Hi Arna,
I think this option is only supported for PythonPackage. This is where
we use it, for example:
https://github.com/ComputeCanada/easybuild-easyconfigs/blob/computecanada-main/easybuild/easyconfigs/s/SciPy-Stack/SciPy-Stack-2022a-GCCcore-9.3.0.eb
Best,
Maxime Boissonneault
Le 2023
Hi Mikael, Loris,
I concur with your user. Boost.MPI is miles ahead of anything MPI you
can do in C++. It is very handy, but not many software packages are
actually written in C++.
I know of RepastHPC, a package written for agent based modeling, which
uses it (https://repast.github.io/repast
Hi Loris,
"extensions"
would be what you are looking for, if you want to pip install specific
packages.
Cheers,
Maxime
Le 21-01-15 à 04:01, Loris Bennett a écrit :
Hi,
I am trying to write an EC for fastText:
https://github.com/facebookresearch/fastText
Using the MakeCp EB I can bui
That’s amazing! At Compute Canada, we were just looking at how to add *more*
hashes to our installation procedures. This comes right on time!
Thanks!
Maxime
>
> On Apr 1, 2020 at 4:19 AM, mailto:kenneth.ho...@ugent.be)>
> wrote:
>
>
> Dear Eas
I did not read he whole bread, but why is this a limitation ? At Compute
Canada, we actually have always had potentially different versions of
numpy/scipy/etc. For python 2.7 vs 3.x, installed in the same module. We
haven’t used the multi_deps for this yet, however.
n here:
https://docs.computecanada.ca/wiki/Accessing_CVMFS
Thanks!
--
---------
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Président - Comité de coordination du soutien à la recherche de Calcul Québec
Team lead - Research Support National Team, Compute Can
with it.
Cheers,
--
-
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Président - Comité de coordination du soutien à la recherche de Calcul Québec
Team lead - Research Support National Team, Compute Canada
Instructeur Software Carpentry
Ph
,
--
-
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Président - Comité de coordination du soutien à la recherche de Calcul Québec
Team lead - Research Support National Team, Compute Canada
Instructeur Software Carpentry
Ph. D. en physique
Oh, and of course, our scipy-stack module supports all versions of python.
Maxime
>
> On Feb 20, 2019 at 6:46 AM, mailto:kenneth.ho...@ugent.be)>
> wrote:
>
>
>
> Dear EasyBuilders,
>
> The next EasyBuild conf call is planned for Wed Feb 20th 2019 (today!),
Hi Kenneth,
I can’t make it to the meeting. However, reading the issue about python to
gcccore, note that on Compute Canada, we have implemented solution #1, and the
bundle module that we propose is based on the scipy stack (and is called
scipy-stack)
https://www.scipy.org/
this user’s statement ?
--
-
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Président - Comité de coordination du soutien à la recherche de Calcul Québec
Team lead - Research Support National Team, Compute Canada
Instructeur Software
l using these toolchains, or if you have another reason
why we should not deprecate them, please come forward ASAP and let us
know.
regards,
Kenneth
--
-
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Président - Comité de coordinati
Giving the announced death of Google+ [1], I feel it's time to start
exploring alternatives to using Google Hangouts for the EasyBuild
conference calls.
I'm open to suggestions, preferably a system where there's no hard cap
on the number of attendees (or at the very least allows at least a
Hard to quantify with an open source project, but there are 127
different forks of the repository, so "a fair number of organizations"
is probably accurate :P
https://github.com/easybuilders/easybuild-framework/network/members
Maxime Boissonneault
On 2018-09-18 3:31 PM, Dey, Jo
Relative to what ?
Relative paths are typically relative to the current working directory.
We can't assume specific working directories.
Regards,
Maxime
On 2018-08-23 7:53 AM, Espen Tangen wrote:
And to set this in easyblocks or easyconfigs?
It seems that there is support for it in framew
Oldeman | bart.olde...@mcgill.ca
<mailto:bart.olde...@mcgill.ca> | bart.olde...@calculquebec.ca
<mailto:bart.olde...@calculquebec.ca>
Scientific Computing Analyst / Analyste en calcul scientifique
McGill HPC Centre / Centre de Calcul Haute Performance de McGill |
http://www.hpc.mcgill
xample that you could share please?
Furthermore, I understand that the scipy-stack is a software stack
providing python2/3 and a host of modules (numpy, scripy, etc) and
there would be a module for this stack. Given that, how do I select
whether I need to use python 2 or 3 in this setup?
