Open tonytonyjan opened 3 years ago
I found an extreme weird behavior when checking how many times wrk would execute the script:
When reading script from file, it executes 2 times, which is unexpected:
bash-3.2$ cat clock.lua print(os.clock()) bash-3.2$ wrk -s clock.lua -t 1 https://example.com 0.004581 0.011374 Running 10s test @ https://example.com 1 threads and 10 connections Thread Stats Avg Stdev Max +/- Stdev Latency 133.24ms 7.33ms 316.11ms 98.47% Req/Sec 74.74 20.84 101.00 64.89% 720 requests in 10.05s, 1.10MB read Requests/sec: 71.63 Transfer/sec: 112.26KB
When using process substitution by <(), it only execute once, which is expected.
<()
bash-3.2$ wrk -s <(echo 'print(os.clock())') -t 1 https://example.com 0.004722 Running 10s test @ https://example.com 1 threads and 10 connections Thread Stats Avg Stdev Max +/- Stdev Latency 134.98ms 4.66ms 153.82ms 84.18% Req/Sec 73.72 18.29 101.00 71.28% 708 requests in 10.01s, 1.08MB read Requests/sec: 70.70 Transfer/sec: 110.68KB
I found an extreme weird behavior when checking how many times wrk would execute the script:
When reading script from file, it executes 2 times, which is unexpected:
When using process substitution by
<(), it only execute once, which is expected.Environment