My experience using loom for a load generator

Wed May 27 10:07:25 UTC 2020

> It might be interesting to compare your results with wrk/2?
I ran wrk2 in WSL, compiled using CFLAGS -march=native and -mtune=native. It reached 13484 requests per second on average (while targeting 20,000), with a standard deviation of 1.1027%.
Note that I could only get 100 threads working for wrk2, in contrast to the 1,000/100,000 my tool simulated.

Kind regards

Eric Ackermann