Your clarif
Hi Julian,
I don't know if this can be of any help, but we have this as our
instructions to build FEniCS:
https://docs.computecanada.ca/wiki/FEniCS
Maxime
On 18-01-31 05:35, Julian Bigge wrote:
Hello fellow Easybuilders,
we are currently working on an Easyconfig for the DOLFIN and Fenics pyt
via
git recursive - given a specific hash. What I have seen so far
indicates that there is none, but I want to be sure before I create an
issue on this.
Regards
Et.
--
-
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Prési
On 18-01-04 04:23, Jakob Schiøtz wrote:
Hi,
I made a TensorFlow easyconfig a while ago depending on Python with the foss
toolchain; and including a variant with GPU support (PR 4904). The latter has
not yet been merged, probably because it is annoying to have something that can
only build on
See
https://github.com/ComputeCanada/easybuild-easyconfigs/tree/computecanada-master/easybuild/easyconfigs/t/Torch
And
https://github.com/ComputeCanada/easybuild-easyblocks/blob/computecanada-master/easybuild/easyblocks/t/torch.py
I have not created a pr yet because
Hi all,
I was at SC last week, and many people shared interests in our setup for
python, in particular the fact that we manage to support multiple
versions of Python for the same module. I thought I would share our
solution here.
1) We do not define PYTHONPATH. PYTHONPATH is too intrusive, a
OpenFOAM.com/.org developpers,
please tell them they're nuts to have done things this way...
--
-
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Président - Comité de coordination du soutien à la recherche de Calcul Québec
Team lead - Res
I will be at SC17 and will be happy to talk with the EB people who are there
about the distributed setup we’ve built for Compute Canada, using Nix,
EasyBuild, CVMFS and Lmod. The stack currently supports almost 2000modules
spanning 3 hardware architectures and is distributed to all n
gen Huthmacher
Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender),
Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
Prof. Dr. Sebastian M. Schmidt
----
--
axime
On 17-10-06 10:49, Maxime Boissonneault wrote:
Not claiming which one is best, but Compute Canada uses this custom
module naming scheme :
https://github.com/ComputeCanada/easybuild-framework/blob/computecanada-master/easybuild/tools/module_naming_scheme/SoftCCHierarchicalMNS.py
Not claiming which one is best, but Compute Canada uses this custom
module naming scheme :
https://github.com/ComputeCanada/easybuild-framework/blob/computecanada-master/easybuild/tools/module_naming_scheme/SoftCCHierarchicalMNS.py
It is a hierarchical naming scheme with modules in lower case a
Hi,
In the Compute Canada repository, we have the following toolchains :
iimkl (intel compiler + intel MKL)
iompi (intel compiler + openmpi)
iomkl (iimkl + openmpi)
iomklc (iomkl + cuda)
iimklc (iimkl + cuda)
and the equivalent for GCC compilers. We build most of our recipes with
those (min
ite.
Like I said, not sure it is helpful here though.
Jack Perdue
Lead Systems Administrator
High Performance Research Computing
TAMU Division of Research
j-per...@tamu.eduhttp://hprc.tamu.edu
HPRC Helpdesk: h...@hprc.tamu.edu
On 08/14/2017 09:58 AM, Maxime Boissonneault wrote:
Hi,
virtualenv can nicely inherit from each other, giving users the ability
to install their own upgraded version of some python packages without us
needing to create a new module each time and try to handle Python
dependencies through modules.
--
-
Maxime Boissonneault
should be looking at.
Regards,
*From:*easybuild-requ...@lists.ugent.be
[mailto:easybuild-requ...@lists.ugent.be] *On Behalf Of *Maxime
Boissonneault
*Sent:* Friday, August 4, 2017 7:12 PM
*To:* Easybuild
*Subject:* [EXTERNAL] Re: [easybuild] NWChem, Amber, OpenBabel, NAMD,
GAMESS
Hi Shahzeb
Hi Shahzeb,
Maybe not quite the same toolchains, but have a look at
https://github.com/computecanada/easybuild-easyconfigs
We have a recipe for each of those, though we have not merged them because we
don't use the same kind of config as upstream.
Maxime
/Torch/
It is not much tested yet, and does not use the default EasyBuild
toolchains, so I'm not ready to make a PR for it.
However, if you need it, feel free to use it.
Maxime
On 17-05-18 03:11, Kenneth Hoste wrote:
Hi Maxime,
On 17/05/2017 22:27, Maxime Boissonneault wrote:
Hi,
Has a
Yes, I am talking about the machine learning on GPU package.
Maxime
>
> On mai 18, 2017 at 3:11 AM, mailto:kenneth.ho...@ugent.be)>
> wrote:
>
>
>
> Hi Maxime,
>
> On 17/05/2017 22:27, Maxime Boissonneault wrote:
> > H
Hi,
Has anyone made an EB recipe for Torch ?
