trex-benchmark-sequence

What the hell?

This is a lil blob to make running repeatable benchmarks and generating consistent graphs easier with Cisco T-Rex.

How use?

- git clone --recursive https://github.com/velocipacktor/trex-benchmark-sequence.git
- go get some coffee. trex-core is big
- npm install (written with node 16.x)
--> ./t-rex-64 -i --astf -c 2 on your t-rex appliance/server
- ./rexbench run -g -r example_full_run sequences/full_run.hjson
- go get another coffee
- edit/create your own sequences / profiles
- ./rexbench run -g sequences/my_sequence.hjson
- ???
- GRAPHS

But how do I setup a t-rex server?

You'll need a system (vm probably works too, but lower perf of course) with at least 3 nics and 4 cores
Put the 1st nic on your lab/mgmt/whatever network
Physically loopback the other 2

copy-pasteables (as root):

mkdir -p /opt/trex
cd /opt/trex
wget --no-cache https://trex-tgn.cisco.com/trex/release/v2.99.tar.gz
tar -xzvf v2.99.tar.gz
cd v.299

  -> list interfaces
./dpdk_setup_ports.py -s

  -> generate config file
./dpdk_setup_ports.py -i
  -> select MAC/Layer2 mode, select your two loopbacked nics
  -> come back later when you (1: know what your hardware can do and 2: have a DUT) and select IP/Layer3 mode

./t-rex-64 -i --astf

see here for more info on install

Help?

Maybe. Do a google first, open an issue second.

Why?

I was only able to find one mention on the entire internet about generating graphs from t-rex.
iirc it was an old Cisco Live thing where they were like "yeah lol we linked the mq to elk and used grafana"

So. I want graphs. So here we are.

Why nodejs?

Because I suck at python and wanted to use a lang I'm comfy with.
This is basically lab equipment, not production code.