Hi Olena,
The problem you are seeing might be related to the use of the virtual
machine (VM) - these can sometimes be much slower than the hardware
they are running on (VirtualBox comes to mind here). For slow
model-free calculations, I would highly recommend you try running on a
cluster or multi-core system using OpenMPI via Gary Thompson's
multi-processor framework in relax
(http://www.nmr-relax.com/manual/multi_processor_framework.html). For
this you can set everything up in the GUI, save the relax state, and
then copy and modify the
sample_scripts/model_free/dauvergne_protocol.py script to load that
state, deleting all of the other user functions. Otherwise you could
run relax directly on MS Windows.
From the current information, it's a bit difficult to work out if this
is a bug or just a very complex optimisation problem coupled with a
slow VM. What do you see in the relax controller window? Is anything
changing? What is the current data pipe and global model? And what
were the last messages? It might be useful to copy and paste those.
Are the messages changing? Could you also run 'relax -i' and copy and
paste that output, as that might help me work out if it is a GUI
freezing bug.
Cheers,
Edward
On 22 July 2014 14:13, Olena Dobrovolska <olena.dobrovolska@xxxxxxxx> wrote:
Hello,
I am a new user of Relax. I started the calculation of Relax on June 12 on
the virtual machine Ubuntu installed on Windows. The Relax control window
does not report any Errors or Warnings, and the last output
folder 'Prolate' with the 6th round is dated by July 3. It has been more
than a month of the calculation already, while in the manual it says no
more than two weeks. I typed 'top i' in the Terminal, and the program is
still running. Can anyone help? I do not see it producing more output since
July 3.
Would appreciate any advice.
Olena
_______________________________________________
relax (http://www.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
Re: Too long computational time