The article outlines five common cache design problems—penetration, concurrency, avalanche, hot‑data management, and business‑cache coupling—illustrates each with a diagram, and provides concrete Redis‑based solutions such as placeholder values, setNX locks, random TTLs, sorted‑set queues, and a Binlog‑Canal‑MQ pipeline.
This article details JD Advertising's journey from a 1 PB Apache Doris data lake to a multi‑level hot‑cold tiering architecture, describing two tiering strategies, the performance and schema‑change challenges faced during the upgrade to Doris 2.0, and the optimizations that reduced storage costs by about 87% while boosting query throughput.
JD-Hotkey is a backend framework that detects and pushes hot keys at millisecond precision, handling billions of keys daily with up to 370,000 detection tasks per second and 600,000 push operations per second, dramatically reducing data‑layer pressure and improving application performance across various caching and rate‑limiting scenarios.
The article explains the principle of hot‑cold data separation, when it should be applied, how to distinguish hot versus cold data, and three practical implementation approaches—code modification, binlog listening, and scheduled scanning—to improve database performance and maintain consistency.
The article explains the hot‑cold data separation pattern, describing its purpose, when to use it, how to distinguish hot versus cold data, and three practical implementation approaches—code modification, binlog listening, and scheduled scanning—to improve performance and maintain data consistency in large‑scale systems.
This article explains the concept of hot data, categorizes its types, discusses caching trade‑offs, introduces LFU, LRU and TinyLFU algorithms, describes the HotCaffeine system architecture and features, and presents performance improvements and practical deployment considerations for Java backend services.
This article explains how to configure Elasticsearch nodes for hot and cold data, assign special node attributes, adjust index templates, and use API calls to migrate data, demonstrating significant query speed improvements through real‑world performance tests.
This article examines the evolution of database architectures, introduces the hot‑data and cold‑data concepts for a weather‑data service, and details a practical implementation using MySQL, TiDB, and PolarDB with synchronization, expiration procedures, and suggestions for future improvements.
The article explains the concept of hot and cold data, why separating them reduces cost, and presents heterogeneous and homogeneous architectural solutions—including Elasticsearch, HBase, AWS S3, and cloud‑based UltraWarm—illustrated with network‑behavior and e‑commerce order system case studies.
