[GRASS-user] Optimized compiling of GRASS 7 for hydrological analysis on CentOS

GRASS GRASS user
1

Hi,

At the moment a colleague and me are trying to compile GRASS 7 on a CentOS server / cluster for watershed modeling (including raster-vector-transformation).

We are having different compilation problems which likely are caused by the setup of the server (we will try to figure out).

Now, we would like to minimize the dependencies and remove all flags/options which are not necessary for watershed analysis, while we like to use all flags/options which may speed-up the hydrological analysis (because resources on the server / cluster are shared and processing time is limited for each user).

In this context we would like to ask for suggestions on what flags and options to use/not to use?
We do not need a gui at all (as we won`t be able to access it on the server anyway), and gui-modules are among those making trouble, can we turn off compilation of gui modules?

Any hints regarding compiler flags are welcome too...

Thanks for helping,
Stefan

2

Blumentrath, Stefan wrote:

We do not need a gui at all (as we won`t be able to access it on the
server anyway), and gui-modules are among those making trouble, can
we turn off compilation of gui modules?

There shouldn't be any need to. If a module fails to compile, it won't
affect the compilation of other modules (even without make's -k
switch). In the worst case, you just some noise in the build output.

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