Like any good sysadmin, I strongly prefer to install software using my OS's
built-in package management system. Unfortunately, a lot of the software
used in scientific high-performance computing doesn't make this easy.
Many popular software projects, including both user applications and common
libraries, don't distribute packages but expect their users to build from source.
And many of these projects have non-standard, arcane, or just-plain-weird
build processes.
This post consists of some notes on a workflow I've used for building scientific software and
producing RPMs which can be used to distribute that software in the future.
I use this workflow mostly for personal projects -- building and testing
clusters on Amazon EC2, for example. (Disclaimer: This is not the workflow in production use
at the Day Job, where the software management problems are much more challenging!)
The core components of this workflow are EasyBuild --
which automates building HPC software projects -- and FPM --
which makes it easy to build OS packages from a directory.
Build server configuration
For building my packages I typically use a c4.xlarge server on
Amazon EC2. The c4.xlarge has the advantage of being
relatively powerful (typically I get 4 Sandy Bridge CPUs when I launch one)
while also being relatively cheap (spot price is currently hovering around
$0.032/hr). I run CentOS 6.x on the server as this is typical of the HPC
environments I generally use.
The following Ansible script installs a set of
basic development tools (the "@development" yum group), EasyBuild, and FPM.
It also pulls down my "module2pkg" wrapper script for packaging modules
with FPM. (More on that later.)
{% codeblock %} {% raw %}
---
- hosts: builder
user: root
tasks:
- name: install dev and packaging tools
yum: name={{item}} state=present
with_items:
- "@development"
- "environment-modules"
- "python-pip"
# Set up module2pkg script
- name: install fpm
gem: name=fpm state=present
- name: pull down module2pkg script
git: repo="https://github.com/ajdecon/module2pkg" dest=/opt/module2pkg
- name: symlink for running module2pkg
file: src="/opt/module2pkg/module2pkg" dest="/usr/local/bin/module2pkg" state=link
# Set up and configure EasyBuild
- name: easybuild user for building software
user: name=easybuild createhome=yes state=present
- name: /opt/easybuild where EB sw will go
file: path="/opt/easybuild" owner=easybuild mode=0755 state=directory
- name: install easybuild
pip: name=easybuild state=present
- name: ensure easybuild user .config dir exists
file: path="/home/easybuild/.config/easybuild" owner=easybuild mode=0755
state=directory
- name: ensure ~easybuild/.config/config.cfg exists
file: path="/home/easybuild/.config/easybuild/config.cfg" owner=easybuild
mode=0644 state=touch
- name: configure prefix
ini_file: dest="/home/easybuild/.config/easybuild/config.cfg" section="config"
option="prefix" value="/opt/easybuild"
{% endraw %} {% endcodeblock %}
Building Modules with EasyBuild
EasyBuild is a framework for building
and installing scientific applications on HPC systems. It comes with a huge
library of build scripts for commonly-used compilers, support libraries, and
simulations -- all of which can be built and installed just by running the
appropriate EasyBuild command.
(This is hugely useful because many of these applications have horrible
build processes. While many can be built with a regular "configure; make; make
install" workflow, many others include custom build scripts or have a long
list of required-but-arcane build options.)
EasyBuild also generates Module files for each software package it builds, which
is very convenient in HPC environments where we may have many users who want to
easily use different versions of each software package. (For example, it's not
uncommon to have strict dependencies on specific compiler or MPI versions -- so
a good HPC platform should make it easy to use those versions.)
