distributed: Event loop was unresponsive in Worker

[marvin-robot]

Archived from the Prefect Public Slack Community

tsang.yong: hey, I’m getting these in the logs. is this something I in my workflow code or dask config or prefect? dask-worker[65048]: distributed.core - INFO - Event loop was unresponsive in Worker for 3.21s. This is often caused by long-running GIL-holding functions or moving large chunks of data. This can cause timeouts and instability.

jeremiah: Hi <@UPV0FJA3D> - that’s a Dask log which you’ll usually see if your task has seized the CPU entirely (so no background processes can run) — are you doing something particularly intensive?

jeremiah: (Another way you can see this is if you’re running a dask worker / scheduler on something like a laptop and you sleep the computer; when you turn it back on it’ll show this message)

tsang.yong: yes it’s a CPU intensive task

jeremiah: Ok, that could be normal — just Dask warning you that it wasn’t able to run any background task while your task was running.

josh: <@ULVA73B9P> archive “distributed: Event loop was unresponsive in Worker”