If you are using minimal toolchains you should be able to just set 
`--use-existing-modules`

On Wed, 5 Dec 2018 at 16:52, Pinches Simon 
<simon.pinc...@iter.org<mailto:simon.pinc...@iter.org>> wrote:
Dear EB experts,

I have an extensive set of modules in wide use and built using minimal 
toolchains.  Since upgrading to EasyBuild/3.7.1 I find that when I try to build 
a new foss/2018a module that depends upon Python/3.6.4, it wants to install 
Python/3.6.4-golf-2018a rather than using the existing Python/3.6.4-foss-2018a. 
 This then leads to conflicts between foss/2018a modules which already load 
Python/3.6.4-foss-2018a as a dependency.  I guess that the golf/2018a version 
of Python has recently appeared as a new more minimal version but what is the 
recommended way to proceed?  Can I persuade EB to use my existing Python 
installation, or do I need to re-install everything with a Python dependency to 
use the golf variant instead?  (Sounds fraught with danger).  Alternatively, 
should I revert and stick with EasyBuild/3.6.2 which seems to not have this 
problem?

Thanks,

Simon



--
Dr. Alan O'Cais
E-CAM Software Manager
Juelich Supercomputing Centre
Forschungszentrum Juelich GmbH
52425 Juelich, Germany

Phone: +49 2461 61 5213
Fax: +49 2461 61 6656
E-mail: a.oc...@fz-juelich.de<mailto:a.oc...@fz-juelich.de>
WWW:    http://www.fz-juelich.de/ias/jsc/EN


------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------
Forschungszentrum Juelich GmbH
52425 Juelich
Sitz der Gesellschaft: Juelich
Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher
Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender),
Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
Prof. Dr. Sebastian M. Schmidt
------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------

Reply via email to