Kolibri Studio load testing experiment

Expected behaviour

• None of the requests will return a 500, 502 or 503 error
• None of the requests will take longer than 20 seconds to return a response

Steps to run

To run against your local studio instance, simply run:

$ make run

This pings your localhost:8080 server with 100 clients and a 20 users/sec hatch rate. You can set each of these parameters for a run.

You can also specify the duration to run the tests for, along with the results file(s), using the following parameters:

make run TIME_ARG="-t5m" RESULTS_FILE="--csv=profile_results"

To run this script against the Studio staging server, change the URL parameter:

$ make run URL=https://develop.studio.learningequality.org
USERNAME=<yourusername> PASSWORD=<yourpassword>

To customize the number of clients and the hatch rate, set the NUM_CLIENTS and the HATCH_RATE respectively:

$ make run NUM_CLIENTS=5 HATCH_RATE=1

After stopping it, Locust slaves must be stopped:

$ make stop_slaves