--
-
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Président - Comité de coordination du soutien à la recherche de Calcul Québec
Team lead - Research Support National Team, Compute Canada
s one.
Maxime
On 17-04-18 07:35, Kenneth Hoste wrote:
On 12/04/2017 20:40, Maxime Boissonneault wrote:
That's interesting, but I'm thinking a more generic answer
(extra_args flag) might be useful. Surely, NAMD is not the only
software which might require that kind of special flags.
ust need a fix in framework since it's doing things wrong in my opinion.
On 04/12/2017 08:12 PM, Maxime Boissonneault wrote:
Hi,
For the latest version of NAMD, we need to add "--std=c++11 -ip" to the
CXXFLAGS when compiling with Intel compilers.
How would we do that from the Ea
ut it is a hack that relies on having the value of cstd not being
quoted when transformed into CXXFLAGS.
What would be the proper way ?
I would expect having something like :
toolchainopts = {'extra_args':'whatever we need'}
Thoughts ?
--
sitzender), Prof. Dr.-Ing. Harald Bolt,
Prof. Dr. Sebastian M. Schmidt
----
--
-
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Président - Comité de coordination du soutien
arch and Education (ACCRE)
Vanderbilt University - Hill Center 201
(615)-875-9137
www.accre.vanderbilt.edu
On Mar 16 2017, at 9:35 am, Maxime Boissonneault
wrote:
If you want an example for this, you can have a look at our easyconfig
for iccifort :
https://github.com/ComputeCanada/e
pplication Developer
Adjunct Assistant Professor of Chemical and Biomolecular Engineering
Advanced Computing Center for Research and Education (ACCRE)
Vanderbilt University - Hill Center 201
(615)-875-9137
www.accre.vanderbilt.edu
On Mar 16 2017, at 9:35 am, Maxime Boissonneault
wrote:
If you want
Karl Eugen Huthmacher
Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender),
Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
Prof. Dr. Sebastian M. Schmidt
-----
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
--------
---
with it already and absolute best
> performance isn't always more important than getting the compilation done in
> less time.
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> On Thu, Mar 2, 2017 at 6:55 PM Maxime Bois
al and Biomolecular Engineering
Advanced Computing Center for Research and Education (ACCRE)
Vanderbilt University - Hill Center 201
(615)-875-9137
www.accre.vanderbilt.edu
On Mar 2 2017, at 5:30 pm, Maxime Boissonneault
wrote:
Hi,
I've seen a couple emails about CUDA recently, and I was a
ersonally don't really care, I just want the best performance for my
users (which is why we don't go with FOSS in the first place, since MKL
gives better performances than OpenBLAS).
I just thought I'ld raise the question.
--
-----
Maxime Boissonneau
Hi Shahzeb,
I am putting you in contact with my colleague Oliver Stueker. He is
actually working to modify the recipe for GROMACS GPU. Maybe he can help.
Maxime
On 17-03-01 14:38, Siddiqui, Shahzeb wrote:
Hello,
I am curious how I would build GROMACS with GPU support, I only see
GROMACS bu
Thanks Alan,
This explains that. We'll look into using the gcc/system toolchain.
Maxime
On 17-03-01 09:59, Alan O'Cais wrote:
Afraid so, the toolchainopts do nothing for the dummy toolchain since
EasyBuild doesn't "know" anything about the options dummy an accept.
If you want to wrap the syst
PIC).
I discovered that only when adding --enable-shared to the configopts
were the flags actually added (I suspect by the configure, rather than
by EasyBuild).
Is this expected ?
--
-----
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Hi,
We are not using version suffixes. We are using an alternative name for
the module (i.e. hdf5-serial/1.8.18, not hdf5/1.8.18-serial)
Maxime
On 17-02-28 15:21, Åke Sandgren wrote:
('hdf5', '1.8.5', '-serial')
depending on actual version
On 02/28/2017
Hi,
I installed HDF5 with the name 'hdf5-serial' (modaltsoftname)
I can't figure out how to get EasyBuild to find the module (i.e. to have
this as a dependency for another module). Whether I specify "hdf5" or
"hdf5-serial", it does not find it. Any tip ?
Maxime
the install directory you could use
"postinstallcmds" in your easyconfig to remove the build directory.
Thanks,
Cormac.
On Thu, Feb 23, 2017 at 2:34 PM, Maxime Boissonneault
<mailto:maxime.boissonnea...@calculquebec.ca>> wrote:
Ok, I tried it this aft
Ok, I tried it this afternoon.
