NoisePage is a relational database management system developed by the Carnegie Mellon Database Group. The research goal of the NoisePage project is to develop high-performance system components that support autonomous operation and optimization as a first-class design principle.
- Integrated machine learning components to forecast, model, and plan the system's behavior.
- Postgres compatible wire-protocol, SQL, and catalogs.
- Apache Arrow compatible in-memory columnar storage.
- Lock-free multi-version concurrency control.
- Just-in-time query compilation using the LLVM.
- Vectorized execution using relaxed-operator fusion (ROF).
- 100% Open-Source (MIT License)
The NoisePage project is built and tested on Ubuntu 20.04 or macOS 10.14+. No other environments are officially supported.
git clone https://github.com/cmu-db/noisepage.git
cd noisepage
sudo ./script/installation/packages.sh
mkdir build
cd build
cmake -GNinja -DCMAKE_BUILD_TYPE=Release -DNOISEPAGE_USE_JEMALLOC=ON -DNOISEPAGE_UNITY_BUILD=ON ..
ninja noisepage
./bin/noisepage
If you are on a Mac, you may need to add this to your .zshrc
:
export PATH="/usr/local/opt/llvm@8/bin:$PATH"
export LIBRARY_PATH=$LIBRARY_PATH:/usr/local/opt/libpqxx/lib/
export CC=/usr/local/Cellar/llvm@8/8.0.1_3/bin/clang
export CXX=/usr/local/Cellar/llvm@8/8.0.1_3/bin/clang++
export LLVM_DIR=/usr/local/Cellar/llvm@8/8.0.1_3
export ASAN_OPTIONS=detect_container_overflow=0
You can now connect to NoisePage over the Postgres client psql
.
psql -h localhost -U noisepage -p 15721
Additional Notes:
- If you have less than 16 GB of RAM, use
-DNOISEPAGE_UNITY_BUILD=OFF
in thecmake
commands above. - If you know what you're doing, install the prerequisite packages from
./script/installation/packages.sh
manually.
Please see the docs.
If you are a current student at CMU,
- See the New Student Guide.
- Consider enrolling in one of the database courses.
Contributions from non-CMU students are also welcome!