Is 68 your last slice?
If so, the Unexpected EOF is an old buglet that is fixed in the new
FSL, but had no effect on the output data. All should be well.
T
On 30 Aug 2007, at 15:38, Sheyna Gifford wrote:
> Fsl list-
>
> We're trying to run bedpost as a parallel process on a cluster with
> 16 dual
> processor servers. The process runs happily, creating the
> appropriate files
> and directories, until it gets to slice 68, whereupon one of two
> things occurs:
>
> 1) the process persists in trying to process slice 68 and will
> continue to
> do so until killed
> OR
> 2) “Unexpected end of file” for bsp/cluster/local/bedpost
>
> There are also sometimes associated memmory errors, but these are not
> consistant:
> ----
> Cannot allocate memory
>
> \x00\x00...
> ------
>
> Has anyone else been the victim of these slice 68 errors running
> bedpost in
> parallel?
>
> Thank you very much,
>
> shey
|