: 2017年6月16日星期五 18:29

收件人: Song Lifeng(宋丽风); trex...@lists.fd.io
抄送: trex...@googlegroups.com

主题: RE: I am Trex user, I got 2 questions need help.

Song Lifeng(宋丽风)

unread,

Jun 19, 2017, 8:48:53 AM6/19/17

to Song Lifeng(宋丽风), Hanoch Haim (hhaim), trex...@lists.fd.io, trex...@googlegroups.com

Hi Hanoh：

As following the diagram. Drop_rate is always 0 during the 60s testing.

I don’t understand why the total_rx always less than total_tx_L2 ？

Also ipackets less than opackets in ‘Port_Statistics’ section .

On more question : what’s the mean ‘queue_full ‘ ?

Thanks,

Song

发件人: Hanoch Haim (hhaim) [mailto:hh...@cisco.com]

发送时间: 2017年6月16日星期五 18:29

收件人: Song Lifeng(宋丽风); trex...@lists.fd.io
抄送: trex...@googlegroups.com

主题: RE: I am Trex user, I got 2 questions need help.

Song Lifeng(宋丽风)

unread,

Jun 19, 2017, 8:52:42 AM6/19/17

to Song Lifeng(宋丽风), Hanoch Haim (hhaim), trex...@lists.fd.io, trex...@googlegroups.com

Hi Hanoh：

As following the diagram. Drop_rate is always 0 during the 60s testing.

I don’t understand why the total_rx always less than total_tx_L2 ？

Also ipackets less than opackets in ‘Port_Statistics’ section .

Thanks,

Song

发件人: Hanoch Haim (hhaim) [mailto:hh...@cisco.com]

发送时间: 2017年6月16日星期五 18:29

收件人: Song Lifeng(宋丽风); trex...@lists.fd.io
抄送: trex...@googlegroups.com

主题: RE: I am Trex user, I got 2 questions need help.

Yaroslav Brustinov (ybrustin)

unread,

Jun 19, 2017, 9:20:42 AM6/19/17

to Song Lifeng(宋丽风), Hanoch Haim (hhaim), trex...@lists.fd.io, trex...@googlegroups.com

Hi,

Under traffic drop rate is approximation.

You need to stop traffic to verify drops.

Thanks,

Yaroslav.

Hanoch Haim (hhaim)

unread,

Jun 19, 2017, 9:41:02 AM6/19/17

to Song Lifeng(宋丽风), trex...@lists.fd.io, trex...@googlegroups.com

Hi Song,

First for 1gbps you don’t need 4 cores. You can have 1 core (-c 1) and it can generate 15-30MPPS per core

Queue_full means that the NIC can’t handle the rate (maybe due to hw flow-control from the DUT)

Reduce the speed in this case

Hanoch Haim (hhaim)

unread,

Jun 19, 2017, 9:43:12 AM6/19/17

to Song Lifeng(宋丽风), trex...@googlegroups.com

Hanoh

From: Hanoch Haim (hhaim)
Sent: Monday, June 19, 2017 4:41 PM
To: 'Song Lifeng(宋丽风

)'; trex...@lists.fd.io

Cc: trex...@googlegroups.com
Subject: RE: I am Trex user, I got 2 questions need help.

Hi Song,

First for 1gbps you don’t need 4 cores. You can have 1 core (-c 1) and it can generate 15-30MPPS per core

Queue_full means that the NIC can’t handle the rate (maybe due to hw flow-control from the DUT)

Reduce the speed in this case

Thanks,

Song Lifeng(宋丽风)

unread,

Jun 19, 2017, 9:58:09 AM6/19/17

to Hanoch Haim (hhaim), trex...@googlegroups.com

Hi Hanoch，

About queue_full，Which NIC？ TX port or RX port？

Thanks，

Song

在 2017年6月19日，21:43，Hanoch Haim (hhaim) <hh...@cisco.com> 写道：

Hanoh

From: Hanoch Haim (hhaim)
Sent: Monday, June 19, 2017 4:41 PM
To: 'Song Lifeng(宋丽风)'; trex...@lists.fd.io
Cc: trex...@googlegroups.com
Subject: RE: I am Trex user, I got 2 questions need help.

