This article examines Youku’s large‑scale system architecture, detailing its server fleet, LAMP‑based front‑end framework, evolving MySQL database designs from master‑slave replication to vertical partitioning and sharding, and its caching strategies and CDN deployment that together support over 1 billion daily users.
Meituan’s group‑buy order system was refactored from a monolithic, single‑point‑failure design into a sharded, micro‑service architecture with connection‑pool middleware, dual‑cluster databases, and robust messaging, delivering stable, high‑performance handling of millions of daily orders even during peak events.
This article walks through the typical evolution of a growing website—from separating web and database servers to adding caching layers, load balancing, sharding, and distributed systems—explaining why each change is needed and what knowledge areas it touches.
The article details the technical challenges encountered during the service‑oriented decomposition of Qunar's payment platform, covering Dubbo and HTTP service conventions, database sharding and read/write separation, asynchronous processing, multi‑system management, and comprehensive monitoring and alerting solutions.
Facing a 200 GB order table that strained performance during flash‑sale spikes, Meituan first vertically split the schema, then horizontally sharded orders by user and merchant IDs using a 32 × 32 hash scheme with embedded timestamp‑based IDs, enabling scalable clusters, flexible migration stages, and simplified queries while preserving transaction integrity.
The article examines the technical challenges encountered during the service‑oriented decomposition of Qunar's payment platform—including development efficiency, interface conventions, concurrency, security, monitoring, database sharding, read‑write separation, and asynchronous processing—and presents concrete solutions and best‑practice recommendations.
This article explains the principles, architecture, and performance of Sharding-JDBC, covering sharding scenarios, rule configuration, JDBC rewriting, SQL parsing and routing, result merging, and future roadmap, while comparing it with other open‑source sharding solutions.
The article outlines how mature large‑site architectures evolve from a single‑server setup to multi‑layered systems featuring separated application, data and file tiers, caching, load‑balancing, database sharding, CDN, distributed file systems, NoSQL/search, business‑level service decomposition and distributed service frameworks.
This article details JD's One‑Yuan Grab Treasure platform redesign, covering business growth drivers, database sharding estimation, hash‑plus‑range routing implementation, Elasticsearch aggregation, Canal‑based sync, historical data migration, and downgrade mechanisms to ensure high‑throughput, reliable order processing during massive sales events.
The article outlines how a large website’s architecture progresses from a single‑server setup through service separation, caching, clustering, read/write splitting, CDN, distributed databases, NoSQL, business decomposition and cloud platforms, while highlighting common pitfalls.
This article traces the step‑by‑step evolution of large‑scale website architectures—from single‑server setups to distributed services—highlighting key techniques such as server clustering, caching, load balancing, database sharding, CDN usage, and the adoption of NoSQL and micro‑service frameworks.
This article explains how a rapidly growing order database exceeding 200 GB can be scaled using vertical and horizontal partitioning, various sharding strategies, custom ID generation, migration phases, and practical considerations for transactions and complex queries.
The article outlines a comprehensive database sharding solution for a massive order system, describing vertical and horizontal partitioning, various sharding strategies, a custom ID scheme, transaction handling, complex query constraints, and a three‑stage data migration plan to achieve long‑term scalability.
The article explains how to build a highly scalable internet application by adopting stateless session handling, effective caching, service decomposition with remote call frameworks, database sharding, asynchronous messaging, unstructured data storage, comprehensive monitoring, and unified configuration management.
Taobao’s massive B2C platform achieves high scalability and reliability by employing stateless session management with multi‑value cookies, leveraging Tair caching, splitting services via HSF, partitioning databases with TDDL, adopting asynchronous messaging, using distributed file storage, and implementing comprehensive monitoring and unified configuration management.
This article explains two practical horizontal sharding techniques for large databases—using an MD5 hash of the user UID to distribute rows across 256 tables, and right‑shifting the UID to create scalable table ranges—while discussing their capacity limits and extensibility.
This article explains two practical horizontal sharding techniques for large‑scale projects—MD5‑based hashing and bit‑shifting of user IDs—detailing their PHP implementations, scalability limits, and how to extend table counts as data grows.
The article examines the increasing storage bottleneck in large-scale web systems, explains vertical and horizontal database sharding, discusses the technical difficulties they introduce—including join rewrites, transaction handling, and ORM integration—and proposes practical approaches to minimize service impact and manage distributed transactions.
The article chronicles Xiaomi Web's architectural journey from a simple three‑engineer monolith in 2011 through systematic service decomposition, asynchronous messaging, database sharding with Cobar, cloud‑native scaling, advanced caching, virtual inventory allocation, and sophisticated monitoring, illustrating practical lessons for building high‑performance e‑commerce platforms.
Horizontal sharding splits a large table into multiple databases using hash‑modulo routing, improving performance and scalability; the article explains the basic algorithm, shows example calculations for a 4‑shard setup, and demonstrates how to expand shards from four to eight.
Facing explosive traffic growth in 2014, Didi’s ride‑hailing platform tackled critical challenges by redesigning its LBS architecture, replacing unstable long‑connection services with an AIO‑based framework, partitioning databases, adopting Dubbo and RocketMQ for distributed processing, and building a real‑time monitoring and data center using Storm, HBase, and custom SQL‑to‑HBase translation.
This article analyzes the complete flash‑sale workflow, identifies its unique characteristics and high‑concurrency technical challenges, and presents a layered architecture—including frontend static pages, request interception, queue design, database sharding, caching, optimistic locking, and anti‑cheating measures—to ensure reliability, scalability and data safety.
The article describes how Alibaba scaled its massive data platform by horizontally sharding Oracle tables to MySQL, employing multi‑data‑center synchronization tools such as Erosa, Eromanga, and Otter, and designing multi‑level caching and proxy solutions like Cobar to ensure high availability and performance.
This article analyzes YouPai.com's evolution from a simple master‑slave setup to horizontal sharding using a mapping table, detailing migration steps, access flow, and challenges such as cross‑database joins, consistency, and ID generation.
The article details how Alibaba tackled explosive data growth by migrating from a single Oracle instance to a multi‑source architecture using horizontal and vertical sharding, MySQL clusters, KV stores, and proprietary tools such as Erosa, Otter, and Cobar to achieve high availability, consistency, and performance across multiple IDC sites.
This article chronicles the evolution of Sina's news comment platform from a single‑server C++/MySQL prototype to a multi‑node, asynchronous, Python‑based architecture, highlighting the caching, sharding, and queueing techniques that enabled the system to survive massive traffic spikes without downtime.
Vertical sharding splits large tables across multiple servers by whole tables, offering simple partitioning but limited scalability, while horizontal sharding distributes rows of high‑traffic tables based on key rules, improving performance and scalability yet introducing complex partition logic and maintenance challenges.
This article explains why many companies over‑invest in Hadoop, outlines how to evaluate data size, growth, and relevance, and presents practical alternatives such as archiving, data sampling, database sharding, and hiring business‑savvy analysts before committing to a Hadoop deployment.
The article traces web architecture’s evolution from simple single‑server Java/JSP sites through memory caching, load‑balanced multi‑machine logic, read/write separation, master‑slave replication or message‑queue syncing, horizontal/vertical sharding, and finally multi‑server web‑server deployments behind load balancers to achieve scalability and reliability.
