JCTools
High-performance concurrent data structures for the JVM, supplementing the Java standard library.
What is JCTools?
JCTools is a Java library that provides high-performance concurrent data structures, specifically queues, which are not available in the standard JDK. It solves the need for efficient, lock-free multi-threaded communication in applications where the built-in Java collections fall short on performance or concurrency patterns.
Java developers building high-throughput, low-latency systems such as networking frameworks, data processing pipelines, or real-time applications that require optimized multi-producer/multi-consumer queues.
Developers choose JCTools for its proven performance, stability, and adoption by major projects like Netty and RxJava. It offers a free, open-source alternative with specialized queue implementations that outperform standard JDK options in concurrent scenarios.
Overview
JCTools provides a collection of concurrent data structures that are missing from the Java Development Kit (JDK). It offers high-performance, lock-free queues and utilities designed for demanding multi-threaded applications, used by major frameworks like Netty and RxJava.
Key Features
- SPSC/MPSC/SPMC/MPMC Queues — Wait-free or lock-less concurrent queues for various producer-consumer patterns (Single/Multi Producer, Single/Multi Consumer).
- Linked Array Queues — Bounded and unbounded queues balancing performance, memory allocation, and footprint.
- XAdd-Based Queues — Unbounded linked array queues using XADD to reduce producer contention and pooled chunks to minimize allocation.
- Expanded Queue Interface —
MessagePassingQueuewith relaxed offer/peek/poll for performance trade-offs and batch drain/fill methods for higher throughput. - Implementation Variations — Queues available in
Unsafe(default),Atomic(usingAtomicFieldUpdater), andUnpadded(reduced memory footprint) versions.
Philosophy
JCTools focuses on delivering excellent performance at zero cost (free under Apache 2.0 License) while maintaining stability and community-driven development.
Use Cases
Best For
- Implementing high-performance message passing between threads in networking servers
- Building low-latency data processing pipelines with multiple producers or consumers
- Replacing JDK concurrent queues where higher throughput is required
- Developing real-time applications needing wait-free or lock-less synchronization
- Reducing contention in multi-threaded producer scenarios using XADD-based queues
- Minimizing memory footprint in concurrent systems with unpadded queue variations
Not Ideal For
- Projects where Java's built-in ConcurrentLinkedQueue or ArrayBlockingQueue already meet performance and simplicity needs
- Environments with strict policies against using sun.misc.Unsafe due to JVM portability concerns
- Teams prioritizing extensive out-of-the-box documentation and beginner-friendly tutorials over raw performance
- Low-concurrency applications where lock-based synchronization is sufficient and easier to debug
Pros & Cons
Pros
Offers wait-free or lock-less implementations that benchmark faster than JDK concurrent queues, as evidenced by adoption in high-throughput frameworks like Netty and RxJava.
Supports SPSC, MPSC, SPMC, and MPMC queue variations, catering to diverse multi-threaded scenarios with optimized designs for different producer-consumer configurations.
Provides unpadded queue versions to reduce false sharing and minimize memory footprint, balancing performance with resource usage.
MessagePassingQueue interface includes drain and fill methods for high-throughput batch processing, reducing contention and improving throughput in demanding applications.
Widely used by major projects like Netty, indicating reliability and community trust, with steady updates and issue tracking on GitHub.
Cons
Default implementations rely on sun.misc.Unsafe, an internal JVM API that may not be portable across all Java versions or vendors, risking future compatibility.
For zero-dependency projects, requires shading or manual jar inclusion via Maven/Gradle plugins, adding build step overhead as noted in the README's 'ShadeJCToolsSamples' reference.
Beyond the brief README, detailed usage guides are sparse, with discussion primarily on GitHub issues, which can hinder quick onboarding.
Some implementations are in experimental modules and may not be stable for production, with potential breaking changes or bugs as development evolves.
Frequently Asked Questions
Related Projects
AgronaHigh Performance data structures and utility methods for Java
Eclipse CollectionsEclipse Collections is a collections framework for Java with optimized data structures and a rich, functional and fluent API.
KolobokeJava Collections till the last breadcrumb of memory and performance
DisruptorInter-thread messaging library
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.