present setup() function breaks iteration_duration metric

[dvob]

Brief summary

If a setup function is present the iteration_duration metrics are wrong.
It seems that the time used to run setup() is incorporated into the iteration_duration metric.

Maybe related to #2718

k6 version

v0.51.0

OS

Linux

Docker version and image (if applicable)

No response

Steps to reproduce the problem

• script.js

import { sleep } from 'k6';

export function setup() {
}

export default function() {
  sleep(1);
}

Expected behaviour

iteration_duration metrics (avg, min, med, max, p(90), p(95)) are all 1s.

     data_received........: 0 B 0 B/s
     data_sent............: 0 B 0 B/s
     iteration_duration...: avg=1s min=1s med=1s max=1s p(90)=1s p(95)=1s
     iterations...........: 1   0.999474/s
     vus..................: 1   min=1      max=1
     vus_max..............: 1   min=1      max=1

Actual behaviour

iteration_duration min shows a very low value:

     data_received........: 0 B 0 B/s
     data_sent............: 0 B 0 B/s
     iteration_duration...: avg=500.52ms min=1.67µs med=500.52ms max=1s p(90)=900.94ms p(95)=950.99ms
     iterations...........: 1   0.998617/s
     vus..................: 1   min=1      max=1
     vus_max..............: 1   min=1      max=1

The time used to run setup() is incorporated into the iteration_duration metric.
If put a sleep(1) into the setup function the metric appears correct.

[codebien]

Hey @dvob,
thanks for reporting it.

We are aware of it, you can check details in #3377 (comment) and #1605.

I'm going to close as it is a duplicate.

[dvob]

@codebien thank you for pointing me to the right issue