Hi Song,

First for 1gbps you don’t need 4 cores. You can have 1 core (-c 1) and it can generate 15-30MPPS per core

Queue_full means that the NIC can’t handle the rate (maybe due to hw flow-control from the DUT)

Reduce the speed in this case

Thanks,

Hanoh

From: Song Lifeng(宋丽风) [mailto:songl...@twsz.com]
Sent: Monday, June 19, 2017 3:49 PM
To: Song Lifeng(宋丽风); Hanoch Haim (hhaim); trex...@lists.fd.io
Cc: trex...@googlegroups.com
Subject: 答复: I am Trex user, I got 2 questions need help.

Hi Hanoh：

As following the diagram. Drop_rate is always 0 during the 60s testing.

I don’t understand why the total_rx always less than total_tx_L2 ？

Also ipackets less than opackets in ‘Port_Statistics’ section .

<image001.png>

On more question : what’s the mean ‘queue_full ‘ ?

<image002.png>

Thanks,

Song

发件人: Hanoch Haim (hhaim) [mailto:hh...@cisco.com]
发送时间: 2017年6月16日星期五 18:29
收件人: Song Lifeng(宋丽风); trex...@lists.fd.io
抄送: trex...@googlegroups.com
主题: RE: I am Trex user, I got 2 questions need help.

SimPy is running under the hood. You don’t need to know what it

Just follow the example of DHCP, you need to create state-machine that implement PPoE server for one client.

Something like that

Wait_for_ppoe_request

Answer_ppoe_response

Another example is ARP service,,in this example it is client side (REQ/RES)

https://github.com/cisco-system-traffic-generator/trex-core/blob/master/scripts/automation/trex_control_plane/stl/trex_stl_lib/services/trex_stl_service_arp.py

Thanks,

Hanoh

From: Song Lifeng(宋丽风) [mailto:songl...@twsz.com]
Sent: Friday, June 16, 2017 1:19 PM
To: Song Lifeng(宋丽风); Hanoch Haim (hhaim); trex...@lists.fd.io
Cc: trex...@googlegroups.com
Subject: 答复: I am Trex user, I got 2 questions need help.

Hi Hanoh,

The ‘Behind the hood’ section said “TRex services are implemented using SimPy “ on the link you sent to me .

so my question is, does it have to use Simpy when I want to create pppoe server service instances following the diagram ?

<image003.png>

Thanks,

Song

本电子邮件（包括任何的附件）为本公司保密文件。本文件仅仅可为以上指定的收件人或公司使用，如果阁下非电子邮件所指定之收件人，那么阁下对该邮件部分或全部的泄漏、阅览、复印、变更、散布或对邮件内容的使用都是被严格禁止的。如果阁下接收了该错误传送的电子邮件，敬请阁下通过回复该邮件的方式立即通知寄件人，同时删除你所接收到的文本。 This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.

Hanoch Haim (hhaim)

unread,

Jun 19, 2017, 9:58:26 AM6/19/17

to Song Lifeng(宋丽风), trex...@googlegroups.com

Tx

Song Lifeng(宋丽风)

unread,

Jun 19, 2017, 10:26:52 AM6/19/17

to Hanoch Haim (hhaim), trex...@googlegroups.com

Hi Hanoch,

About 1st question:

Drop_rate is always 0 from start to finish.

ipackets less than opackets in 'Statistics' section during in the testing , wait until the testing stop, ipackets less than opackets too. Usually，DUT have packet dropout， Drop _rate is not 0.

I can't understand why drop _rate is 0.

