The article provides a comprehensive overview of IBM's General Parallel File System (GPFS), detailing its physical and logical architecture, key components such as NSD and quorum mechanisms, scalability, load balancing, and fault‑tolerance features for parallel and serial applications.
The article examines why software architectures inevitably degrade over time in long‑running projects, analyzes the stages of decay, and presents practical strategies—including environment upgrades, phased and distributed builds, component isolation, and platform‑plus‑app designs—to keep systems maintainable and performant.
Will Larson’s insights on scalable system architecture emphasize linear horizontal scalability, redundancy, load balancing strategies, caching layers, CDN usage, offline processing with message queues, and platform layering, providing practical guidance for building robust, high‑capacity backend systems.
This article describes the design, architecture, and key features of a Linux‑based distributed file system that emphasizes horizontal scalability, data consistency, fault tolerance, and efficient handling of both small and large files without traditional metadata services.
Uber’s Schemaless is a highly available, horizontally scalable MySQL‑based key‑value store that abandons fixed schemas, supports JSON blobs, triggers, and global secondary indexes, and was created to meet five critical requirements for trip data storage, including linear expansion, write throughput, and reliable operations.
This article examines Uber's rapidly growing real‑time dispatch platform, detailing its geo‑spatial indexing, microservice architecture, fault‑tolerant design, and scaling techniques that enable millions of writes per second and high availability across thousands of nodes.
The article chronicles LinkedIn’s journey from a single‑server Leo monolith to a highly distributed, service‑oriented backend architecture, detailing the introduction of member graphs, read‑only replicas, caching layers, Kafka pipelines, Rest.li APIs, super‑blocks, and multi‑data‑center deployments to support billions of daily requests.
LinkedIn grew from a single monolithic Leo server handling all web requests to a complex ecosystem of over 750 independent services, employing graph databases, read replicas, caching layers, Kafka pipelines, Rest.li APIs, and multi‑data‑center deployments to support billions of daily queries.
This article explains the concept of microservices, contrasts it with monolithic architectures, outlines nine characteristic principles such as componentization, business‑capability‑oriented teams, product thinking, smart endpoints, decentralized governance, polyglot persistence, infrastructure automation, fault‑tolerance, and evolutionary design, and discusses the benefits, challenges, and practical considerations of adopting this style.
This article outlines Sogou's infrastructure evolution under rapid business iteration, detailing stages of compute and storage horizontal scaling, serviceization, and stream computing, while sharing the practices, principles, lessons learned, and reflections that guided the platform's architectural transformation.
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 details the design and implementation of 58’s real‑time messaging platform, covering its business background, overall architecture, protocol design for extensibility and debugging, and solutions for fast TCP reconnection to ensure stable, high‑throughput delivery for intra‑city freight services.
This article presents a pattern‑based approach to performance tuning in large‑scale backend services, describing common degradation anti‑patterns, a set of optimization patterns such as horizontal and vertical partitioning, runtime 3NF, data locality, and degradation strategies, and illustrates each with real Meituan case studies.
This article examines Ctrip's Hermes asynchronous messaging system, detailing its evolution from a simple Mongo‑backed queue to a broker‑centric, partitioned architecture with lease‑based cluster management, and shares practical techniques for building high‑performance, low‑latency message queues in large‑scale distributed environments.
This article provides a detailed overview of Stack Overflow's 2015 architecture, covering daily traffic growth, hardware upgrades, redundancy principles, DNS and ISP routing, HAProxy load balancing, IIS/ASP.NET web layer, Redis caching, WebSocket services, Elasticsearch search, SQL Server databases, and the open‑source tools that support the platform.
This talk explores how Meipai designed a scalable, reliable database architecture that balances SQL and NoSQL technologies to support rapid product iteration and high‑peak traffic, highlighting key considerations for storage service reliability and performance optimization.
The article outlines how Meipai’s rapidly growing platform built a backend system that guarantees high availability, handles massive traffic, supports evolving media formats, and adopts a multi‑language development model with clear principles for language selection and systematic integration.
This article explains how large‑scale elastic architecture designs achieve high availability and scalability by automatically allocating resources during traffic spikes such as Spring Festival ticket sales and Double‑11 shopping festivals, and outlines key techniques, trends, and disaster‑recovery strategies.
Load balancing distributes incoming traffic across multiple servers to enhance performance, ensure high availability, and enable horizontal scaling, with various implementations such as DNS, HTTP, IP, link‑layer, and hybrid methods, each employing algorithms like round‑robin, random, least connections, hash, and weighted distribution, plus hardware options.
