curl IP/stop does not actually terminate Locust swarms

Expected Per suggestion in #831, I have a script that starts Locust through a POST request, then terminates Locust through a GET /stop.

There are two issues with this:

1. Depending on the spawn_rate, calling GET /stop only terminates httpusers that are already spawned. I expected this would terminate the entire test’s users.
2. If you initiate a swarm with a HTTP request, the UI doesn’t have a “STOP” button. The only way to terminate that swarm is with a GET /stop.

To repro

1. curl -XPOST -d "user_count=100&spawn_rate=10" http://$LOADGEN_IP/swarm where user_count/spawn_rate are our custom fields
2. sleep for 5 seconds, then curl http://$LOADGEN_IP/stop
3. Notice, after running step 2 UI indicates STATUS: STOPPED, and HTTPResponse is successful:

{
  "message": "Test stopped",
  "success": true
}

1. However, according to our reliable telemetry, Locust is still swarming

Environment

• Locust version: 1.3.1

[mboutet]

From my understanding of the codebase, I believe that these problems could be alleviated using synchronisation primitives offered by gevent (e.g. Synchronized queues, Locking primitives). For instance, some events such as start, stop, spawn events could be queued up and locking primitives could be used to ensure that only one of those mutually exclusive events (start and stop are two exclusive events IMO) is happening at a given time. Right now, these greenlets all run concurrently without real knowledge of each others resulting in flaky situations like the one in this issue.

[github-actions[bot]]

This issue was closed because it has been stalled for 10 days with no activity. This does not necessarily mean that the issue is bad, but it most likely means that nobody is willing to take the time to fix it. If you have found Locust useful, then consider contributing a fix yourself!