The article explains why a large product cache moved from an in‑heap Caffeine cache to an off‑heap OHC solution, detailing serialization trade‑offs, per‑dimension cache configuration, singleflight loading, and the performance gains observed during high‑traffic promotions.
This article details practical Spark performance optimizations—including RDD reuse, broadcast variables, Kryo serialization, parallelism settings, shuffle tuning, and JVM tweaks—while also presenting systematic solutions for data skew, shuffle failures, serialization errors, and YARN mode issues, all illustrated with concrete code snippets and examples.
This article explains five common serialization techniques—Java native serialization, Hessian, JSON, Google Protobuf, and Kryo—detailing their principles, performance characteristics, code examples, and when to choose each for Java backend systems.
Fury is a JIT‑compiled, multi‑language native serialization framework that fully implements JDK custom serialization, delivering 20‑200× speed improvements over JDK, Hessian, and Kryo, while preserving compatibility across Java, Python, Go, C++, and JavaScript, as demonstrated by detailed protocol analysis and performance benchmarks.
This article explains how to improve Spark and Spark Streaming performance by tuning serialization, broadcast variables, parallelism, batch intervals, memory usage, garbage collection, and Kafka integration, providing practical code examples and real‑world optimization results.