This article presents a comprehensive e‑commerce website architecture case study, detailing functional and non‑functional requirements, initial single‑server design, capacity estimation, and progressive evolution toward a distributed, high‑availability architecture with clustering, caching, load balancing, service‑orientation, and database sharding.
This article explains how large‑scale internet applications can overcome database bottlenecks by using load‑balancing clusters, implementing read/write separation, and applying vertical or horizontal sharding, while outlining each method's principles, implementation steps, advantages, and drawbacks.
The article outlines how MySQL architectures evolve from a single-instance setup to vertical splitting, master‑slave replication, horizontal sharding, and cloud‑native SaaS solutions, discussing scalability strategies, sharding techniques, operational challenges, and automation for seamless expansion and high availability.
The article explains the core concepts of message queues and distributed systems, describes how queues act as intermediaries to decouple producers and consumers, outlines benefits such as faster response, reliability, and scalability, and details why a distributed architecture is needed for multi‑system collaboration and fault tolerance.
This article examines the critical issues of massive data handling, concurrency, file storage, relational design, indexing, distributed processing, AJAX usage, security, clustering, and OpenAPI trends that developers must address when architecting large, high‑interaction web sites.
This article explains how Weibo’s massive image‑hosting platform uses a cross‑IDC distributed object storage system, optimized upload/download pipelines, and a custom compression library to handle billions of images and extreme traffic spikes during events like the Chinese New Year.
This article explores the importance of database clustering in modern information systems, outlines the challenges of performance, availability, and scalability, and compares Share‑Disk and Share‑Nothing architectures along with their advantages, drawbacks, and real‑world implementations.
This article examines the technical challenges and solutions involved in designing, deploying, and evolving a cross‑border e‑commerce platform, covering system architecture, network setup, database scaling, service modularization, and real‑world pitfalls encountered during rapid growth.
This article chronicles LinkedIn's evolution from a monolithic Leo application to a massive micro‑service ecosystem, detailing the introduction of member graphs, read‑only replicas, caching layers, Kafka pipelines, Rest.li APIs, super‑blocks, and multi‑data‑center strategies that enable handling billions of requests daily.
This article examines how Tumblr handled rapid growth—processing 5 billion daily page views, 40 k requests per second, and terabytes of data—by evolving from a LAMP stack to a Scala‑based, Finagle‑driven distributed system with HBase, Redis, Kafka, and a cell architecture that supports massive real‑time dashboards.
This article explains the benefits and drawbacks of performance optimization, distinguishes between single‑application and structural optimization approaches, outlines common steps, tools, and techniques for each, and presents case studies illustrating architectural evolution for improved scalability and stability.
The article explains the concepts of vertical and horizontal database splitting, compares their technical difficulties, discusses the impact on joins and transactions, and proposes practical approaches such as careful SQL refactoring, ORM decoupling, and selective use of distributed‑transaction techniques to maintain system stability and scalability.
This article shares practical insights on achieving high availability by understanding goals, decomposing requirements, designing resilient architectures, ensuring operability, testing rigorously, and reducing release risk, illustrated through the multi‑stage evolution of a transaction system.
The article outlines the three core traits of internet architecture—agility, scalability, and openness—illustrates them with Alibaba’s supply‑chain and product examples, and explains how cloud‑based, distributed systems enable enterprises to become operators of their own core capabilities.
The article outlines the design requirements, sharding strategies, capacity planning, cost trade‑offs, and indexing techniques for building a highly available, scalable distributed storage solution for Weibo, focusing on storage‑layer decisions without involving caching or ID generation.
The article explains how large‑scale website architectures evolve from simple LAMP stacks to modern distributed, service‑oriented, virtualized systems, emphasizing the need for scalability, cost efficiency, middleware reuse, and continuous architectural transformation to meet growing user and development demands.
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.
This article shares practical design principles for building scalable read services, covering stateless architecture, data closed‑loop processing, multi‑layer caching strategies, concurrency optimization, degradation switches, rate limiting, traffic switching, and other operational best practices.
This article traces the development of game server architectures from early text‑based MUD systems through first‑generation HTTP‑based services for simple card and runner games, to the complex, distributed backend designs used by modern MMORPGs, highlighting key technologies, design choices, and scaling challenges.