Hanoch Haim (hhaim)

unread,

Jun 19, 2017, 12:15:58 PM6/19/17

to Song Lifeng(宋丽风), trex...@googlegroups.com

The drop rate is just an estimation, it could be zero still there could be a drops

You should use the total counters at the end of the test.

Hanoh

From: Song Lifeng(宋丽风) [mailto:songl...@twsz.com]

Sent: Monday, June 19, 2017 5:27 PM
To: Hanoch Haim (hhaim)
Cc: trex...@googlegroups.com

Song Lifeng(宋丽风)

unread,

Jun 20, 2017, 5:06:20 AM6/20/17

to Song Lifeng(宋丽风), Hanoch Haim (hhaim), trex...@lists.fd.io, trex...@googlegroups.com

Hi Hanoch,

I power off DUT1 and power on DUT2. Trex report a error when then I restart TRex tool. How to recover from this error.

----------------------------------------------------------------------------------------------------------------------------------------------------------------

Logs:

root@dell-1:/home/trex/v2.26# ./t-rex-64 -i --cfg tw/slf_stateless_br.yaml

Starting Scapy server.... Scapy server is started

The ports are bound/configured.

Starting TRex v2.26 please wait ...

EAL: Error - exiting with code: 1

Cause: Requested device 0000:01:00.3 cannot be used

root@dell-1:/home/trex/v2.26# dpkg-s

dpkg-split dpkg-statoverride

root@dell-1:/home/trex/v2.26# ./dpdk_setup_ports.py

Trying to bind to igb_uio ...

/usr/bin/python dpdk_nic_bind.py --bind=igb_uio 0000:03:00.0 0000:03:00.1

The ports are bound/configured.

root@dell-1:/home/trex/v2.26#

-------------------------------------------------------------------------------------------------------------------------------------------------------------

Thanks,

Song

Song Lifeng(宋丽风)

unread,

Jun 20, 2017, 5:20:14 AM6/20/17

to Song Lifeng(宋丽风), Hanoch Haim (hhaim), trex...@lists.fd.io, trex...@googlegroups.com

Hi Hanoch,

So sorry , I lost some message.

----------------------------------------------------------------------------------------------------------------------------------------------------------------

Logs:

root@dell-1:/home/dpdk/dpdk-stable-16.11.1/tools# ./dpdk-setup.sh

……(Omit some message)

Option: 21

Network devices using DPDK-compatible driver

============================================

0000:01:00.2 'I350 Gigabit Network Connection' drv=igb_uio unused=igb

0000:01:00.3 'I350 Gigabit Network Connection' drv=igb_uio unused=igb

0000:03:00.0 'Ethernet 10G 2P X520 Adapter' drv=igb_uio unused=ixgbe

0000:03:00.1 'Ethernet 10G 2P X520 Adapter' drv=igb_uio unused=ixgbe

Network devices using kernel driver

===================================

0000:01:00.0 'I350 Gigabit Network Connection' if=eno1 drv=igb unused=igb_uio *Active*

0000:01:00.1 'I350 Gigabit Network Connection' if=eno2 drv=igb unused=igb_uio *Active*

Other network devices

=====================

-------------------------------------------------------------------------------------------------------------------------------------------------------------

Thanks,

Song

发件人: Song Lifeng(宋丽风)
发送时间: 2017年6月20日星期二 17:06
收件人: Song Lifeng(宋丽风); 'Hanoch Haim (hhaim)'; 'trex...@lists.fd.io'
抄送: 'trex...@googlegroups.com'
主题: 答复: I am Trex user, I got 2 questions need help.

Hi Hanoch,

I power off DUT1 and power on DUT2. Trex report a error when I restart TRex tool. How to recover from this error.

----------------------------------------------------------------------------------------------------------------------------------------------------------------

Logs:

root@dell-1:/home/trex/v2.26# ./t-rex-64 -i --cfg tw/slf_stateless_br.yaml

Starting Scapy server.... Scapy server is started

The ports are bound/configured.

Starting TRex v2.26 please wait ...

