Skip to content

Latest commit

 

History

History
executable file
·
60 lines (37 loc) · 2.36 KB

artifact_readme.md

File metadata and controls

executable file
·
60 lines (37 loc) · 2.36 KB
Raw

Description of included files

  • analysis: directory where log files will be stored
  • helper-scripts: useful script for running several benchmarks or extracting information from log files
  • squares: source code for the tool
  • tests-examples: directory that contains all benchmarks
  • tests: directory that is used to specify which benchmarks should be run when using the helper scripts. Sub-directories from tests-examples can either be symlinked to here, or copied over. The TODO!!!!!!
  • sequential.py, portfolio.py, cubes.py: files to run the corresponding configurations.
  • miniconda.sh, env.tar: setup files containing all required dependencies

Installation

Note: all commands should be run in the CUBES directory

Install and setup conda:

./miniconda.sh -b -p $HOME/anaconda
eval "$(/home/tacas21/anaconda/bin/conda shell.bash hook)"
conda init

Extract conda environment:

tar -xvf env.tar

Activate conda environment (this should be run every time a new shell is opened):

conda activate ./env

Running

Note: all commands should be run in the CUBES directory

Running a single benchmark

python sequential.py tests-examples/55-tests/1.yaml
python portfolio.py tests-examples/55-tests/1.yaml TODO!!!
python cubes.py tests-examples/55-tests/1.yaml

Parameters

You can use -h or --help in any of the 3 files to see what parameters are available.

Important parameters

By default, cubes.py uses the number of available logical processors minus two. This can be changed using the -j option. For example, if you wish to use 8 threads, you would use -j 8.

It should be taken into account that without disabling some features (TODO say which?), cubes.py should not be run with less than 4 threads.

Running all benchmarks present in the tests directory

python helper-scripts/benchmark.py RUN_IDENTIFIER
python helper-scripts/benchmark.py --portfolio RUN_IDENTIFIER
python helper-scripts/benchmark.py --cubes RUN_IDENTIFIER

You should replace RUN_IDENTIFIER with a string that will be used as a name for the directory containing the log files for the benchmarks, as well as a CSV file containing the results.

Important parameters

  • -t specifies the time limit for each benchmark in seconds (600 seconds by default)
  • -m specifies the memory limit for each benchmark in MB (57344 MB by default (56GB))