Indeed, buildininstalldir = True does keep the downloaded packages
there... but it also keeps the build directory in there, which clutters
the installation directory for no good reason.
Maxime
On 17-02-23 12:06, Maxime Boissonneault wrote:
I did not know
,
Cormac.
On Wed, Feb 22, 2017 at 6:52 PM, Maxime Boissonneault
<mailto:maxime.boissonnea...@calculquebec.ca>> wrote:
Ok, indeed, after digging in the code, I see that the install_step
does delete the install directory before installation. This breaks
an installation of PETSc wh
can follow.
Maxime
On 17-02-22 18:04, Maxime Boissonneault wrote:
Some more info I discovered.
The files that PETSc downloads are actually correctly installed in the
final installation directory during the configure phase. However,
somewhere between the configure and the install phase, something (
EasyBuild deletes
the content of the installation directory between the configure and the
install phase ?
Maxime
On 17-02-22 14:13, Maxime Boissonneault wrote:
On 17-02-22 14:08, Kenneth Hoste wrote:
Hi Maxime,
On 22/02/2017 17:22, Maxime Boissonneault wrote:
Actually, looking through the log
On 17-02-22 14:08, Kenneth Hoste wrote:
Hi Maxime,
On 22/02/2017 17:22, Maxime Boissonneault wrote:
Actually, looking through the log file, I see PETSc which says to run
make install, but easybuild runs "make" without the install option:
3693 make[2]: Leaving directory
'/d
d " make
PETSC_DIR=/dev/shm/mboisson/PETSc/3.7.5/iompi-2016.4/petsc-3.7.5/ "
exited with exitcode 0 and output:
== 2017-02-22 15:08:34,357 run.py:443 INFO cmd "make test" exited with
exitcode 0 and output:
== 2017-02-22 15:08:34,762 run.py:443 INFO cmd " make install &quo
ke install" of PETSc will copy the external packages as well as PETSc
itself, but in this case, it is not for some reason.
Maxime
On 17-02-22 11:12, Alan O'Cais wrote:
Hi Maxime,
You are probably missing:
sourceinstall = True
in your easyconfig.
Alan
On 22 February 2017
On 17-02-22 01:53, Åke Sandgren wrote:
On 02/22/2017 03:40 AM, Maxime Boissonneault wrote:
Thanks Robert,
While I understand and am all in favor of reusing dependencies, most
packages with dependencies do not provide download hooks for specific
versions of their dependencies. PETSc is this way
nload-spai=1
- --download-chaco=1 --download-sundials=1 --download-triangle=1
- --download-parms=1 --download-hdf5=1'
and that works.
Inge
On 02/22/17 07:53, Åke Sandgren wrote:
On 02/22/2017 03:40 AM, Maxime Boissonneault wrote:
Thanks Robert, While I understand and am all in favor
more like examples
and helpful defaults.
As a related aside, this is a pretty big deal with things like
tensorflow. These packages do end up being more like "Bundles" of
software and dependencies. Disk space is relatively cheap,
On Tue, Feb 21, 2017 at 1:41 PM Maxime Boissonneault
&l
.
# cd ..
# tar cfJ ${ARCHIVE} ${SRCDIR} && xz -t ${ARCHIVE} && rm -rf ${SRCDIR}
Maxime
On 17-02-21 13:31, Maxime Boissonneault wrote:
Hi,
I am looking at installing PETSc through EasyBuild. I am surprised to
see that the EasyBlock relies heavily on other recipes. My personal
exp
--download-ml --download-suitesparse --download-hypre
Is this contrary to other people's experience ?
--
-
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Président - Comité de coordination du soutien à la recherche de Calcul Q
Yes. Though my question is not how to restrict access, but whether
people do it or not for NAMD at their site.
Maxime
On 17-02-20 13:40, Bennet Fauber wrote:
Robert's scheme would apply to any software module one wished to
restrict access, to, yes?
-- bennet
municate with the NAMD
developers which users have access to the software, even though I
might recall wrong and that might be for another software.
I hope that helps!
Damian
On 20/02/17 17:11, "Maxime Boissonneault"
mailto:maxime.boissonnea...@calculquebec.ca>&
site ?
--
-
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Président - Comité de coordination du soutien à la recherche de Calcul Québec
Team lead - Research Support National Team, Compute Canada
Instructeur Software Carpentry
Ph. D. en physique
In any case: Maxime, please open a feature request to support this
more easily via --list-toolchains or some derivative.
Done here:
https://github.com/hpcugent/easybuild-framework/issues/2118
Maxime
Hi Kenneth,
The proper way would be making --list-toolchains a bit more
intelligent, and maybe support the notion of
--list-available-toolchains or something like that.