EasyBuild resolves dependencies, and can be run with a "dry run" option which will show you what
software it will build to support any given application. For example, this is the output
when you request a dry run for running a particular recipe for
GROMACS:
[easybuild@ip-10-0-0-88 ~]$ eb GROMACS-4.6.5-gmpolf-1.4.8-hybrid.eb --robot --dry-run
== temporary log file in case of crash /tmp/eb-2U1YxO/easybuild-jmFjre.log
Dry run: printing build status of easyconfigs and dependencies
* [ ] /usr/easybuild/easyconfigs/g/GCC/GCC-4.8.1.eb (module: GCC/4.8.1)
* [ ] /usr/easybuild/easyconfigs/m/MPICH/MPICH-3.0.4-GCC-4.8.1.eb (module: MPICH/3.0.4-GCC-4.8.1)
* [ ] /usr/easybuild/easyconfigs/g/gmpich/gmpich-1.4.8.eb (module: gmpich/1.4.8)
* [ ] /usr/easybuild/easyconfigs/o/OpenBLAS/OpenBLAS-0.2.6-gmpich-1.4.8-LAPACK-3.4.2.eb (module: OpenBLAS/0.2.6-gmpich-1.4.8-LAPACK-3.4.2)
* [ ] /usr/easybuild/easyconfigs/f/FFTW/FFTW-3.3.3-gmpich-1.4.8.eb (module: FFTW/3.3.3-gmpich-1.4.8)
* [ ] /usr/easybuild/easyconfigs/s/ScaLAPACK/ScaLAPACK-2.0.2-gmpich-1.4.8-OpenBLAS-0.2.6-LAPACK-3.4.2.eb (module: ScaLAPACK/2.0.2-gmpich-1.4.8-OpenBLAS-0.2.6-LAPACK-3.4.2)
* [ ] /usr/easybuild/easyconfigs/g/gmpolf/gmpolf-1.4.8.eb (module: gmpolf/1.4.8)
* [ ] /usr/easybuild/easyconfigs/n/ncurses/ncurses-5.9-gmpolf-1.4.8.eb (module: ncurses/5.9-gmpolf-1.4.8)
* [ ] /usr/easybuild/easyconfigs/c/CMake/CMake-2.8.12-gmpolf-1.4.8.eb (module: CMake/2.8.12-gmpolf-1.4.8)
* [ ] /usr/easybuild/easyconfigs/g/GROMACS/GROMACS-4.6.5-gmpolf-1.4.8-hybrid.eb (module: GROMACS/4.6.5-gmpolf-1.4.8-hybrid)
== temporary log file /tmp/eb-2U1YxO/easybuild-jmFjre.log has been removed.
== temporary directory /tmp/eb-2U1YxO has been removed.
EasyBuild will first build all the modules which GROMACS depends on, all the way
down to the compiler stack. This may seem a little unnecessary -- after all, doesn't
the system come with a perfectly good compiler? However, it provides an easy way to
decouple your HPC software stack from the underlying OS, so that you can perform
OS upgrades for security or other issues while having less of an impact on your
users' software. It also means that you can build those lower level components
with options which are useful for HPC, but which your system vendor may not
have used in their stock build.
(A note on EC2: When you launch a particular
instance type, like c4.xlarge, you may get Sandy Bridge CPUs... or you may get
Westmeres or Ivy Bridges. The CPUs may also not actually support all the options
that they should: I initially tried to build software with EasyBuild's default
optimization levels which include -march=native and got "instruction not
supported" errors for some of the AVX instructions. Given that I am going to be
building RPMs anyway, I want them to be as general
as possible. So when building on EC2, I typically use the EasyBuild option
--optarch='' to turn off CPU-specific optimizations.)
Now execute without the dry run:
[easybuild@ip-10-0-0-88 ~]$ eb GROMACS-4.6.5-gmpolf-1.4.8-hybrid.eb --robot --optarch=''
And wait a long time, as we're starting with GCC:
== temporary log file in case of crash /tmp/eb-AGI9Yh/easybuild-DLAzfR.log
== resolving dependencies ...
== processing EasyBuild easyconfig /usr/easybuild/easyconfigs/g/GCC/GCC-4.8.1.eb
== building and installing GCC/4.8.1...
== fetching files...
== creating build dir, resetting environment...
== unpacking...
== patching...
== preparing...
== configuring...
== building...
