Hey team - I am running prefect jobs within a docker container, and I still cannot seem the get the logging to the API to turn off when I am just running locally. Is there a better way to do this? Nothing seems to prevent it from showing up in the cloud and triggering our alarms for failed runs?

if __name__ == '__main__':
    import os
    os.environ["PREFECT_LOGGING_ORION_ENABLED"] = "False"
    # with temporary_settings(set_defaults={PREFECT_LOGGING_TO_API_ENABLED : False}):
    with use_profile('prod'):
        run_my_flow(x)

Hey Mitch - to clarify, are you trying to only disable logging or are you trying to avoid sending all run state updates to the Cloud API?

trying to disable sending run states to the API just for local testing. I guess it would be fine if it logged to API, but when I am testing new applications, we get a bunch of notifications due to the runs failing. In prefect one all of the runs locally didn't log into the cloud API. We have an Enterprise subscription for prefect

great, that makes sense - the most robust way to do this is through profiles. For example I have a profile that configures my API URL to be a locally running instance of

prefect server start

(so it's not talking to Cloud at all), so I'll do:

prefect profile use local
## python run-my-flow.py or whatever

if you are heavily relying on cloud features within the flow itself, this is where a separate workspace comes in handy (i.e. a "Sandbox" workspace where anything goes), and then

prefect profile use sandbox

to switch APIs

Yeah that makes sense - we rely on workspace specific blocks such as secrets in many of the runs however. Hmm

gotcha, yea that makes sense; we probably could expose a flag for disabling all create/update activity and instead make it a "read-only" run that never gets an ID. I'm not immediately sure if there are edge cases or "gotchas" in that approach, but I'd be game to look into it.