GNU Guix is a transactional
package manager, with support for per-user package installations.
Users can install their own packages without interfering with each
other, yet without unnecessarily increasing disk usage or rebuilding
every package. Users can in fact create as many software environments
as they like—think of it
as VirtualEnv but not limited
to Python, or
modules but not
limited to your sysadmin-provided modules.
The software environments created with Guix are fully reproducible: a package built from a specific Guix commit on your laptop will be exactly the same as the one built on the HPC cluster you deploy it too, usually bit-for-bit.
We believe this makes Guix a great foundation for reproducible software deployment in high-performance computing (HPC). Here’s how to get started.
You can install Guix on your laptop in 5 minutes: just follow the binary install instructions.
You’re a cluster sysadmin and would like to have a cluster-wide install? Read this article.
Say you’re searching for a sparse solver among the 20,000+ packages that come with Guix:
$ guix search sparse solver name: mumps version: 5.0.2 outputs: out systems: x86_64-linux i686-linux armhf-linux aarch64-linux + mips64el-linux dependencies: gfortran-5.4.0 metis-5.1.0 openblas-0.2.19 + scotch-6.0.4 location: gnu/packages/maths.scm:1550:2 homepage: http://mumps.enseeiht.fr license: CeCILL-C synopsis: Multifrontal sparse direct solver description: MUMPS (MUltifrontal Massively Parallel sparse + direct Solver) solves a sparse system of linear equations + A x = b using Guassian elimination. relevance: 12 …
To install it along with the latest GNU compiler tool chain:
$ guix install mumps gcc-toolchain The following packages will be installed: mumps 5.2.1 gcc-toolchain 10.2.0 The following derivation will be built: /gnu/store/kipa9k61zkhw4s3frs92w683ps23hpjj-profile.drv 3.1 MB will be downloaded … hint: Consider setting the necessary environment variables by running: GUIX_PROFILE="$HOME/.guix-profile" . "$GUIX_PROFILE/etc/profile" Alternately, see `guix package --search-paths -p "$HOME/.guix-profile"'.
Spawning One-Off Environments
Sometimes all you want is to try out a program without installing it in
your profile. That’s
guix environment comes
in. To create an environment containing Python 3.x, NumPy, and
$ python3 bash: python3: Command not found $ guix environment --ad-hoc python@3 python-numpy python-scikit-learn The following derivation will be built: /gnu/store/2g3mj1xdlq2rj8j0crl4sa68bqhmfsmd-profile.drv building directory of Info manuals... building database for manual pages... [env]$ python3 Python 3.5.3 (default, Jan 1 1970, 00:00:01) [GCC 5.4.0] on linux Type "help", "copyright", "credits" or "license" for more information. >>> import numpy >>> import sklearn >>>
Occasionally you’ll want to customize the way packages are
From the command line,
you can apply transformations, such as replacing one dependency
with another one in the dependency graph. The example below replaces
openmpi-thread-multiple in the dependency graph of
$ guix install mumps-openmpi \ --with-input=openmpi=openmpi-thread-multiple
The expressivity of the command line is limited, but you can go further by writing your own package definitions.
To add a package, you
generate a template from a third-party repository using
write a package definition,
which looks like this:
(define-public scalapack (package (name "scalapack") (version "2.0.2") (source (origin (method url-fetch) (uri (string-append "http://www.netlib.org/scalapack/scalapack-" version ".tgz")) (sha256 (base32 "0p1r61ss1fq0bs8ynnx7xq4wwsdvs32ljvwjnx6yxr8gd6pawx0c")))) (build-system cmake-build-system) (inputs (list openmpi gfortran lapack)) (arguments `(#:configure-flags `("-DBUILD_SHARED_LIBS:BOOL=YES"))) (home-page "http://www.netlib.org/scalapack/") (synopsis "Library for scalable numerical linear algebra") (description "ScaLAPACK is a Fortran 90 library of high-performance linear algebra routines on parallel distributed memory machines. ScaLAPACK solves dense and banded linear systems, least squares problems, eigenvalue problems, and singular value problems.") (license (license:bsd-style "file://LICENSE" "See LICENSE in the distribution."))))
You can have your own package collection published as a channel.
Sending Packages to Guix-less Machines
laptop$ scp `guix pack -RR hwloc -S /bin=bin` supercomputer:hwloc.tar.gz … supercomputer$ mkdir -p ~/.local supercomputer$ (cd ~/.local; tar xf ~/hwloc.tar.gz) supercomputer$ ~/.local/bin/lstopo
Other options include building Singularity or Docker images.
Read about on-going Guix-HPC developments on our blog.