Why are my parallel virtual users taking so long to complete?

Grafana k6 OSS Support
1

Hi, everyone.

I have a question about parallel vus. I have really huge script with a lot of requests. I use these options:

export const options = {
    vus: 500,
    iterations: 500,
    httpDebug: "full"
};

One virtual user is taking between 20-25 seconds. But if I start 500 vus parallel they works for 5 minutes at least, so why 500 vus is working for so long, is it a problem with long responses from our webapp or I’m missing some options.

We also used groups in the script and I saw that before all 500 users finished the first group, the second group would not start (the same situation with all groups in the script).

Thanks in advance for your help.

2

Hello!

You’ll definitely want to have a look at your response times. A VU will proceed to the next group as soon as the previous one has finished.

There’s not enough information in what you’ve shared to provide a more complete diagnosis, but I will say that you should only be using httpDebug: “full” during script debugging and not running with 500 VUs… that’s a lot of data going to stdout and that in itself may be slowing down k6!