Once everything has finished building, we can see the new software
available in Modules:
[easybuild@ip-10-0-0-88 ~]$ module use /opt/easybuild/modules/all/
[easybuild@ip-10-0-0-88 ~]$ module av
---------------------------------------------------- /opt/easybuild/modules/all/ -----------------------------------------------------
CMake/2.8.10.2-gmpolf-1.4.8 ScaLAPACK/2.0.2-gmpich-1.4.8-OpenBLAS-0.2.6-LAPACK-3.4.2
CMake/2.8.12-gmpolf-1.4.8 bzip2/1.0.6-gmpolf-1.4.8
FFTW/3.3.3-gmpich-1.4.8 gettext/0.18.2-gmpolf-1.4.8
GCC/4.8.1 gmpich/1.4.8
GLib/2.34.3-gmpolf-1.4.8 gmpolf/1.4.8
GROMACS/4.6.5-gmpolf-1.4.8-hybrid libffi/3.0.13-gmpolf-1.4.8
HPL/2.1-gmpolf-1.4.8 libreadline/6.2-gmpolf-1.4.8
MPICH/3.0.4-GCC-4.8.1 ncurses/5.9-gmpolf-1.4.8
OpenBLAS/0.2.6-gmpich-1.4.8-LAPACK-3.4.2 zlib/1.2.7-gmpolf-1.4.8
Python/2.7.3-gmpolf-1.4.8
--------------------------------------------------- /usr/share/Modules/modulefiles ---------------------------------------------------
dot module-git module-info modules null use.own
Packaging the Modules with FPM
FPM is a wonderful tool which
makes it very easy to build Linux packages such as RPMs and DEBs. While I'll
sometimes still go to the trouble of writing a real RPM SPEC file if I'm putting
together a maintainable build process, for ad-hoc work I will always prefer to
use FPM.
For working with Modules, I've written a simple helper script called
module2pkg which makes it a little
easier to build RPMs when you already have Modules. When you give it the name of a
Module, it
- Reads the Modulefile to get the Module's root directory
- Resolves the dependencies of that Module recursively, by checking to see which modules it
either loads or lists as prerequisites
- Runs FPM to package each Module in the dependency tree, including specifying its dependencies
(EasyBuild is actually working on integrating fpm support
for building packages, at which point I will mostly be able to retire module2pkg.
The only advantage of my approach is that it depends only on the Modulefiles themselves,
so it can be used to package hand-compiled Modules as well as those
built by EasyBuild.)
module2pkg also includes an option for adding a prefix to each package name --
in this case "eb". This lets me package things like GCC without worrying about
any potential conflict with the system packages.
Note that module2pkg prints out the fpm command it uses for each module.
[root@ip-10-0-0-88 ~]# module2pkg -p eb HPL/2.1-gmpolf-1.4.8
fpm -s dir -t rpm -n eb-GCC -v 4.8.1 -p eb-GCC-VERSION_ARCH.rpm -C / /opt/easybuild/software/GCC/4.8.1 /opt/easybuild/modules/all/GCC/4.8.1
no value for epoch is set, defaulting to nil {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
Created package {:path=>"eb-GCC-4.8.1_x86_64.rpm"}
fpm -s dir -t rpm -n eb-MPICH -v 3.0.4-GCC-4.8.1 -p eb-MPICH-VERSION_ARCH.rpm -d "eb-GCC = 4.8.1" -C / /opt/easybuild/software/MPICH/3.0.4-GCC-4.8.1 /opt/easybuild/modules/all/MPICH/3.0.4-GCC-4.8.1
Package version '3.0.4-GCC-4.8.1' includes dashes, converting to underscores {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