The article chronicles Yirendai’s four‑stage backend evolution—detailing the bottlenecks, architectural redesigns, logging improvements, and service‑splitting strategies that transformed a single‑server system into a scalable, multi‑service platform.
This article details the architectural evolution of Kuaidi Dache from 2013‑2014, covering LBS bottlenecks, MongoDB scaling, long‑connection stability, distributed system refactoring, a wireless open platform, real‑time monitoring with Storm and HBase, and a data‑center built on sharding and synchronization.
This article explains how microservices emerged to address low development efficiency, technology incompatibility, and resource waste, and describes their benefits such as faster deployment, technology flexibility, optimal resource use, and improved stability, while also noting performance challenges.
This article explains why serving many images strains web servers, outlines the challenges of I/O bottlenecks, and presents a practical, scalable solution that uses distributed image servers, dynamic load balancing, random selection algorithms, and health monitoring to improve performance for medium‑size business sites.
This article outlines a comprehensive design for a high‑concurrency distributed backend system that handles tens of millions of page views, detailing group‑based data partitioning, master‑slave roles, consistency models, heartbeat services, and global coordination to achieve scalability and fault tolerance.
This article chronicles JD.com's POP platform evolution from a simple monolithic system in 2010 to a micro‑service‑based, highly scalable architecture by V3.0, highlighting the technical challenges, solutions, and the introduction of service‑oriented APIs, diversified data stores, and unified operational tools.
This article details the rapid traffic growth challenges faced by Kuaidi Dache from 2013‑2014 and presents representative architectural bottlenecks and the engineering solutions—including LBS optimization, long‑connection redesign, distributed refactoring, a wireless open platform, real‑time monitoring, and data layer transformation—that enabled stable, scalable, high‑performance ride‑hailing services.
The article details Xiaomi Push Service's XMPP‑based protocol, multi‑layer server architecture, client SDK design, scalability measures, security mechanisms, performance metrics during the 11.11 promotion, major refactorings, and practical insights for building high‑throughput push systems.
This article shares the architecture, design principles, challenges, and performance‑optimizing solutions behind a ten‑hundred‑thousand‑scale server monitoring system, covering data collection agents, distributed pipelines, real‑time alerts, high throughput, multi‑platform support, and practical lessons learned.
This article details Haier's e‑commerce platform architecture, explaining how a SOA approach built on Dubbo enables high scalability, a flexible product service layer with a DSL query language, and performance optimizations that handle massive traffic spikes during major shopping events.
This article explains how JD.com’s Order Fulfillment Center (OFC) was built, re‑engineered, and continuously optimized to handle massive order volumes during major sales events, covering its architecture, migration from .Net to Java, distributed task queues, flow control, and operational practices that ensure reliability and scalability.
This article examines the technical challenges of massive web sites—such as billions of users, extreme concurrency, and petabyte‑scale data—and explains how architectural evolution, including service separation, caching, clustering, load balancing, CDN, distributed storage, NoSQL, and micro‑service decomposition, enables scalable, highly available, and secure operations.
This article examines the inherent drawbacks of Redis Cluster—such as gossip overhead, upgrade challenges, client protocol complexities, and implementation gaps—and proposes a set of architectural enhancements, including proxy layers, dashboards, and agents, to improve scalability, reliability, and operational efficiency.
During a 2015 tech salon, Weibo platform architect Wei Xiangjun dissected the essence of system architecture, illustrating how Sina Weibo scales to billions of users through multi‑layered services, sharding, CDN, multi‑datacenter caching, orthogonal decomposition, and robust monitoring, offering practical insights for large‑scale backend design.
The article explains the internal and external drivers behind website architecture evolution, outlines the U‑N‑S‑C model, describes scaling steps such as service separation, caching, clustering, load balancing, database read/write splitting, CDN, and highlights the operational and organizational challenges of each phase.
This article summarizes Chen Ming’s 2015 ArchSummit talk on WeChat Moments, detailing its massive posting volume, core database tables, publishing and reading workflows, and the multi‑data‑center disaster‑recovery architecture that ensures seamless, loss‑less service worldwide.
This article examines the push and pull models used by micro‑blogging platforms such as Twitter and Sina Weibo, analyzes their scalability challenges, and proposes a time‑partitioned pull approach that reduces database load while maintaining fast feed retrieval for active users.
This article presents a pattern‑based approach to performance optimization, describing common degradation anti‑patterns and corresponding optimization patterns—such as horizontal and vertical partitioning, runtime 3NF, data locality, and degradation—to help engineers improve response time, throughput, and availability in large‑scale backend services.
