Session based high bandwidth throughput testing. - - PowerPoint PPT Presentation

▶

Jun 04, 2023 491 likes •678 views

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001 Session based high bandwidth throughput testing.

SLIDE 1

Session based high bandwidth throughput testing.

RP2

System and Network Engineering System and Network Engineering

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001 010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

SLIDE 2

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Research Question

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

SLIDE 3

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

UDP vs TCP

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

UDP connectionless 8 byte header no ordering lightweight TCP connection oriented 20 byte header

rdering

heavyweight reliable congestion control

SLIDE 4

T h e C h a n n e l

Research Question

What is needed to perform high bandwidth session based throughput tests and how to go beyond pure network infrastructure testing?

010101010101101010101110110100010010110111101101010101001001010010100100010010010100101010100101001010100101010010101001010100101010 100101010101010101001010100101001010101010101011100101011011000100101011010101101100011001101001001001110100101001010010101001001001

The term ”high bandwidth” references to at least 40Gb/s. The term ”session based” references to TCP traffic.

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

SLIDE 5

Test environment

5 010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

A B C E D S

40Gb/s 100Gb/s 40Gb/s 100Gb/s 40Gb/s

MGMT M

100Gb/s 1Gb/s 1Gb/s 1Gb/s 1Gb/s Vlan100: 10.10.10.0/24 Vlan 66: 10.60.66.0/24

Internet

4x10Gb/s

SLIDE 6

Tools

6 010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

BoNeSi pktgen-kernel MoonGen pktgen-DPDK

SLIDE 7

First important result

7 010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

DPDK is the way to go

SLIDE 8

Use cases

8 010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

NR Use case DUT Explanation UC1 Bandwidth generation Client The goal is to see if the client is capable of filling up the link and to reach the maximum amount of pps UC2 Throughput Switch/Router Generate the maximum amount of bidirectional data to make sure the intermediate hardware is able to forward  at line rate UC3 Session per second Client/Server Get the hardware limitations of the end hosts. UC4 Application Server and intermediate devices The clients will try to overload the server with requests at application level

SLIDE 9

Experiment UC1 bandwidth

9 010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

SLIDE 10

Experiment UC1 PPS

10 010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

SLIDE 11

Experiment UC3 HTTP link usage

11 010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

SLIDE 12

Experiment UC3 sessions

12 010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

SLIDE 13

Real world test

13 010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

AE112 4x10Gb/s AE113 4x10Gb/s Router1B Router1A Firewall1 Firewall2 DC1-1A DC1-1B DC2-1A DC2-1B

SURFnet

VCP 5x10Gb/s

1 G b/s 40Gb/s 4 G b/s 4 G b/s 40Gb/s SERVER A 40Gb/s 40Gb/s 40Gb/s 40Gb/s 2x40Gb/s Active - passive 40Gb/s 2x40Gb/s AE0 AE0

sync

1 G b

/ s Router 4x10Gb/s QFX10k2

CLIENT 40Gb/s

A P

SLIDE 14

Real world result - bandwidth

Student: Bram ter Borch

NGINX RAW TCP HTTP

SLIDE 15

Conclusion

15 010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

DPDK is the way to go for high bandwidth session based throughput testing The use cases are valid The weakest link in a path can be found Monitoring is very important

SLIDE 16

Future work

16 010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

DPDK tests have to be run using 100Gb/s interfaces IPv6 support has to be added to WARP Other layer 7 protocols have to be added to WARP

SLIDE 17

Thank you

Questions

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001

Student: Bram ter Borch

010101010101101010101110110100010010110111101101100101011011000100101011010101101100011001101001001001110100101001010010101001001001