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 reviews the complete C‑end product system built for billions of users, covering demand analysis, Java migration, high‑concurrency read service design, launch safeguards, and a detailed list of proven caching and traffic‑isolation techniques validated in production.
This article explains how using an off‑heap local cache (OHC) can dramatically lower GC pauses and cut interface latency by up to tenfold, covering the underlying principles, configuration, custom serializers, performance testing, monitoring metrics, and practical optimization recommendations for high‑traffic Java backend services.
This article details the performance tuning of Vivo's Java-based push recommendation service, covering background challenges, measurement metrics, hotspot code and JVM GC optimizations, custom split utilities, map key improvements, off‑heap caching, and the resulting throughput and latency gains.
The Vivo Push recommendation service was dramatically accelerated by replacing regex‑based String.split with a custom utility, converting massive String‑key maps to long keys, and moving large caches off‑heap, which together cut latency by about 32%, boosted throughput roughly 45% and nearly doubled overall performance.
This article explains the principles and practical implementation of the Java off‑heap cache framework OHC, detailing its architecture, memory allocation, serialization, configuration, and real‑world performance results within the MaFengWo recommendation engine, illustrating how it reduces latency and improves cache hit rates.
