Hi,
Do you "qsub bedpostx"? You do not need that and I am not sure how queueing dependencies will be handled. Bedpostx calls internally fsl_sub (a wrapper of qsub) and should be called directly on the cluster.
From your logs, it seems bedpostx in your case runs serially on one core. How long does it take to run? If in parallel it should not take more than 2-3 hours using typical diffusion data (2x2x2 mm3, 60 directions).
Cheers
Stam
On 5 Aug 2013, at 17:26, Rodrigo Perea wrote:
> Hello,
> So I run my bedpost analyses in a cluster however when I look at the log files from the cluster some bedpostx processing finish like this:
>
> tail QSUBS/tee_11883_122_01_BEDPOST.qsub
> 59 slices processed
> 60 slices processed
> 61 slices processed
> 62 slices processed
> 64 slices processed
> 65 slices processed
> 67 slices processed
> Queuing post processing stage
> 75 slices processed
> All slices processed
>
>
>
>
> But others look like this:
>
> [rperea@hpc BED_XXvsXX]$ tail QSUBS/tee_11969_053_01_BEDPOST.qsub
> 55 slices processed
> 56 slices processed
> 57 slices processed
> 58 slices processed
> 59 slices processed
> 61 slices processed
> 62 slices processed
> 64 slices processed
> 67 slices processed
> Queuing post processing stage
>
>
>
> So, does this mean that the processing has not finished? Could that be RAM/cpu errors? I know if I re-run it eventually it will finish but I dont know why this happens :/
>
|