HeavyDB is an in-memory, column store, SQL relational database designed from the ground up to run on GPUs.
The repository includes a number of third party packages provided under separate licenses. Details about these packages and their respective licenses is at ThirdParty/licenses/index.md.
This project is licensed under the Apache License, Version 2.0.
The repository includes a number of third party packages provided under separate licenses. Details about these packages and their respective licenses is at ThirdParty/licenses/index.md.
Follow the instructions noted in CONTRIBUTIONING.md
Follow the instructions noted in Build From Source. This will build the entire HeavyAI platform which includes Immerse, WebServer, GEOS DSOs, and HeavyDB into a usable docker image.
Contributed code should compile without generating warnings by recent compilers on most Linux distributions. Changes to the code should follow the C++ Core Guidelines.
A .clang-format style configuration, based on the Chromium style guide, is provided at the top level of the repository. Please format your code using a recent version (9.0+ preferred) of ClangFormat before submitting.
To use:
clang-format -i File.cpp
A .clang-tidy configuration is provided at the top level of the repository. Please lint your code using a recent version (6.0+ preferred) of clang-tidy before submitting.
clang-tidy requires all generated files to exist before running. The easiest way to accomplish this is to simply run a full build before running clang-tidy. A build target which runs clang-tidy is provided. To use:
make run-clang-tidy
Note: clang-tidy may make invalid or overly verbose changes to the source code. It is recommended to first commit your changes, then run clang-tidy and review its recommended changes before amending them to your commit.
Note: the clang-tidy target uses the run-clang-tidy.py script provided with LLVM, which may depend on PyYAML. The target also depends on jq, which is used to filter portions of the compile_commands.json file.
Source files must carry the NVIDIA SPDX license header. For C/C++/CUDA/Java this is a leading comment block:
/*
* SPDX-FileCopyrightText: Copyright (c) 2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
* SPDX-License-Identifier: Apache-2.0
*/
(#-style comments are used for Python, shell, and CMake.) The header — including its copyright year range — is added and kept up to date automatically by a pre-commit hook. Install it once:
pip install pre-commit
pre-commit install
After that, every commit checks the files you changed; any missing or out-of-date header is fixed in place (re-stage and commit). Legacy HEAVY.AI/OmniSci/MapD headers are replaced with the NVIDIA header when a file is touched. Enforcement is going forward — only files you add or modify are checked, so the existing tree is not rewritten en masse. The same check runs in CI (the license-headers job in .github/workflows/pr-required-checks.yml).
Vendored third-party code keeps its upstream license and is exempt (ThirdParty/, Archive/, ODBC/, and the individual third-party files listed in .pre-commit-config.yaml).
To run the check manually against the files changed on your branch:
scripts/check_license_headers.sh # compares against origin/master
The maintained dependency scripts support Ubuntu 22.04 and Rocky
Linux 8.x. They install under /usr/local/mapd-deps and generate
/usr/local/mapd-deps/mapd-deps.sh, which must be sourced before configuring
HeavyDB:
source /usr/local/mapd-deps/mapd-deps.sh
To install a prepared dependency archive supplied with the developer or CI build context, use:
sudo scripts/mapd-deps-prebuilt.sh --static
Ubuntu also supports --shared; Rocky Linux supports static dependencies.
Pass --enable to install a profile entry that loads the generated
environment in future login shells.
To build dependency archives from source:
cd scripts
./mapd-deps-ubuntu.sh --static --compress
or, on Rocky Linux:
cd scripts
./mapd-deps-rockylinux.sh --compress
GPU builds also require a compatible NVIDIA driver and CUDA toolkit. Follow the NVIDIA CUDA installation instructions for the target distribution. For complete dependency-script details, see the dependency quickstart.
Warning
Do not report security vulnerabilities through public GitHub issues!
NVIDIA takes security seriously. If you discover a vulnerability in heavydb, DO NOT open a public issue. Use one of the private reporting channels described in SECURITY.md.
Join the HeavyAI GitHub Discussions to ask questions, share feedback, and report issues. HeavyAI maintainers review issues, discussions, and pull requests on a best effort basis without guaranteed response timelines.