srun_fastsurfer.sh exits early

[ClePol]

Description

srun_fastsurfer.sh does not work as before in the latest stable release. An unknown errors causes the script to exit slinently. Removing "set -e" (exit on error) seems to restore the previous functionality.

Steps to Reproduce

./srun_fastsurfer.sh \
  --singularity_image /fastsurfer_stable.sif \
  --sd /output_dir \
  --data / \
  --subject_list /missing_fastsurfer_cases.csv \
  --subject_list_delim ',' \
  --fs_license /software/fs732/.license \
  --num_cases_per_task 1 \
  --work $HPCWORK/fastsurfer-output/debug_srun \
  --skip_cleanup \
  --debug

Expected Behavior

slurm_jobs are started as before

Log Files /Screenshots

Last executed commands

+ i=1
+ IFS=' 
'
+ [[ -n , 8bca23c8-57ac-4be6-9b06-4ff7f45ef6a5, dca3bb34-7300-411c-a1b7-ffc9dedf90af, b5f1fb67-5c38-4f1d-9080-d4a3b0af7866, a7cccd61-5b58-4342-b89c-bd210f3986b4, a67bdf44-01d2-4bbe-8414-34a99f3d9f66, 227aef5b-542b-49b0-aa9c-58ae050afd56, 5812c757-19ba-4750-9145-23dd8991ab78 ]]
++ [[ -n '' ]]
+ condition=

I can provide the full outputs and commands on request

Environment

Singularity image built from stable (0553f0a, v2.4.2)

Additional Context

I agree that setting the -e flag is a good option, we should look into why the script fails and maybe into producing a stacktrace, so next time it's easy to see where the error occured. Maybe in the future we can change this into a python script?

[dkuegler]

Am I assuming correctly, that you are using a v2.4.2 image, but the dev srun_fastsurfer.sh?
The relevant set -e (line 198 of srun_fastsurfer I assume) has been in srun since 2023.

In reality, the output implies, that for the listed cases/participants see 3 of the quote, there are no files stored, which may have different reasons. I think I would need to investigate this.

[dkuegler]

Could you check, if some of your images are symlinks, i.e. if the files you are passing to be processed are symlinks via the file or even some folder being a symlink?

[dkuegler]

Issue solved in #741