Hi Jon
I haven't seen that before, can you give us some more details?
- How did you run bedpostx? "bedpostx subject"? Which FSL version?
- Did processing finished? Bedpostx output looks OK?
- Which are the huge logfiles? log*, bedpostx.e* or bedpostx.o*? Or some other files?
Thanks
Stam
On 10 May 2013, at 11:16, Jonathan Brooks wrote:
> Dear All,
>
> We recently had a problem with running bedpostX via SGE. The problem was that a number of *HUGE* folders were left over in our subject.bedpostX folders after processing, which caused our file system to fill up.
>
> FYI we were running 16 subjects with default bedpostX options. Before we started the process there was 2TB of free space on the system, and about 8 hours into processing we were down to 0bytes.
>
> The problem was that the logs folder totalled ~268GB per subject, and contained several huge logfiles (10s of GBs each).
>
> The raw data were not huge, 2mm isotropic (96x96), 50 slices, b=0 (x1), b=1000 (64 directions), 2 averages.
>
> Any suggestions on how to trouble shoot this so I can avoid the situation happening again would be much appreciated.
>
> Cheers, Jon
> --
> Jonathan Brooks (PhD)
> CRiCBristol Senior Research Fellow
> University of Bristol, St Michael's Hill, Bristol, UK
> T: +44(0)117-342-1512 W: http://www.cricbristol.ac.uk
>
|