EcoQuery: Sustainability Benchmark for Databases

EcoQuery and ATLAS for assessing the environmental impact of analytical databases et al.

EcoQuery is a benchmarking tool designed to measure the environmental impact of different database systems. It currently supports benchmarking with TPC-H and TPC-DS for DuckDB, Hyper, StarRocks, and MonetDB.

Overview

EcoQuery allows you to:

• Compare the energy consumption and environmental impact of different database systems
• Benchmark using industry-standard TPC-H and TPC-DS workloads
• Easily deploy and configure multiple database systems for testing

Database Installation

StarRocks

1. Download the StarRocks binary distribution package:

   wget https://releases.starrocks.io/starrocks/StarRocks-3.1.4.tar.gz
   

2. Extract the package:

   tar -xzvf StarRocks-3.1.4.tar.gz

3.Configure the installation path in the benchmark scripts (details below).

MonetDB

1. Follow the installation instructions at the official MonetDB repository:

   # Add the MonetDB repository
   sudo apt-get install -y apt-transport-https
   wget -O - https://www.monetdb.org/downloads/MonetDB-GPG-KEY | gpg --dearmor | sudo tee /etc/apt/trusted.gpg.d/monetdb.gpg > /dev/null
   sudo add-apt-repository "deb https://www.monetdb.org/downloads/deb/ $(lsb_release -cs) monetdb"
   
   # Update and install
   sudo apt-get update
   sudo apt-get install -y monetdb5-server monetdb-client
   

2. Configure the MonetDB dbfarm path in the benchmark scripts (details below).

DuckDB

1. Install DuckDB using pip:

   pip install duckdb

2. For additional installation options, refer to the DuckDB Python installation guide.

Hyper

1. Install Hyper using pip:

   pip install tableauhyperapi

2. For additional information, refer to the Hyper installation documentation.

Running Benchmarks

TPC-H Benchmark

To run the TPC-H benchmark, use the following command:

./run_tpch.sh {data_directory} {dbgen_tool_directory} {python_code_directory}   {data_generation_flag}

Parameters

• data_directory: Directory to store generated data files.
• dbgen_tool_directory: Directory containing the TPC-H dbgen tool.
• python_code_directory: Directory containing the EcoQuery Python scripts.
• data_generation_flag: Set to true to generate new data, or false to use existing data.

TPC-DS Benchmark

To run the TPC-DS benchmark, use the following command:

./run_tpcds.sh {data_directory} {dbgen_tool_directory} {python_code_directory}   {data_generation_flag}

Parameters

• data_directory: Directory to store generated data files.
• dbgen_tool_directory: Directory containing the TPC-H dbgen tool.
• python_code_directory: Directory containing the EcoQuery Python scripts.
• data_generation_flag: Set to true to generate new data, or false to use existing data.

Configuration

Database Paths

Before running the benchmarks, you need to configure the paths for the databases in the script files:

1. For MonetDB, define the dbfarm path in the ile=/etc/systemd/system/tpch_benchmark.env file, which holds the variables of service files:

   # Example: Modify this line in the script
   DATA_DIR="/path/to/monetdb/dbfarm"

2.For StarRocks, define the installation path:

# Example: Modify this line in the script
local starrocks_path=="/path/to/starrocks/installation"