The purpose of this program is to benchmark the sortedness performance on various indexes.
This research project is part of the Data-intensive Systems and Computing (DiSC) lab at Boston University.
-
Clone this repo
git clone https://github.com/BU-DiSC/bliss_benchmark.git
-
Create the build directory
cd bliss_benchmark && mkdir build
-
Create CMake configuration
cmake ..
-
Compile the program
make
To run the program, use:
./bliss_bench [OPTIONS]
The program currently accepts the following parameters:
-d, --data_file arg Path to the data file
-p, --preload_factor arg Preload factor (default: 0.5)
-w, --write_factor arg Write factor (default: 0.25)
-r, --read_factor arg Read factor (default: 0.1)
-m, --mixed_read_write_ratio arg Read write ratio (default: 0.5)
-s, --seed arg Random Seed value (default: 0)
-v, --verbosity [=arg(=1)] Verbosity [0: Info| 1: Debug | 2: Trace] (default: 0)
-i, --index arg Index type (alex|lipp) (default: btree)