Engine to generate POP perfomance metrics and translate Paraver (.prv, .row, .pcf) files into HDF5 file format.
To take a look to the available options and funcionalities:
$python kapitanPOP.py --help
To generate a .csv file with POP metrics:
$python kapitanPOP.py trace.prv
To translate a Paraver trace to HDF5:
$python kapitanPOP.py -p trace1.prv
Before using KapitanPOP you will need to install some dependencies and compile a small binary
- HDF5 library 1.12.0 release - C/Fortran library to store data on disk in HDF5 file format
- Pandas - Python module for data analysis
- Numpy - Package for scientific computing in Python
- Dask the version from GitHub's master branch - Parallel computing library for Python
- H5py - HDF5 for Python
- Dimemas - Network simulator tool for message-passing programs
You should be able to install Pandas, Numpy and h5py on your computer very easyly using pip install, for example.
pip install pandas numpy h5py
To install HDF5-1.12.0 the best approach is to download the sources and install them on your computer:
For Dask it is advisable to install it from source from its master branch on GitHub because the latest releases have some issues related to HDF5.
Dimemas is not a key dependency. You can run KapitanPOP without it, but you will miss some performance metrics (transfer and serialization efficiency).
After installing C/Fortran HDF5 library, you can run the Makefile to compile KapitanPOP's Paraver traces parser.
make install
It will generate an error if your HDF5 installation is not in a standard system directory. You can export the variable HDF5_HOME with the path where the library was installed, or modify the include and library paths in the Makefile to avoid the error.
Once the Paraver parser is generated and Python's dependencies satisfied, you are set up to execute kapitanPOP.py.
- Parsing of Paraver trace files (currently only .prv and .row) into the HDF5 file format.
- Generation of a CSV file with the available POP metrics from a list of Paraver or parsed HDF5 files.