Created package {:path=>"eb-MPICH-3.0.4_GCC_4.8.1_x86_64.rpm"}
fpm -s dir -t rpm -n eb-OpenBLAS -v 0.2.6-gmpich-1.4.8-LAPACK-3.4.2 -p eb-OpenBLAS-VERSION_ARCH.rpm -d "eb-MPICH = 3.0.4-GCC-4.8.1" -d "eb-gmpich = 1.4.8" -d "eb-GCC = 4.8.1" -C / /opt/easybuild/software/OpenBLAS/0.2.6-gmpich-1.4.8-LAPACK-3.4.2 /opt/easybuild/modules/all/OpenBLAS/0.2.6-gmpich-1.4.8-LAPACK-3.4.2
Package version '0.2.6-gmpich-1.4.8-LAPACK-3.4.2' includes dashes, converting to underscores {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
Created package {:path=>"eb-OpenBLAS-0.2.6_gmpich_1.4.8_LAPACK_3.4.2_x86_64.rpm"}
fpm -s dir -t rpm -n eb-gmpich -v 1.4.8 -p eb-gmpich-VERSION_ARCH.rpm -d "eb-MPICH = 3.0.4-GCC-4.8.1" -d "eb-GCC = 4.8.1" -C / /opt/easybuild/software/gmpich/1.4.8 /opt/easybuild/modules/all/gmpich/1.4.8
no value for epoch is set, defaulting to nil {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
Created package {:path=>"eb-gmpich-1.4.8_x86_64.rpm"}
fpm -s dir -t rpm -n eb-ScaLAPACK -v 2.0.2-gmpich-1.4.8-OpenBLAS-0.2.6-LAPACK-3.4.2 -p eb-ScaLAPACK-VERSION_ARCH.rpm -d "eb-MPICH = 3.0.4-GCC-4.8.1" -d "eb-OpenBLAS = 0.2.6-gmpich-1.4.8-LAPACK-3.4.2" -d "eb-GCC = 4.8.1" -d "eb-gmpich = 1.4.8" -C / /opt/easybuild/software/ScaLAPACK/2.0.2-gmpich-1.4.8-OpenBLAS-0.2.6-LAPACK-3.4.2 /opt/easybuild/modules/all/ScaLAPACK/2.0.2-gmpich-1.4.8-OpenBLAS-0.2.6-LAPACK-3.4.2
Package version '2.0.2-gmpich-1.4.8-OpenBLAS-0.2.6-LAPACK-3.4.2' includes dashes, converting to underscores {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
Created package {:path=>"eb-ScaLAPACK-2.0.2_gmpich_1.4.8_OpenBLAS_0.2.6_LAPACK_3.4.2_x86_64.rpm"}
fpm -s dir -t rpm -n eb-HPL -v 2.1-gmpolf-1.4.8 -p eb-HPL-VERSION_ARCH.rpm -d "eb-FFTW = 3.3.3-gmpich-1.4.8" -d "eb-gmpolf = 1.4.8" -d "eb-OpenBLAS = 0.2.6-gmpich-1.4.8-LAPACK-3.4.2" -d "eb-ScaLAPACK = 2.0.2-gmpich-1.4.8-OpenBLAS-0.2.6-LAPACK-3.4.2" -d "eb-gmpich = 1.4.8" -d "eb-MPICH = 3.0.4-GCC-4.8.1" -d "eb-GCC = 4.8.1" -C / /opt/easybuild/software/HPL/2.1-gmpolf-1.4.8 /opt/easybuild/modules/all/HPL/2.1-gmpolf-1.4.8
Package version '2.1-gmpolf-1.4.8' includes dashes, converting to underscores {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
Created package {:path=>"eb-HPL-2.1_gmpolf_1.4.8_x86_64.rpm"}
fpm -s dir -t rpm -n eb-FFTW -v 3.3.3-gmpich-1.4.8 -p eb-FFTW-VERSION_ARCH.rpm -d "eb-MPICH = 3.0.4-GCC-4.8.1" -d "eb-gmpich = 1.4.8" -d "eb-GCC = 4.8.1" -C / /opt/easybuild/software/FFTW/3.3.3-gmpich-1.4.8 /opt/easybuild/modules/all/FFTW/3.3.3-gmpich-1.4.8
Package version '3.3.3-gmpich-1.4.8' includes dashes, converting to underscores {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
Created package {:path=>"eb-FFTW-3.3.3_gmpich_1.4.8_x86_64.rpm"}
fpm -s dir -t rpm -n eb-gmpolf -v 1.4.8 -p eb-gmpolf-VERSION_ARCH.rpm -d "eb-FFTW = 3.3.3-gmpich-1.4.8" -d "eb-OpenBLAS = 0.2.6-gmpich-1.4.8-LAPACK-3.4.2" -d "eb-ScaLAPACK = 2.0.2-gmpich-1.4.8-OpenBLAS-0.2.6-LAPACK-3.4.2" -d "eb-gmpich = 1.4.8" -d "eb-MPICH = 3.0.4-GCC-4.8.1" -d "eb-GCC = 4.8.1" -C / /opt/easybuild/software/gmpolf/1.4.8 /opt/easybuild/modules/all/gmpolf/1.4.8
