Async Timer

This timer can be started in one scenario, on one worker, and be stopped in another scenario on another worker.

It is possible to stop a timer only based on tid and version, but in that case that combination can only be used for one timer. This is useful if a timers is started in many different scenarios, but are stopped in one scenario which does not have any other information than the id (which it might have received from a queue or topic), and hence the timer name is not know at the time it is being stopped.

A name must always be provided when starting the timer.

name, tid and version supports Templating.

Step implementations

• step_task_async_timer_start

• step_task_async_timer_stop

Statistics

When running distributed, the timer statistics piggyback with normal locust statistics which is sent from workers (default every 3 seconds) via the event locust.event.report_to_master, and is then handled on master with the locust.event.worker_report event. If running local, it will be started and stopped in "real-time".

When the stop-part of this task is executed, a "request" with method DOC and name of the timer will be added to the request statistics. If anything goes wrong when executing the task, the error will also visible in the locust failure summary.

Any timers that has not been stopped when the load test is finished, will be listed in the behave failure summary, and the test will be marked as failed.