A real‑world incident at JD Tech shows that a single 1.5 MB Redis cache key caused a full‑site outage during a high‑traffic event, and the article explains the underlying cache‑breakdown and bandwidth traps, then details three emergency mitigations and long‑term preventive practices.
This article explains InfiniFS's cache organization for directory metadata, its optimistic cache usage, and the lazy invalidation mechanism that broadcasts rename updates to a few metadata servers, enabling scalable and efficient metadata services in large‑scale distributed file systems.
The article analyzes a Meituan interview question about caching 10 million MySQL rows when Redis can store only 200 thousand rows, presenting a five‑step three‑tier cache architecture that separates cold and hot data, selects appropriate eviction policies, uses proactive hotspot detection, adds multi‑level defense, and employs both scheduled and real‑time pre‑heating, with performance numbers showing an 85% hit rate and 100 ms latency.
This article analyzes the high‑concurrency contract‑query workload of JD Logistics, examines call sources, query dimensions and volume patterns, and proposes a three‑layer protection architecture—Bloom filter, Redis cache, and database fallback—along with strategies for cache penetration, breakdown, and avalanche mitigation.
This article explains the fundamentals, characteristics, and use‑cases of distributed caching, compares Memcached and Redis, and provides a step‑by‑step Spring Boot implementation with code samples, testing guidance, and solutions to common cache pitfalls such as hot keys, penetration, and avalanche.
This article explores common cache challenges such as centralized expiration, cache penetration, avalanche, hot keys, large keys, data consistency, and concurrent pre‑warming, offering practical design patterns and mitigation strategies to build robust, high‑performance backend systems.
This article explains how to design Redis cache for relational list data such as user timelines or news feeds, covering fixed-length caching, consistency, resource utilization, using ZSET structures, handling additions, deletions, queries, and strategies like preloading, retry mechanisms, and asynchronous rebuilding.
This article examines the essential concepts, common pitfalls such as cache stampede, penetration, avalanche, hot keys, large keys, consistency, and concurrent pre‑heating, and presents practical design patterns and mitigation techniques for building a robust Redis cache architecture that can handle billion‑scale traffic.
This article provides a comprehensive overview of caching in large‑scale distributed systems, covering cache fundamentals, classification, major implementations such as CDN, reverse‑proxy, local, and distributed caches, detailed analyses of Memcached and Redis, common design challenges, and real‑world industry solutions.
This article explains the evolution from memcache to Redis, outlines seven classic cache problems such as centralized expiration, cache penetration, avalanche, hot keys, large keys, data consistency, and concurrent pre‑heating, and provides practical mitigation strategies for each scenario.
This article details the architectural design, technical challenges, and iterative evolution of Zhihu's massive read‑service, covering early Bloom‑filter solutions, the RBase cache‑through system, high‑availability and performance optimizations, and the final migration to TiDB for cloud‑native scalability.
This article explores the fundamentals and practical details of server‑side caching in distributed web systems, covering local vs. distributed caches, Redis instance planning, key‑value modeling, persistence, eviction policies, and CRUD operations with performance‑focused recommendations.
This comprehensive guide explains cache fundamentals, covering CDN, reverse‑proxy, distributed, and local caches, their design considerations, common pitfalls like consistency, high availability, cache avalanche and penetration, and provides practical architecture examples and mitigation techniques for robust high‑traffic systems.
ScyllaDB, a Cassandra‑compatible NoSQL database, achieves over tenfold performance improvements through a thread‑per‑core design, asynchronous I/O, custom caching, self‑tuning schedulers, a user‑space TCP/IP stack, and LLVM‑JIT query execution, making it a compelling study for high‑performance database engineering.
