Cannot find lh.white.H about openneuro HOT 12 CLOSED

@chrisfilo I think this is fixed by the disk changes (maybe the move to Overlay2). Please reopen if you're still running into it.

from openneuro.

I recently stumbled upon this again: https://openneuro.org/datasets/ds000110/versions/00001?app=FreeSurfer&version=26&job=79e6d717-ab91-4590-bb17-63f5d574d445

from openneuro.

Does this occur randomly? If you retry the job it sometimes succeeds?

Since the file is missing in the results and symlinks are dereferenced when it's copied out, I think it's actually missing when it reaches this part of the app.

from openneuro.

from openneuro.

I'm seeing more of such and other similar error in this job https://openneuro.org/datasets/ds001043/versions/00001?app=FreeSurfer&version=27&job=5a2aec58fce2fd0001047fc3

from openneuro.

I'm getting similar looking error when running freesurfer as part of FMRIPREP:

Loading label /output/data/freesurfer/fsaverage5/label/rh.cortex.label
Standard error:
No such file or directory
mri_vol2surf: could not open label file /output/data/freesurfer/fsaverage5/label/rh.cortex.label
Invalid argument
mri_vol2surf: could not load label file /output/data/freesurfer/fsaverage5/label/rh.cortex.label
Invalid argument
Return code: 255

https://openneuro.org/datasets/ds001076/versions/00001?app=FMRIPREP&version=45&job=7090f9ff-30e9-41ed-968d-e540e5cd03f2 (Subject control07)

from openneuro.

I rerun the previously failing job at https://openneuro.org/datasets/ds001043/versions/00001?app=FreeSurfer&version=27&job=5a2aec58fce2fd0001047fc3 and it hit the same snag.

This might be a good datasets to test potential solutions.

from openneuro.

I tried with 30Gb instances and this is still failing (on ds001043): https://openneuro.org/datasets/ds001043/versions/00001?app=FreeSurfer&version=28&job=5a5f7a7324c4760001b50c89

from openneuro.

I'm begging to think that this might have something to do with the fact that this error happens on datasets with more than one session. Investigating.

from openneuro.

It fails even if only one session is selected: https://openneuro.org/datasets/ds001043/versions/00001?app=FreeSurfer&version=30&job=5a600929bcef5e0001c394a1

from openneuro.

Noticed that production is still using an out of date host container and it looks like this is because the dockerhub build did not complete. This is fixed now.

I'm running some more tests to see if I can get this to fail outside of ECS as that will make it easier to pause a container and look at what might have failed.

from openneuro.

I'm experiencing this error when running freesurfer on ds000201: https://openneuro.org/datasets/ds000201/versions/00004

Are there any workarounds?

from openneuro.