[aspect-devel] Writing output in parallel
Thomas Geenen
geenen at gmail.com
Tue Feb 28 09:08:17 PST 2012
yep that looks like it
but rather system ("mv tmp_filename filename &");
beeing able to set the /scratch dir from the input(prm) file would be nice
as well.
i have an implementation lying around somewhere. i implemented it and
tested it on our local cluster
an added benefit of doing it like this is that you do not have to wait till
the end of the run to inspect your results
for curious people like me thats a huge advantage.....
cheers
Thomas
On Tue, Feb 28, 2012 at 6:02 PM, Wolfgang Bangerth
<bangerth at math.tamu.edu>wrote:
>
> > something like:
> > 1. ln -s /a/local/tmp/filesystem output
> > 2. run normally: mpirun ...
> > 3. mv output/* /some/shared/filesystem/
> >
> > Thomas was thinking of doing step 3 concurrently with the code running
> > (for example spawn a shell script after every visualization).
>
> Thinking about this, isn't step 3 *after* the job is done creating the
> very same bottleneck we wanted to avoid? I.e., isn't doing it in the
> background the only real option?
>
> Thomas, at least for this second scheme, what would you think if I
> modified the background_writer() function we currently have to the
> following:
> ..........................
> void Visualization<dim>::background_writer (const std::string *filename,
> const std::string *file_contents)
> {
> char *tmp_filename = tempnam(...) // get name of a tmp file
>
> ...write data into tmp_filename...
>
> system ("mv tmp_filename filename");
> }
> ..........................
> Is this what you had in mind?
>
> Best
> W.
>
> --
> ------------------------------------------------------------------------
> Wolfgang Bangerth email: bangerth at math.tamu.edu
> www: http://www.math.tamu.edu/~bangerth/
>
> _______________________________________________
> Aspect-devel mailing list
> Aspect-devel at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://geodynamics.org/pipermail/aspect-devel/attachments/20120228/9c9f4fbf/attachment.htm
More information about the Aspect-devel
mailing list