EAL: Error - exiting with code: 1

Cause: Requested device 0000:01:00.3 cannot be used

root@dell-1:/home/trex/v2.26# dpkg-s

dpkg-split dpkg-statoverride

root@dell-1:/home/trex/v2.26# ./dpdk_setup_ports.py

Trying to bind to igb_uio ...

/usr/bin/python dpdk_nic_bind.py --bind=igb_uio 0000:03:00.0 0000:03:00.1

The ports are bound/configured.

root@dell-1:/home/trex/v2.26#

-------------------------------------------------------------------------------------------------------------------------------------------------------------

Thanks,

Song

Hanoch Haim (hhaim)

unread,

Jun 20, 2017, 6:00:37 AM6/20/17

to Song Lifeng(宋丽风), trex...@lists.fd.io, trex...@googlegroups.com

could you send output of this

./t-rex-64 -i --cfg tw/slf_stateless_br.yaml –v 7

Hanoh

Song Lifeng(宋丽风)

unread,

Jun 20, 2017, 6:13:14 AM6/20/17

to Hanoch Haim (hhaim), trex...@lists.fd.io, trex...@googlegroups.com

Hi Hanoch,

The attach document is log.

By the way: If eno3 and eno4 was bound, eno1 connect to SW by Ethernet cable, Which IP address input in GUI ‘ Connections’ text boxes or other values？

Best regards

产品技术中心&测试部宋丽风

T&W深圳市共进电子股份有限公司

SHENZHEN GONGJIN ELECTRONICS CO.,LTD

中国广东深圳市南山区蛇口南海大道1019号南山医疗器械园二、三、四、五楼

2F/3F/4F/5F Baiying Building, 1019#Nanhai RD, Nanshan Dist., Shenzhen, Guangdong,China

直线(TEL)：(86-755)26681222 分机(EXT)：10113

发件人: Hanoch Haim (hhaim) [mailto:hh...@cisco.com]

发送时间: 2017年6月20日星期二 18:00

收件人: Song Lifeng(宋丽风); trex...@lists.fd.io
抄送: trex...@googlegroups.com

主题: RE: I am Trex user, I got 2 questions need help.

log.txt

Song Lifeng(宋丽风)

unread,

Jun 29, 2017, 10:53:04 AM6/29/17

to Hanoch Haim (hhaim), trex...@googlegroups.com, Liu Shengwen(刘圣文), Huang Wenjun(黄文君)

Hi Hanoh，

I have a pppoed.py that be tested ok. I want to use this pppoed.py in trex tool.

I followed this guide. Guide link is https://communities.cisco.com/community/developer/trex/blog/2017/05/04/using-trex-services-to-enhance-your-tests

Add some code to trex_stl_client.py whick path is /home/trex/v2.26/automation/trex_control_plane/stl/trex_stl_lib/

I get the request on port 1, But have no response when I test. What is wrong with it ?

Code:

By the way : My topology is TREX PORT 0 ------ETH cable --------[DUT LAN PORT](DUT is Route) [ DUT WAN PORT is PPPOE Client]-------------ETH cable --------- TREX PORT 1

Thanks，

Song

发件人: Hanoch Haim (hhaim) [mailto:hh...@cisco.com]

发送时间: 2017年6月19日星期一 21:58
收件人: Song Lifeng(宋丽风)

Yaroslav Brustinov

unread,

Jun 29, 2017, 2:37:38 PM6/29/17

to Song Lifeng(宋丽风), Hanoch Haim (hhaim), trex...@googlegroups.com, Liu Shengwen(刘圣文), Huang Wenjun(黄文君)

The Simpy is used to gain pseudo-parallelism of many state machines.

As you read from blog, it's being used under the hood, and as Hanoch said, you do not have to deal with those low level details.

Basic usage for services is to define some filter for your state machine (UDP src port for example), and to use following functions:

async_wait_for_pkt

async_tx_pkt

async_wait

Look at the examples provided.

