[GRASSLIST:5778] Spawning processes

GRASS GRASS user
1

Hi!
I need to process a large data set with s.surf.rst (I
went back to using
5.4 due to Bug#2903). The idea is to spawn several
s.sur.rst processes
on an SMP machine, each with a different region (set
to have some
overlapping). The individual rasters are then patched
together into a
single large raster file.

The idea is to have a script doing this
g.region north=.... south=...
s.surf.rst <params> &
g.region north=... south=...
s.surf.rst <params>&

This of course, assumes that once s.surf.rst is
called, the region
information is no longer used, and can be changed for
the next process
to use a different region.

Can anyone see any showstoppers in this?

Cheers!
José
--
Jose L Gomez-Dans, Research Assistant
Bristol Glaciology Centre, Geographical Sciences/CPOM
University of Bristol, Bristol, UK

___________________________________________________________
ALL-NEW Yahoo! Messenger - all new features - even more fun! http://uk.messenger.yahoo.com

2

Jose Luis Gomez Dans wrote:

I need to process a large data set with s.surf.rst (I
went back to using
5.4 due to Bug#2903). The idea is to spawn several
s.sur.rst processes
on an SMP machine, each with a different region (set
to have some
overlapping). The individual rasters are then patched
together into a
single large raster file.

The idea is to have a script doing this
g.region north=.... south=...
s.surf.rst <params> &
g.region north=... south=...
s.surf.rst <params>&

This of course, assumes that once s.surf.rst is
called, the region
information is no longer used, and can be changed for
the next process
to use a different region.

Can anyone see any showstoppers in this?

At a minimum, you need to introduce a delay between starting
s.surf.rst in the background and changing the region, otherwise the
region might get changed before s.surf.rst actually reads it.

--
Glynn Clements <glynn@gclements.plus.com>