WorkerLost: Task failure due to loss of worker Failure during join data creation using HTE Parsl config with SQ00014613.sqlite

[d33bs]

This issue occurred while trying to confirm the fix for #38 . After initial data extraction and concatenation (per compartment) CytoTable attempts to join the data in chunks and then concatenates the chunks. Using a Google Colab environment I noticed that I was unable to complete the join chunk creation step when using the HTE (HighThroughputExecutor) from Parsl (a default for CytoTable currently). The error is WorkerLost: Task failure due to loss of worker # with little other details on why.

Using a ThreadPoolExecutor for the parsl config results in a successful run within the same environment.

It may be that memory resources are inadequate within Google Colab environments (~12GB) to complete this work with a larger dataset.

Possibly related within the runinfo/*.err file. The parsl.log file shows the "WorkerLost" error but no other details.

/usr/lib/python3.10/multiprocessing/resource_tracker.py:224: UserWarning: resource_tracker: There appear to be 7 leaked semaphore objects to clean up at shutdown

Example Google Colab notebook and backup Gist.