no value for epoch is set, defaulting to nil {:level=>:warn}
no value for epoch is set, defaulting to nil {:level=>:warn}
Created package {:path=>"eb-gmpolf-1.4.8_x86_64.rpm"}
[root@ip-10-0-0-88 ~]# ls *rpm
eb-FFTW-3.3.3_gmpich_1.4.8_x86_64.rpm eb-HPL-2.1_gmpolf_1.4.8_x86_64.rpm
eb-GCC-4.8.1_x86_64.rpm eb-MPICH-3.0.4_GCC_4.8.1_x86_64.rpm
eb-gmpich-1.4.8_x86_64.rpm eb-OpenBLAS-0.2.6_gmpich_1.4.8_LAPACK_3.4.2_x86_64.rpm
eb-gmpolf-1.4.8_x86_64.rpm eb-ScaLAPACK-2.0.2_gmpich_1.4.8_OpenBLAS_0.2.6_LAPACK_3.4.2_x86_64.rpm
Installing the resulting packages
To demonstrate that the new packages actually work, you can spin up a new server and
show you can install them. In this case, let's install GROMACS:
Running Transaction
Installing : eb-GCC-4.8.1-1.x86_64 1/8
Installing : eb-MPICH-3.0.4_GCC_4.8.1-1.x86_64 2/8
Installing : eb-gmpich-1.4.8-1.x86_64 3/8
Installing : eb-OpenBLAS-0.2.6_gmpich_1.4.8_LAPACK_3.4.2-1.x86_64 4/8
Installing : eb-ScaLAPACK-2.0.2_gmpich_1.4.8_OpenBLAS_0.2.6_LAPACK_3.4.2-1.x86_64 5/8
Installing : eb-FFTW-3.3.3_gmpich_1.4.8-1.x86_64 6/8
Installing : eb-gmpolf-1.4.8-1.x86_64 7/8
Installing : eb-GROMACS-4.6.5_gmpolf_1.4.8_hybrid-1.x86_64 8/8
Verifying : eb-MPICH-3.0.4_GCC_4.8.1-1.x86_64 1/8
Verifying : eb-ScaLAPACK-2.0.2_gmpich_1.4.8_OpenBLAS_0.2.6_LAPACK_3.4.2-1.x86_64 2/8
Verifying : eb-GROMACS-4.6.5_gmpolf_1.4.8_hybrid-1.x86_64 3/8
Verifying : eb-FFTW-3.3.3_gmpich_1.4.8-1.x86_64 4/8
Verifying : eb-OpenBLAS-0.2.6_gmpich_1.4.8_LAPACK_3.4.2-1.x86_64 5/8
Verifying : eb-gmpich-1.4.8-1.x86_64 6/8
Verifying : eb-GCC-4.8.1-1.x86_64 7/8
Verifying : eb-gmpolf-1.4.8-1.x86_64 8/8
Installed:
eb-GROMACS.x86_64 0:4.6.5_gmpolf_1.4.8_hybrid-1
Dependency Installed:
eb-FFTW.x86_64 0:3.3.3_gmpich_1.4.8-1 eb-GCC.x86_64 0:4.8.1-1
eb-MPICH.x86_64 0:3.0.4_GCC_4.8.1-1 eb-OpenBLAS.x86_64 0:0.2.6_gmpich_1.4.8_LAPACK_3.4.2-1
eb-ScaLAPACK.x86_64 0:2.0.2_gmpich_1.4.8_OpenBLAS_0.2.6_LAPACK_3.4.2-1 eb-gmpich.x86_64 0:1.4.8-1
eb-gmpolf.x86_64 0:1.4.8-1
Complete!
And then, as a test, run one of the GROMACS benchmarks:
[easybuild@ip-10-0-0-75 ~]$ wget ftp://ftp.gromacs.org/pub/benchmarks/gmxbench-3.0.tar.gz
--2015-04-12 16:50:33-- ftp://ftp.gromacs.org/pub/benchmarks/gmxbench-3.0.tar.gz
=> `gmxbench-3.0.tar.gz'
Resolving ftp.gromacs.org... 130.238.41.205
Connecting to ftp.gromacs.org|130.238.41.205|:21... connected.
Logging in as anonymous ... Logged in!
==> SYST ... done. ==> PWD ... done.
==> TYPE I ... done. ==> CWD (1) /pub/benchmarks ... done.
==> SIZE gmxbench-3.0.tar.gz ... 4060558
==> PASV ... done. ==> RETR gmxbench-3.0.tar.gz ... done.
Length: 4060558 (3.9M) (unauthoritative)
100%[============================================================================================>] 4,060,558 803K/s in 5.6s
2015-04-12 16:50:40 (712 KB/s) - `gmxbench-3.0.tar.gz' saved [4060558]
[easybuild@ip-10-0-0-75 ~]$ tar xzf gmxbench-3.0.tar.gz
[easybuild@ip-10-0-0-75 ~]$ cd d.dppc/
[easybuild@ip-10-0-0-75 d.dppc]$ grompp_mpi
...
