Spurious CancelledError on LocalCluster.close()

The following repro produces a spurious log error from tornado/asyncio on closing the LocalCluster:

[00:36:18.413 ERROR  ] Exception in Future <Future cancelled> after timeout
Traceback (most recent call last):
  File "../lib/python3.5/site-packages/tornado/gen.py", line 970, in error_callback
    future.result()
  File "../lib/python3.5/asyncio/futures.py", line 286, in result
    raise CancelledError
concurrent.futures._base.CancelledError

Here is the repro:

from joblib import parallel_backend

import distributed.joblib
from distributed import Client, LocalCluster


def run_example():
    lc = LocalCluster(n_workers=1)
    client = Client(lc)
    with parallel_backend('dask.distributed', scheduler_host=client.scheduler.address):
        pass
    client.close()
    lc.close()


if __name__ == "__main__":
    run_example()

Couldn’t figure out how to silence or what exactly is wrong here. Appreciate any guidance.

Distributed versions is 1.23.1 on python 3.5.6 and joblib 0.12.3

dask                      0.19.1                   py35_0  
dask-core                 0.19.1                   py35_0  
tornado                   5.1              py35h14c3975_0 

[ericness]

I was also getting the same error message. I couldn’t figure out how to stop the error message so I manually closed the client before exiting and caught the exception.

from concurrent.futures._base import CancelledError

...

try:
    client.close()
except CancelledError:
    print('Dask distributed processing did not shut down cleanly.')

[somewacko]

Yes, installing @danpf’s patch on my client machine seems to have fixed the issue.