Re: 'ellipsoid' crash -- November 06, 2008 - 10:08

That really is a strange problem.  I wouldn't run relax as root
though.  We've used safe routines, security wise, in relax but running
as root is probably not a good idea.  This is likely to be an issue
with the computer set up and the non-local mount points.

Regards,

Edward



On Thu, Nov 6, 2008 at 2:56 AM, Tyler Reddy <TREDDY@xxxxxx> wrote:
> Yeah it's a permission problem. The same thing happens with XPLOR-NIH 
> writing
> files to the mount directory. The command:
>
> sudo relax full_analaysis.py
>
> does the trick. I'm pretty sure I could also chmod a+w the parent directory 
> or
> just execute the script as root.
>
> Tyler
>
> Quoting Edward d'Auvergne <edward@xxxxxxxxxxxxx>:
>
> > Hmmm, that's quite strange.  Does the 'oblate/init' directory exist
> > while the results file is missing?
> >
> > Regards,
> >
> > Edward
> >
> >
> > On Wed, Nov 5, 2008 at 9:42 PM, Tyler Reddy <TREDDY@xxxxxx> wrote:
> > > I just got the same problem with 'oblate', and that folder was removed.
> > > I wonder
> > > if there's a problem with doing this via the NFS mount. Maybe a 
> > > permissions
> > > thing with directory creation from the python script run on the CPU
> > > of another
> > > machine writing results to the mount.
> > >
> > > Tyler
> > >
> > >
> > >
> > > Quoting Edward d'Auvergne <edward@xxxxxxxxxxxxx>:
> > >
> > > > Tyler,
> > > >
> > > > This is an easy one.  Just delete the ellipsoid directory!
> > > >
> > > > The problem is that the full_analysis.py scipt is not very robust.
> > > > It's checking for the 'ellipsoid/init' directory to determine which
> > > > round of optimisation it is up to.  Maybe it would be an idea to
> > > > modify the script to be more robust to handle the directory existing
> > > > yet the final results file being missing for any round.  I would guess
> > > > that relax was killed during its execution, hence the directory exists
> > > > yet the results are missing.  I can't do this at the moment but if you
> > > > would like play with some python coding to implement it yourself, the
> > > > changes could be incorporated back into relax for the benefit of all.
> > > > This may not be simple though, so just deleting the offending
> > > > directory will do the trick.
> > > >
> > > > Regards,
> > > >
> > > > Edward
> > > >
> > > >
> > > > On Wed, Nov 5, 2008 at 6:59 PM, Tyler Reddy <TREDDY@xxxxxx> wrote:
> > > > > Hi,
> > > > >
> > > > > Using the full analysis script, optimization of the 'ellipsoid'
> > > > > model crashes
> > > > > out early with the following output:
> > > > >
> > > > > relax> results.write(file='results', dir='ellipsoid/init/', force=True,
> > > > > compress_type=1)
> > > > > Opening the file 'ellipsoid/init//results.bz2' for writing.
> > > > > RelaxError: Cannot open the file 'ellipsoid/init//results.bz2'.  No
> > > > > such file or
> > > > > directory.
> > > > >
> > > > > I'm not sure why it would try to open this directory though. Isn't
> > > > > it odd that
> > > > > there are 2 forward slashes?
> > > > >
> > > > > Tyler
> > > > >
> > > > >
> > > > > _______________________________________________
> > > > > relax (http://nmr-relax.com)
> > > > >
> > > > > This is the relax-users mailing list
> > > > > relax-users@xxxxxxx
> > > > >
> > > > > To unsubscribe from this list, get a password
> > > > > reminder, or change your subscription options,
> > > > > visit the list information page at
> > > > > https://mail.gna.org/listinfo/relax-users
> > > >
> > > > _______________________________________________
> > > > relax (http://nmr-relax.com)
> > > >
> > > > This is the relax-users mailing list
> > > > relax-users@xxxxxxx
> > > >
> > > > To unsubscribe from this list, get a password
> > > > reminder, or change your subscription options,
> > > > visit the list information page at
> > > > https://mail.gna.org/listinfo/relax-users
> > >
> > >
> > >
> > >
> > > _______________________________________________
> > > relax (http://nmr-relax.com)
> > >
> > > This is the relax-users mailing list
> > > relax-users@xxxxxxx
> > >
> > > To unsubscribe from this list, get a password
> > > reminder, or change your subscription options,
> > > visit the list information page at
> > > https://mail.gna.org/listinfo/relax-users
>
>
>
>
> _______________________________________________
> relax (http://nmr-relax.com)
>
> This is the relax-users mailing list
> relax-users@xxxxxxx
>
> To unsubscribe from this list, get a password
> reminder, or change your subscription options,
> visit the list information page at
> https://mail.gna.org/listinfo/relax-users