Where should simulating load test scripts/workflows live?

[pzrq]

@aherlihy Developed some Python scripts that simulate load and thus help test the real-time nature of the RTSS graphs:

• https://gist.github.com/aherlihy/87c94f1b5051d8d2ef5ce39a5c9ba406 and
• https://gist.github.com/aherlihy/3e187ca66a9b6457e30c965f09971d54

The Compass Sydney Team recently found them useful as part of testing the COMPASS 1.5 Feature Matrix.

My initial thoughts - options are:

1. Retain the existing manual workflow and just document that they exist, e.g. in the README?
2. Convert them to node/JS to work with the npm run storybook ? Or...
3. Is it at such a different level that you'd need to mock or capture/replay some raw data?
4. If you're actually at that level, should we write a fuzzer or something similar like the server has done? Would that be likely to yield better enzyme or similar tests?
5. Any other suggestions?