For now, this seems to be doing what you're after:
module avail $(eb --list-toolchains | grep -v "List of known" |
cut
On 17-02-14 14:13, Kenneth Hoste wrote:
On 06/02/2017 17:58, Maxime Boissonneault wrote:
Hi all,
Is there a way for easybuild to tell me what are the toolchains which
it currently sees as installed on the system ?
Note that we have a custom naming scheme, and that we hide toolchains
to
Thanks.
Are you planning to do some video edition (of Bart's presentation in
particular) ?
Maxime
On 17-02-10 11:35, Kenneth Hoste wrote:
On 10/02/2017 16:40, Maxime Boissonneault wrote:
Hi Alan and Kenneth,
Are the streams from yesterday available ?
Wednesday:
opening+Singul
f. Dr. Sebastian M. Schmidt
--
---------
Maxime Boissonneault
Analyst
Hi all,
Is there a way for easybuild to tell me what are the toolchains which it
currently sees as installed on the system ?
Note that we have a custom naming scheme, and that we hide toolchains to
users.
Thanks,
--
-
Maxime Boissonneault
Analyste de
just add a u to the beginning of '\n'
so u'\n'.join(test_report)
there must be a non-ascii string somewhere in there. Not sure where
exactly, but by default it would be smarter to handle these strings as
unicode anyway.
On Tue, Jan 24, 2017 at 10:14 AM
sting.py",
line 231, in create_test_report
return '\n'.join(test_report)
UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position
364: ordinal not in range(128)
Any idea how to fix this ? As a workaround, I'm setting LANG=en.
--
---
ok like it would require a modification to the easyblock.
There are probably some other places this would be useful too.
On Tue, Jan 24, 2017 at 9:12 AM Maxime Boissonneault
<mailto:maxime.boissonnea...@calculquebec.ca>> wrote:
Hi all,
I need to install Intel TBB as a depend
be created, or should it be a variation
of the existing easyblock ?
Thank you for your guidance,
--
-
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Université Laval
Président - Comité de coordination du soutien à la recherche de Calcul Québec
Team
On 16-08-24 11:48, Vanzo, Davide wrote:
Maxime,
I agree that the current toolchain version numbering is quite confusing.
One question about date numbering: what if somebody wants to downgrade
one of the toolchain elements? Maybe a micro version number would
still be useful for this task.
Tool
ay, Inc.
--
Eliot Eshelman, Vice President
Strategic Accounts and HPC Initiatives
Microway, Inc.
12 Richards Road, Plymouth, MA 02360
(508) 732-5534
el...@microway.com
http://www.microway.com
--
-
Maxime Boissonneault
Analyste de calcul - Calcul Québec, Universit
On 16-08-23 07:10, Kenneth Hoste wrote:
On 22/08/16 15:59, Maxime Boissonneault wrote:
Thanks, but what I meant is to have EasyBuild generate a wheel house,
rather than install them from a wheel (i.e. build the wheels rather
than install them as modules).
Hmm, what would that involve
On 16-08-22 09:50, Kenneth Hoste wrote:
For Python packages, R packages, Perl packages, Lua packages, etc.,
we usually instruct our users to either install them in their home
(using pip --user for Python for example), or using a virtual
environment (for Python).
In order to avoid build headac
Don't Numpy and Scipy require MKL or some optimized BLAS version though
(it should) ?
Also, doesn't mpi4py require MPI ?
How are these not "additional deps" ?
For Python packages, R packages, Perl packages, Lua packages, etc., we
usually instruct our users to either install them in their home (
Hi Todd,
Looking at what Singularity is (http://singularity.lbl.gov/), that is a
container technology, I would say that this is not the kind of thing
that you want to install through EasyBuild...
This is something that should be installed by an admin, directly on the
node.
Maxime
On 16-08
axime
On 16-08-11 09:37, Kenneth Hoste wrote:
Hi Maxime,
On 11/08/16 15:15, Maxime Boissonneault wrote:
Hi,
I am looking at modules generated with EasyBuild. I have the
following config parameters:
modules-tool: Lmod
module-naming-scheme=HierarchicalMNS
module-syntax=Lua
For some reason, th
Hi,
I am looking at modules generated with EasyBuild. I have the following
config parameters:
modules-tool: Lmod
module-naming-scheme=HierarchicalMNS
module-syntax=Lua
For some reason, the GCC module contains a conflict over GCC:
[mboisson@colosse1 easybuild]$ cat modules/all/Core/GCC/4.9.2.
85 matches
Mail list logo