NOTE 1 [file grompp.mdp]:
The Berendsen thermostat does not generate the correct kinetic energy
distribution. You might want to consider using the V-rescale thermostat.
Generated 91 of the 91 non-bonded parameter combinations
Generating 1-4 interactions: fudge = 1
Generated 91 of the 91 1-4 parameter combinations
Excluding 3 bonded neighbours molecule type 'DPPC'
turning all bonds into constraints...
Excluding 2 bonded neighbours molecule type 'SOL'
turning all bonds into constraints...
Analysing residue names:
There are: 1024 Other residues
There are: 23552 Water residues
Analysing residues not classified as Protein/DNA/RNA/Water and splitting into groups...
Number of degrees of freedom in T-Coupling group DPPC is 103422.73
Number of degrees of freedom in T-Coupling group SOL is 141310.27
NOTE 2 [file grompp.mdp]:
You are using a plain Coulomb cut-off, which might produce artifacts.
You might want to consider using PME electrostatics.
Largest charge group radii for Van der Waals: 0.190, 0.190 nm
Largest charge group radii for Coulomb: 0.190, 0.190 nm
This run will generate roughly 9 Mb of data
There were 2 notes
gcq#280: "There's Still Time to Change the Rope You're On" (Led Zeppelin)
[easybuild@ip-10-0-0-75 d.dppc]$ mpirun -np 2 mdrun_mpi
:-) G R O M A C S (-:
Great Red Owns Many ACres of Sand
:-) VERSION 4.6.5 (-:
Contributions from Mark Abraham, Emile Apol, Rossen Apostolov,
Herman J.C. Berendsen, Aldert van Buuren, Pär Bjelkmar,
Rudi van Drunen, Anton Feenstra, Gerrit Groenhof, Christoph Junghans,
Peter Kasson, Carsten Kutzner, Per Larsson, Pieter Meulenhoff,
Teemu Murtola, Szilard Pall, Sander Pronk, Roland Schulz,
Michael Shirts, Alfons Sijbers, Peter Tieleman,
Berk Hess, David van der Spoel, and Erik Lindahl.
Copyright (c) 1991-2000, University of Groningen, The Netherlands.
Copyright (c) 2001-2012,2013, The GROMACS development team at
Uppsala University & The Royal Institute of Technology, Sweden.
check out http://www.gromacs.org for more information.
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public License
as published by the Free Software Foundation; either version 2.1
of the License, or (at your option) any later version.
:-) mdrun_mpi (-:
Option Filename Type Description
------------------------------------------------------------
-s topol.tpr Input Run input file: tpr tpb tpa
-o traj.trr Output Full precision trajectory: trr trj cpt
-x traj.xtc Output, Opt. Compressed trajectory (portable xdr format)
-cpi state.cpt Input, Opt. Checkpoint file
-cpo state.cpt Output, Opt. Checkpoint file
-c confout.gro Output Structure file: gro g96 pdb etc.
