After importing a subject’s data (T1 single time point)

I run a script with these commands:

recon-all –s <subject> -autorecon1

recon-all –s <subject> -autorecon2

recon-all –s <subject> -autorecon3

I did this twice, once on each of two linux machines.

I am using the same exact input intensity T1 data and importing the data into separate subject names (e1859_workingBak  and e1858_nonworking) so there is no collision.

Why am I getting many (really 1000’s ) of  v1_1.dat files in the $SUBJECTS_DIR/<subject>/mri   of one of the machines, even though the input data is the same?

They are named vNNN_N.dat where N is a digit . Contents of one of these files is below.

The contents of the subject folders are represented below using the du –k . command. Now in this nonworking case the job terminated early (run time limit on the job which I can fix ) ; however I am curious: are all of these *.dat files  (40GB worth!) written out all the time and then deleted at the end of the recon tasks?

This happens on one red-hat linux machine (cluster ) but not on another for the same subject data

Is this normal behavior? Are the dat files written out normally and then deleted later in the processing (since I don’t see them in the run that completed).

Contents of a subject after a successful run to completion:

[e1858_workingBak]$ du -k .

1176    ./scripts

16      ./mri/transforms/bak

108200  ./mri/transforms

14088   ./mri/orig

176068  ./mri

149940  ./surf

16      ./tmp

35228   ./label

1096    ./touch

544     ./stats

16      ./src

16      ./trash

16      ./bem

364132  .

Contents of same MRI input data to a different subject name, whose processing (same as above) terminated prematurely:

[e1858_nonworking]$ du -k .

800     ./scripts

16      ./mri/transforms/bak

108200  ./mri/transforms

14088   ./mri/orig

41027804        ./mri

29252   ./surf

16      ./tmp

16      ./label

360     ./touch

16      ./stats

16      ./src

16      ./trash

16      ./bem

41058328        .

This is when the processing of e1858_nonworking got terminated:

[scripts]$ tail recon-all-status.log

#@# Intensity Normalization2 Thu Jul 25 19:20:22 EDT 2013

#@# Mask BFS Thu Jul 25 19:24:53 EDT 2013

#@# WM Segmentation Thu Jul 25 19:24:56 EDT 2013

#@# Fill Thu Jul 25 19:27:47 EDT 2013

#@# Tessellate lh Thu Jul 25 19:28:47 EDT 2013

#@# Smooth1 lh Thu Jul 25 19:28:56 EDT 2013

#@# Inflation1 lh Thu Jul 25 19:29:01 EDT 2013

#@# QSphere lh Thu Jul 25 19:29:37 EDT 2013

#@# Fix Topology lh Thu Jul 25 19:35:03 EDT 2013

#@# Make White Surf lh Thu Jul 25 19:55:20 EDT 2013

This is the contents of one (v1_1.dat ) of the many *.dat files  in /mri

[mri]$ tail v1_1.dat

12.014513 223.051927 250.208421

12.052258 222.911943 252.203797

12.090005 222.771944 254.199159

12.127751 222.631960 256.194520

12.165498 222.491960 258.189912

12.203243 222.351976 260.185273

12.240990 222.211977 262.180635

12.278736 222.071992 264.175996

12.316483 221.931993 266.171358

12.354229 221.791994 268.166750