How to Build Distributed Concurrency Primitives

Explore the challenges and solutions for implementing concurrency primitives in distributed systems through this 48-minute conference talk from GeeCON 2024. Discover why traditional Java concurrency tools like Atomics, Lock, CountDownLatch, and Semaphore cannot simply be exposed through REST or RPC APIs for distributed environments. Learn about the additional guarantees required for distributed concurrency primitives to function effectively across microservices and distributed systems. Examine practical examples including distributed Lock services that provide exclusive access to shared resources across multiple client instances, and AtomicReference services that ensure atomic visibility of shared value changes across dependent microservices. Understand the fundamental differences between local and distributed concurrency requirements, and delve into implementation strategies using the Raft consensus protocol as a foundation for building robust distributed concurrency primitives.