-e ener.edr Output Energy file
-g md.log Output Log file
-dhdl dhdl.xvg Output, Opt. xvgr/xmgr file
-field field.xvg Output, Opt. xvgr/xmgr file
-table table.xvg Input, Opt. xvgr/xmgr file
-tabletf tabletf.xvg Input, Opt. xvgr/xmgr file
-tablep tablep.xvg Input, Opt. xvgr/xmgr file
-tableb table.xvg Input, Opt. xvgr/xmgr file
-rerun rerun.xtc Input, Opt. Trajectory: xtc trr trj gro g96 pdb cpt
-tpi tpi.xvg Output, Opt. xvgr/xmgr file
-tpid tpidist.xvg Output, Opt. xvgr/xmgr file
-ei sam.edi Input, Opt. ED sampling input
-eo edsam.xvg Output, Opt. xvgr/xmgr file
-j wham.gct Input, Opt. General coupling stuff
-jo bam.gct Output, Opt. General coupling stuff
-ffout gct.xvg Output, Opt. xvgr/xmgr file
-devout deviatie.xvg Output, Opt. xvgr/xmgr file
-runav runaver.xvg Output, Opt. xvgr/xmgr file
-px pullx.xvg Output, Opt. xvgr/xmgr file
-pf pullf.xvg Output, Opt. xvgr/xmgr file
-ro rotation.xvg Output, Opt. xvgr/xmgr file
-ra rotangles.log Output, Opt. Log file
-rs rotslabs.log Output, Opt. Log file
-rt rottorque.log Output, Opt. Log file
-mtx nm.mtx Output, Opt. Hessian matrix
-dn dipole.ndx Output, Opt. Index file
-multidir rundir Input, Opt., Mult. Run directory
-membed membed.dat Input, Opt. Generic data file
-mp membed.top Input, Opt. Topology file
-mn membed.ndx Input, Opt. Index file
Option Type Value Description
------------------------------------------------------
-[no]h bool no Print help info and quit
-[no]version bool no Print version info and quit
-nice int 0 Set the nicelevel
-deffnm string Set the default filename for all file options
-xvg enum xmgrace xvg plot formatting: xmgrace, xmgr or none
-[no]pd bool no Use particle decompostion
-dd vector 0 0 0 Domain decomposition grid, 0 is optimize
-ddorder enum interleave DD node order: interleave, pp_pme or cartesian
-npme int -1 Number of separate nodes to be used for PME, -1
is guess
-nt int 0 Total number of threads to start (0 is guess)
-ntmpi int 0 Number of thread-MPI threads to start (0 is guess)
-ntomp int 0 Number of OpenMP threads per MPI process/thread
to start (0 is guess)
-ntomp_pme int 0 Number of OpenMP threads per MPI process/thread
to start (0 is -ntomp)
-pin enum auto Fix threads (or processes) to specific cores:
auto, on or off
-pinoffset int 0 The starting logical core number for pinning to
cores; used to avoid pinning threads from
different mdrun instances to the same core
-pinstride int 0 Pinning distance in logical cores for threads,
use 0 to minimize the number of threads per
physical core
-gpu_id string List of GPU device id-s to use, specifies the
per-node PP rank to GPU mapping
-[no]ddcheck bool yes Check for all bonded interactions with DD
-rdd real 0 The maximum distance for bonded interactions with
DD (nm), 0 is determine from initial coordinates
-rcon real 0 Maximum distance for P-LINCS (nm), 0 is estimate
-dlb enum auto Dynamic load balancing (with DD): auto, no or yes
-dds real 0.8 Minimum allowed dlb scaling of the DD cell size
-gcom int -1 Global communication frequency
-nb enum auto Calculate non-bonded interactions on: auto, cpu,
gpu or gpu_cpu
-[no]tunepme bool yes Optimize PME load between PP/PME nodes or GPU/CPU
-[no]testverlet bool no Test the Verlet non-bonded scheme
-[no]v bool no Be loud and noisy
-[no]compact bool yes Write a compact log file
-[no]seppot bool no Write separate V and dVdl terms for each
interaction type and node to the log file(s)
-pforce real -1 Print all forces larger than this (kJ/mol nm)
-[no]reprod bool no Try to avoid optimizations that affect binary
reproducibility
-cpt real 15 Checkpoint interval (minutes)
-[no]cpnum bool no Keep and number checkpoint files
-[no]append bool yes Append to previous output files when continuing
from checkpoint instead of adding the simulation
part number to all file names
-nsteps step -2 Run this number of steps, overrides .mdp file
option
-maxh real -1 Terminate after 0.99 times this time (hours)
-multi int 0 Do multiple simulations in parallel
-replex int 0 Attempt replica exchange periodically with this
period (steps)
-nex int 0 Number of random exchanges to carry out each
exchange interval (N^3 is one suggestion). -nex
zero or not specified gives neighbor replica
exchange.
-reseed int -1 Seed for replica exchange, -1 is generate a seed
-[no]ionize bool no Do a simulation including the effect of an X-Ray
bombardment on your system
Reading file topol.tpr, VERSION 4.6.5 (single precision)
Using 2 MPI processes
starting mdrun 'DPPC in Water'
5000 steps, 10.0 ps.
Writing final coordinates.
Average load imbalance: 0.1 %
Part of the total run time spent waiting due to load imbalance: 0.0 %
Core t (s) Wall t (s) (%)
Time: 1177.360 589.282 199.8
(ns/day) (hour/ns)
Performance: 1.466 16.366
gcq#191: "Ich Bin Ein Berliner" (J.F. Kennedy)
It works!