Hi all, a quick one about running prefect agents on windows, with enterprise cloud server. I am facing an issue where after long-running tasks complete the agents just stop. The last flows exit cleanly, and the next message I get is just

Agent stopped!

- is there any reason why agents would just be stopping? I thought they were meant to run indefinitely. I do not know where to access any logging to see why the agent was stopped.

I ran into https://stackoverflow.com/a/30517482 for general python scripts seeming to have frozen until interaction and changing the setting stopped that, but then you cant copy paste from the terminal... whomever thought that was a good idea....

this is really helpful thank you! The plot has thickened a bit since - something is going on which is quite worrying in a flow that uses

from multiprocessing import Manager, Process

in short, I believe this might be interfering with prefect. When I have deployed 2 agents on the same box in different queues, I have seen logs from the flow which uses multiprocessing present in an agent that is registered on a completely different work queue! This can't be expected behaviour! As a further note to my original question I believe running

manager.shutdown()

within my flow might be having some undesirable effect on the prefect agents, but I have not yet tested this. In the meantime I have disabled the flow using

multiprocessing

I get

Agent stopped

as well with the following error:

Failed the last 3 attempts.  Please check your environment and configuration.
Examples of recent errors:

Traceback (most recent call last):
  File "miniconda3/envs/prefect2-test/lib/python3.9/site-packages/h2/connection.py", line 224, in process_input
    func, target_state = self._transitions[(self.state, input_)]
KeyError: (<ConnectionState.CLOSED: 3>, <ConnectionInputs.SEND_HEADERS: 0>)

This error shows up after the agent was running for more than 24h. Would you say this is similar to your issue or completely different?

yes, that's exactly the error

My agent was dead again this morning. This is getting a bit annoying.

same here

Now it just died again after < 3h on a linux machine.

thanks for submitting this - surprised more people aren't encountering this