Re-runs of genmod on the same cases when using singularity containers were crashing due to OSError
Turns out this error is very likely to be caused by multiprocessing.Manager(), because its default socket address on Unix machines is not very random. So multiple containers on the same machine might be trying to access the same address resulting in memory conflicts. Changes proposed in this PR should fix that.
Re-runs of genmod on the same cases when using singularity containers were crashing due to OSError
Turns out this error is very likely to be caused by multiprocessing.Manager(), because its default socket address on Unix machines is not very random. So multiple containers on the same machine might be trying to access the same address resulting in memory conflicts. Changes proposed in this PR should fix that.
More info: https://stackoverflow.com/questions/74058016/how-do-you-run-multiple-aws-batch-jobs-with-multiprocessing-mp-manager-witho
Review:
This version is a: