preprocessing Bus error

[lshoheisel]

Hi folks,
I’m trying to use rabies to preprocess some data from the StandardRat dataset. The pipeline works if I run it on up to 8 scans, but if I try to run it on 9 or more, I get a Bus error:

subprocess.CalledProcessError: Command 'antsApplyTransforms -i /preprocess_outputs/main_wf/bold_main_wf/bold_commonspace_trans_wf/_scan_info_subject_id302009.sessionNone.runNone_split_name_sub-302009_task-rest_bold/bold_transform/bold_vol884.nii.gz -t /preprocess_outputs/main_wf/commonspace_reg_wf/atlas_reg/template_sharpen_shapeupdate_output_1Warp.nii.gz -t /preprocess_outputs/main_wf/commonspace_reg_wf/atlas_reg/template_sharpen_shapeupdate_output_0GenericAffine.mat -t /preprocess_outputs/main_wf/commonspace_reg_wf/generate_template/nlin/1/transforms/sub-302009_task-rest_bold_RAS_inho_cor_1Warp.nii.gz -t /preprocess_outputs/main_wf/commonspace_reg_wf/generate_template/nlin/1/transforms/sub-302009_task-rest_bold_RAS_inho_cor_0GenericAffine.mat -t motcorr_vol884.mat -n BSpline[5] -r /preprocess_outputs/main_wf/bold_main_wf/bold_commonspace_trans_wf/_scan_info_subject_id302009.sessionNone.runNone_split_name_sub-302009_task-rest_bold/bold_transform/resampled.nii.gz -o /preprocess_outputs/main_wf/bold_main_wf/bold_commonspace_trans_wf/_scan_info_subject_id302009.sessionNone.runNone_split_name_sub-302009_task-rest_bold/bold_transform/deformed_volume884.nii.gz' died with <Signals.SIGBUS: 7>.

(Full error file: err.txt)

Rabies call:
srun --nodelist=$1 docker run --rm -v $PWD/atlas/:/atlas -v $PWD/StandardRat:/StandardRat:ro -v $PWD/preprocess_outputs:/preprocess_outputs/ rabies -p MultiProc preprocess /StandardRat/ /preprocess_outputs/ --anat_template /atlas/SIGMA_InVivo_Brain_Template.nii --brain_mask /atlas/SIGMA_InVivo_Brain_Mask.nii --WM_mask /atlas/SIGMA_InVivo_WM_mask.nii --CSF_mask /atlas/SIGMA_InVivo_CSF_mask.nii --vascular_mask /atlas/SIGMA_InVivo_CSF_mask.nii --labels /atlas/SIGMA_InVivo_Anatomical_Brain_Atlas.nii --TR 1.0 --commonspace_resampling 0.3x0.3x0.3 --bold_only

log file: rabies_preprocess.log

I’m running rabies in docker on a CentOS server using slurm. I managed to run 20 scans from a different dataset without any issues, but I’m struggling to make this work for this dataset – can you tell me what the issue might be?

Best,
Linnea

[gdevenyi]

Sigbus comes from the supervising operating system, this is not a rabies error, are you on some kind of HPC? You may be violating a resource limit.

[lshoheisel]

I'm working on our lab server - I've tried increasing the amount of memory available to the job, but this didn't help.