worker error, id: 1 Must start tracing before stopping.

[zoe-jobson-IL]

Version info:

    "artillery": "^2.0.21",

Running this command:

Run from GHA

I expected to see this happen:
We recently implemented redis to serve users to our playwright test scripts, at a similar time to that code going in we started to get "worker error, id: 1 Must start tracing before stopping" in the cloud log.
to check if the redis code might be somehow causing this to happen on errors, I ran a test that doesn't use redis as its about creation of new users, I deliberately ran it to the points I got application timeouts and errors, and saw no sign of the "worker error" message.

run without error https://app.artillery.io/share/sh_43ec2f5b8e569a52b63470504fbde0f0a67c504b8e6e159aacced7b49afa841c
most recent run with error
https://app.artillery.io/share/sh_60c9ff04ca44e0efa5255814ee7a4ac4b002f81a0473ff0803dd7883e36fe27a

This is a suggested way of handling users and shared content amongst fargate runners, so I would expect it to work.

I expect errors to be caught and playwright traces produced, I also do not expect all the workers to stop even though they hadn't finished, I would expect the test to continue for a few minutes after users finished being created.

Instead, this happened:

All workers got ""worker error, id: 1 Must start tracing before stopping" at about the same time and stopped even though the test hadn't got to the end point. No playwright traces were saved.

Files being used:

<relevant yaml/js/csv go here>