Knowledge Planet: Old Dock's Tech Chronicles Zhihu: SamDeepThinking A technical manager who still codes heavily on the front line. From junior developer to tech lead, then tech manager, now leading the whole front‑ and back‑end development team—leveling up along the way. I have some insights on programming, career development, and tech management.
This article demonstrates the evolution from synchronized/wait/notify to Lock/Condition for producer-consumer patterns, showing how notifyAll causes spurious wakeups with multiple threads and how Condition enables precise signaling via separate condition queues for producers and consumers.
This article compares Java's synchronized keyword with the Lock interface (ReentrantLock), demonstrating how Lock provides finer-grained control through timeout acquisition (tryLock), interruptible locking (lockInterruptibly), fair/non-fair policies, and manual unlock in finally blocks, with code examples for each scenario.
This article explains Java's CyclicBarrier through two practical examples: coordinating concurrent API stress tests and batch-importing one million Excel rows, then reveals a design flaw where the main thread participates in the barrier and shows how to separate concerns using CountDownLatch for cleaner batch processing.
The article explains why marketing activity tables require a status field (draft, approving, release, cancelled) alongside start_time and end_time, because time fields alone cannot represent manual termination, approval workflows, or operational interventions, and demonstrates SQL queries combining both dimensions.
This article explains the key differences between Java's CyclicBarrier and CountDownLatch, focusing on which threads wait, thread lifecycle after signaling, reusability, and a practical pressure-test example using CountDownLatch to simulate concurrent requests.
The article explains CountDownLatch's implementation as a thin wrapper over AQS, detailing how it overrides tryAcquireShared and tryReleaseShared with CAS for thread-safe countdown, and teaches a high-level-to-detail interview answering strategy.
This article explains Java's CountDownLatch synchronization tool using a meeting analogy, demonstrates its usage with a Java 17 code example simulating attendees arriving at a meeting, and describes a real-world batch translation scenario where the main thread waits for multiple concurrent translation tasks to complete.
This article explains Java Semaphore's core decoupling property—where any thread can acquire or release permits—contrasting it with ReentrantLock's thread-owned locking, demonstrates with Java 17 code examples, and shows real-world rate-limiting applications in Sentinel, Dubbo, and Tomcat.
This article explains thread starvation in Java concurrency, demonstrates it with a code example showing unfair lock acquisition and cumulative wait times, and provides two solutions: moving slow operations outside synchronized blocks and using ReentrantLock tryLock with timeout to bound waiting.
The article explains livelock using a RabbitMQ incident where unacknowledged messages cause endless re-delivery, and describes two production fixes: immediate try-catch with ACK to stop the loop, then bug fix; and a configurable "empty consumption" switch to discard messages when replay is possible.
