HPC jobs completing successfully; error logs not empty

[jhlegarreta]

Summary

I am using QSIPrep to pre-process some BIDS datasets using an HPC. I am seeing that the scheduler is reporting the tasks as being completed successfully and the output files being written successfully, but the err files are showing some error stack.

Additional details

• QSIPrep version: 1.0.0rc1
• Docker version: N/A
• Singularity version: 3.7.0

What were you trying to do?

Run qsiprep to preprocess anat+dwi(+fielmap) data.

What did you expect to happen?

I would expect the err files to be empty.

What actually happened?

The err file written by the SLURM job contains some error stack. The full contents of the file are attached.

qsiprep_err.txt

Reproducing the bug

The sbatch relevant snippet:

#SBATCH --output=output_%A_%a.out
#SBATCH --error=error_%A_%a.err

The singularity command snippet:

cmd="singularity run --cleanenv \
  --bind ${in_bids_dirname} \
  --bind ${out_dirname} \
  --bind ${fs_license_fname}:${mapped_fs_license_fname} \
  --bind ${work_dirname} \
  ${qsiprep_singularity_fname} \
  ${in_bids_dirname} \
  ${out_dirname} \
  participant \
  --participant-label ${participant} \
  --output-resolution ${output_resolution} \
  --fs-license-file ${mapped_fs_license_fname} \
  -w ${work_dirname} \
  --skip-bids-validation"

The stack trace does not seem to contain critical errors, but would be grateful if you could confirm this.

Even if they are not critical, I believe they should be fixed, though.

Thanks for developing and maintaining this tool.