Locust --processes argument generating multiple html reports

[Hchanni]

Prerequisites

• I am using the latest version of Locust
• I am reporting a bug, not asking a question

Description

I'm running Locust in a docker container with the command within ECS, and it works well, I push the html report contents to S3, but for some reason, it's providing me with 5 files instead of just one. It seems that I'm getting a file for each worker, and then the master aswell.

Command line

locust -f ./locustfile.py --headless --processes 4 --users 150 --spwan-rate 150 --run-time 15s --host --html /tmp/report.html

Locustfile contents

Just a very basic hello world api endpoint.

Python version

3.11

Locust version

Locust docker image 2.24.0

Operating system

Linux

[cyberw]

Interesting. Is it not the same if you launch worker processes individually?

[Hchanni]

Interesting. Is it not the same if you launch worker processes individually?

When you mean launch worker processes individually do you mean in it's own seperate ECS task?

It's really weird because I run the Locust script through direct Github Actions with the html report essentially the exact same command above, and I only get one html report.

I also have run a distributed environment where worker tasks run in a separate ECS environment, and that runs with the UI and I can download the one specific HTML script.

[cyberw]

Interesting. Is it not the same if you launch worker processes individually?

When you mean launch worker processes individually do you mean in it's own seperate ECS task?

I mean by running locust … --worker 4 times instead of using --processes. I dont know that much about ecs tasks :) maybe run locally on your own machine first to remove it from the equation.

It's really weird because I run the Locust script through direct Github Actions with the html report essentially the exact same command above, and I only get one html report.

Weird indeed…

[Hchanni]

I've decided to go the distributed way of using --expect_workers and then spawning up multiple workers.