POET
POET is a coupled reactive transport simulator implementing a parallel architecture and a fast, original MPI-based Distributed Hash Table.
External Libraries
The following external header library is shipped with POET:
- argh - https://github.com/adishavit/argh (BSD license)
Installation
Requirements
To compile POET you need several software to be installed:
- C/C++ compiler (tested with GCC)
- MPI-Implementation (tested with OpenMPI and MVAPICH)
- R language and environment
- CMake 3.9+
If you want to build documentation during compilation, doxygenand graphviz must be provided too.
The following R libraries must then be installed, which will get the needed dependencies automatically:
Compiling source code
The generation of makefiles is done with CMake. If you obtained POET from git, you should be able to generate Makefiles by running
mkdir build && cd build
cmake ..
This will create the directory build and processes the CMake files and generate Makefiles from it. You're now able to run make to start build
process.
If POET was obtained from the official SVN repository or the redmine at https://redmine.cs.uni-potsdam.de/projects/poet the branch or tag to be used have to be set via
mkdir build && cd build
cmake -D POET_SET_BRANCH="<BRANCH>" ..
where currently available branches/tags are:
- dev
If everything went well you'll find the executable at build/src/poet, but it is recommended to install the POET project structure to a desired CMAKE_INSTALL_PREFIX with make install.
During the generation of Makefiles, various options can be specified via cmake -D <option>=<value> [...]. Currently there are the following available options:
- DHT_Debug=boolean - toggles the output of detailed statistics about DHT
usage (
cmake -D DHT_Debug=ON). Defaults to OFF. - BUILD_DOC=boolean - toggles the generation of documantiation during compilation process. Defaults to ON.
- only from svn version: POET_SET_BRANCH=string - set branch or tag whose code is used
Example: Build from scratch
Assuming that only the C/C++ compiler, MPI libraries, R runtime environment and CMake have been installed, POET can be installed as follows:
# start R environment
$ R
# install R dependencies
> install.packages(c("devtools", "Rcpp", "RInside"))
> devtools::install_gitlab("delucia/RedModRphree", host="https://git.gfz-potsdam.de")
> devtools::install_gitlab("delucia/Rmufits", host="https://git.gfz-potsdam.de")
> q(save="no")
# cd into POET project root
$ cd <POET_dir>
# Build process
$ mkdir build && cd build
$ cmake -DCMAKE_INSTALL_PREFIX=/home/<user>/poet ..
$ make -j<max_numprocs>
$ make install
This will install a POET project structure into /home/<user>/poet which is
called hereinafter <POET_INSTALL_DIR>. With this version of POET we do not
recommend to install to hierarchies like /usr/local/ etc.
The correspondending directory tree would look like this:
.
└── poet/
├── bin/
│ └── poet
├── data/
│ └── SimDol2D.R
├── docs/
│ └── html/
│ ├── index.html
│ └── ...
└── R_lib/
├── kin_r_library.R
└── parallel_r_library.R
The R libraries will be loaded at runtime and the paths are hardcoded
absolute paths inside poet.cpp. So, if you consider to move bin/poet either
change paths of the R source files and recompile POET or also move R_lib/*
according to the binary.
To display the generated html documentation just open docs/html/index.html
with the browser of your choice.
Running
Before POET is ready to run, a working directory must be created. In this
directory you should find the executable file, the R scripts
<POET_ROOT>/R_lib/kin_r_library.R and <POET_ROOT>/R_lib/parallel_r_library.R
and the simulation description e.g. <POET_ROOT>/data/chem_problems/SimDol2D.R.
Run POET by mpirun ./poet <OPTIONS> <SIMFILE> <OUTPUT_DIRECTORY> where:
- OPTIONS - runtime parameters (explained below)
- SIMFILE - simulation described as R script (currently supported:
<POET_INSTALL_DIR>/data/SimDol2D.R) - OUTPUT_DIRECTORY - path, where all output of POET should be stored
Runtime options
The following parameters can be set:
| Option | Value | Description |
|---|---|---|
| --work-package-size= | 1..n | size of work packages (defaults to 5) |
| --ignore-result | disables store of simulation resuls | |
| --dht | enabling DHT usage (defaults to OFF) | |
| --dht-nolog | disabling applying of logarithm before rounding | |
| --dht-signif= | 1..n | set rounding to number of significant digits (defaults to 5) |
| --dht-strategy= | 0-1 | change DHT strategy. NOT IMPLEMENTED YET (Defaults to 0) |
| --dht-size= | 1-n | size of DHT per process involved in byte (defaults to 1 GiB) |
| --dht-snaps= | 0-2 | disable or enable storage of DHT snapshots |
| --dht-file= | <SNAPSHOT> |
initializes DHT with the given snapshot file |
Additions to dht-signif
Only used if no vector is given in setup file. For individual values per column
use R vector signif_vector in SIMFILE.
Additions to dht-snaps
Following values can be set:
- 0 = snapshots are disabled
- 1 = only stores snapshot at the end of the simulation with name
<OUTPUT_DIRECTORY>.dht - 2 = stores snapshot at the end and after each iteration iteration
snapshot files are stored in
<DIRECTORY>/iter<n>.dht
Example: Running from scratch
We will continue the above example and start a simulation with SimDol2D.R,
which is the only simulation supported at this moment. The required flow velocities
snapshots are included in the R package Rmufits. It's a 2D, 50x50 grid, with 20 time
steps. To start the simulation with 4 processes cd into your previously installed
POET-dir <POET_INSTALL_DIR>/bin and run:
mpirun -n 4 ./poet ../data/SimDol2D.R output
After a finished simulation all data generated by POET will be found in the
directory output.
You might want to use the DHT to cache previously simulated data and
reuse them in further time-steps. Just append --dht to the options of POET to
activate the usage of the DHT. The resulting call would look like this:
mpirun -n 4 ./poet --dht SimDol2D.R output
About the usage of MPI_Wtime()
Implemented time measurement functions uses MPI_Wtime(). Some important
information from the OpenMPI Man Page:
For example, on platforms that support it, the clock_gettime() function will be used to obtain a monotonic clock value with whatever precision is supported on that platform (e.g., nanoseconds).