--
You received this message because you are subscribed to the Google Groups "TRex Traffic Generator" group.
To unsubscribe from this group and stop receiving emails from it, send an email to trex-tgn+unsubscribe@googlegroups.com.
To post to this group, send email to trex...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/trex-tgn/282CE65D74D8974DBAB13B4C0304A14D3286692C%40mailbox1.gj.com.
For more options, visit https://groups.google.com/d/optout.

Song Lifeng(宋丽风)

unread,

Aug 1, 2017, 4:57:11 AM8/1/17

to Hanoch Haim (hhaim), trex...@googlegroups.com, Liu Shengwen(刘圣文), Huang Wenjun(黄文君), Yuan Liangwei(袁良伟)

Hi hhaim,

I have 2 question need help.

Q1: Is there any other way Statistics pkts ? Such as ETH Mac / UDP or TCP Port / IP address ….

Because RFC2544 thoughput test require drop pkts must be zero.

Rx more than tx when I test . following is test log：

trex> clear

trex>start -f tw/udp.py --force -p 0 -m 500.5mbps -t size=64,vm=cached -d 60

Port 0 job done

trex>stats -x

Name: | Port 0: | Port 1:

------------------------------------------------------------------

rx_good_packets | 0 | 298016

tx_good_packets | 298040 | 0

rx_good_bytes | 0 | 17880768

tx_good_bytes | 17882336 | 0

rx_size_64_packets | 0 | 58652384

rx_broadcast_packets | 0 | 37

rx_total_packets | 0 | 58652384

tx_total_packets | 58652344 | 0

rx_total_bytes | 0 | 3519143040

tx_total_bytes | 3519140640 | 0

tx_size_64_packets | 58652344 | 0

trex> clear

trex>start -f tw/udp.py --force -p 0 -m 937.6mbps -t size=64,vm=cached -d 60

2017-07-27 09:03:43 - [server][info] - Port 0 job done

trex>stats -x

Name: | Port 0: | Port 1:

------------------------------------------------------------------

rx_good_packets | 0 | 26650

tx_good_packets | 26674 | 0

rx_good_bytes | 0 | 1598872

tx_good_bytes | 1600312 | 0

rx_size_64_packets | 0 | 109875045

rx_broadcast_packets | 0 | 45

rx_total_packets | 2 | 109875045

tx_total_packets | 109875001 | 1

rx_total_bytes | 120 | 6592502700

tx_total_bytes | 6592500060 | 60

tx_size_64_packets | 109875001 | 1

tx_broadcast_packets | 1 | 1

trex>clear

Q2: cpu to high when 64B tx rate is 1Gbps , whether it will affect result ? the pkts in Queue_full is be sent when I stop test ?

I think the answer is No .but I need trex expert Reconfirm. Thank you !

Look forward to your reply。

From Song

Yaroslav Brustinov

unread,

Aug 1, 2017, 9:21:17 AM8/1/17

to Song Lifeng(宋丽风), Hanoch Haim (hhaim), trex...@googlegroups.com, Liu Shengwen(刘圣文), Huang Wenjun(黄文君), Yuan Liangwei(袁良伟)

Hi,

1) You can use flow stats to get statistics only on sent traffic.

https://trex-tgn.cisco.com/trex/doc/trex_stateless.html#_tutorial_per_stream_statistics

2) You are probably using multiplies 1gbps or similar, which specifies L2 rate of traffic, the linerate of NIC is in terms of L1.

Try using multiplier with "l1" added or percentage of linerate (look at help of start command)

Thanks,

Yaroslav.

--

You received this message because you are subscribed to the Google Groups "TRex Traffic Generator" group.
To unsubscribe from this group and stop receiving emails from it, send an email to trex-tgn+unsubscribe@googlegroups.com.
To post to this group, send email to trex...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/trex-tgn/282CE65D74D8974DBAB13B4C0304A14D32874E1B%40mailbox1.gj.com.

Reply all

Reply to author

Forward