Bazel

Bazel
CC BY-SA 3.0 · source
Name	Bazel
Developer	Google
Released	2015
Programming language	Java, Python, C++
Operating system	Linux, macOS, Windows
License	Apache License 2.0

Bazel is a build and test tool created to automate software compilation, packaging, and continuous integration tasks for large codebases. It originated at Google and is used to produce reproducible, hermetic builds for projects spanning multiple languages and platforms. Designed for performance and scalability, it emphasizes incremental builds, remote execution, and declarative build rules suitable for monorepos and polyglot environments.

Overview

Bazel provides a high-performance build system influenced by earlier systems such as Blaze (software), Make (software), Ninja (build system), Buck (build tool), and concepts from Distributed computing and Meritocracy (organizational model). It targets scenarios found at Google, Facebook, Twitter, Microsoft, and Uber where monolithic repositories and complex dependency graphs require fast, reproducible outputs. Bazel integrates with continuous integration systems like Jenkins, Bazelisk ecosystems, and cloud providers including Google Cloud Platform, Amazon Web Services, and Microsoft Azure through remote caching and remote execution APIs.

History and Development

Development began inside Google as an internal system named Blaze and was later open-sourced in 2015 under the stewardship of engineers who had also worked with Sun Microsystems, Oracle Corporation, and projects influenced by LLVM. Early public attention came from conference presentations at Google I/O, O'Reilly Fluent Conference, and talks at Strange Loop. Adoption grew as major technology companies such as Stripe, Airbnb, Square (company), and LinkedIn evaluated Bazel for monorepo builds. Governance evolved toward a community model involving contributors from Red Hat, Apple Inc., Intel, and independent maintainers coordinating via GitHub and working groups inspired by processes at The Apache Software Foundation and Eclipse Foundation.

Architecture and Concepts

The core of Bazel implements design principles reminiscent of Hermetic build practices used at Google and ideas from Functional programming-influenced build graphs. Key concepts include a directed acyclic graph (DAG) of dependencies like those in Graph theory, remote execution similar to gRPC-based services, and sandboxing techniques akin to Containerization approaches used by Docker and Kubernetes. Bazel rules are written in Starlark, a dialect influenced by Python (programming language) and concepts from Scheme (programming language). Outputs follow content-addressable storage patterns comparable to Git objects and IPFS content addressing. The execution model supports local caches and remote build farms similar to Bazel Remote Execution API deployments used by cloud providers and build accelerators from Google Cloud Build and CircleCI.

Features and Workflow

Bazel emphasizes incremental builds, parallel execution, and reproducibility. Typical workflows interoperate with version control systems like Git, issue trackers such as JIRA, and CI systems including Travis CI and CircleCI. Features include fine-grained dependency analysis derived from techniques used in Static program analysis, remote caching used by Content Delivery Network architectures, sandboxed action execution akin to chroot and seccomp-style isolation, and test selection similar to approaches used in JUnit and Google Test. Tooling integrates with IDEs like IntelliJ IDEA, Visual Studio Code, and Eclipse, and supports distributed build acceleration found in solutions from BuildBarn and RBE (Remote Build Execution) providers.

Language and Platform Support

Bazel has first-class support for languages including Java (programming language), C++, Python (programming language), and Go (programming language), plus extension rules for Scala, Kotlin, Android (operating system), and Objective-C. Platform targets range across Linux, macOS, and Microsoft Windows, and mobile targets such as Android (operating system) and iOS via rules inspired by Bazel Rules community contributions and SDK integrations from Android SDK and Xcode. Language interoperability leverages toolchains and compilers like GCC, Clang, and OpenJDK, and bundling formats mirror those used by Docker images and Debian packages.

Use Cases and Adoption

Bazel is used for large-scale monorepos, microservices compilation, mobile app builds, and embedded systems where repeatability and speed are critical. Notable adopters or evaluators include Google, Pinterest, Spotify, Dropbox, Salesforce, and research projects at MIT, Stanford University, and Carnegie Mellon University. Industry use cases span cloud services, client applications, and systems programming, often combined with CI/CD pipelines leveraging Kubernetes clusters, Jenkins X, and remote caching networks implemented by companies such as Gradle Inc. and consulting firms like ThoughtWorks.

Community and Governance

The Bazel ecosystem is supported by an open-source community of engineers, maintainers, and corporate contributors coordinated through GitHub, mailing lists, special interest groups similar to Wikimedia Foundation working groups, and meetups at conferences such as KubeCon, SREcon, and Google I/O. Stewardship involves corporate sponsors including Google and contributions from companies like Red Hat, Intel, and Apple Inc., with roadmaps discussed in public proposals similar to governance processes at The Linux Foundation. The project hosts an ecosystem of third-party rule sets and integrations maintained by community projects and organizations such as Bazelbuild-affiliated repositories and independent contributors.

Category:Build systems