distributed.worker - ERROR - Worker stream died during communication

[NiklasPhabian]

On an irregular basis, dasks gives me something like this:

distributed.worker - ERROR - Worker stream died during communication: tcp://127.0.0.1:42391
Traceback (most recent call last):
File "/home/griessbaum/.virtualenvs/staremaster_py/lib/python3.8/site-packages/distributed/comm/tcp.py", line 186, in read
n_frames = await stream.read_bytes(8)
tornado.iostream.StreamClosedError: Stream is closed

oddly enough, it seems like the sidecar is created just fine regardless

[NiklasPhabian]

I mainly see this happening when I reduce the chunksize in conversions.py and/or when I set workers to my max number of cores on my machine