From a single‑machine “all‑in‑one” setup to a distributed, service‑oriented architecture, 58.com’s journey illustrates how startups can evolve their backend systems—adopting LAMP, read/write splitting, vertical splitting, caching, Java migration, and automated services—to handle traffic growth from tens of thousands to over a billion visits.
This article chronicles the architectural journey of JD’s Dongdong instant‑messaging platform, detailing its early simple polling design, subsequent performance and scalability challenges, and the progressive shifts toward service‑oriented, micro‑service, and cloud‑native architectures that support massive user growth.
The article presents a pattern‑based catalog of performance‑optimization techniques—such as horizontal and vertical partitioning, constant‑variable separation, locality, and sandboxing—illustrated with Meituan case studies, explaining motivations, trade‑offs, and how each pattern reduces latency, boosts throughput, and enhances availability in high‑throughput backend systems.
Facing massive traffic spikes during events like Double‑11, Haier’s e‑commerce platform adopted a SOA architecture built on Dubbo, designed a flexible product service with a DSL‑driven query engine, and achieved high performance, linear scalability, and extensibility across diverse product categories.
The article chronicles a developer’s journey from a simple personal homepage to managing billion‑visit web platforms, sharing practical insights on architecture evolution, knowledge structures, design principles, infrastructure, and software engineering practices essential for building and operating large‑scale websites.
This article recounts Zhihu's evolution from a tiny Python‑Tornado service on a single Linode to a massive, highly available backend employing custom logging, event‑driven processing, page‑render optimizations, and a service‑oriented architecture that now supports over 100 million users.
This article explores the unique characteristics of e‑commerce search engines, their specialized architecture, core modules, data update processes, and practical solutions for bugs, high concurrency, caching, and cold‑start challenges, offering a comprehensive guide for building robust search systems.
This article examines 58.com’s BangBang system architecture, detailing its four‑layer backend design, evolution from a traditional IM to a mobile management platform, and the push technologies used to support massive O2O traffic and millions of concurrent users.
During Double 11, JD.com’s product detail page handled tens of billions of requests without downtime by introducing a unified service architecture that consolidates promotion, inventory, recommendation, and other functions behind a single Nginx+Lua entry point, employing multi‑level caching, service isolation, and flexible degradation mechanisms.
In this interview, Xiaomi R&D architect Ouyang Chen shares the design philosophy, evolution stages, core modules, pitfalls, and future challenges of the Xiaomi advertising platform, offering practical insights for building scalable backend systems.
This article examines Uber's rapid 38‑fold growth and the engineering choices behind its real‑time market platform, detailing the scheduling system, geographic indexing, microservices, Ringpop, TChannel, and strategies for scalability, availability, and fault tolerance.
This article traces the evolution of large‑scale website architectures—from single‑server setups to distributed services—detailing how separation of concerns, caching, load balancing, database sharding, CDN, reverse proxies, NoSQL, and service decomposition collectively improve performance, scalability, and reliability.
This article details Alibaba's mobile infrastructure evolution that powered a record‑breaking 2015 Double‑11 sale, covering client container design, network pipeline challenges, dynamic rendering, cross‑platform solutions like Weex, and the open‑source plans that enable massive scalability.
This article explains why microservices outperform monolithic architectures by detailing their design, deployment, benefits, and challenges, helping developers decide if the approach suits complex, evolving applications; it also compares traditional single‑code‑base models, discusses scaling strategies, and outlines practical considerations for implementation.
JD’s Order Fulfillment Center (OFC) evolved from a small data‑transfer team into a highly scalable, distributed architecture that handles massive order volumes during events like 618 and Double‑11, employing Java migration, service decomposition, flow control, and robust operations to ensure data consistency and rapid delivery.
This article explains how to design a high‑availability flash‑sale system by separating business and data layers, using Redis queues for pressure isolation, ensuring inventory consistency, handling transaction reconciliation, and addressing anti‑fraud measures to cope with massive concurrent traffic.
The article details how LinkedIn has massively scaled Kafka usage over several years, addressing quotas, a new ZooKeeper‑free consumer, reliability enhancements, security features, monitoring frameworks, fault testing, and ecosystem integrations to support its massive data‑driven operations.
This article outlines the typical progression of website architecture from a single‑server setup through service separation, caching, clustering, read/write splitting, CDN, reverse proxy, distributed storage, NoSQL, and business‑level service decomposition, providing a clear roadmap for scaling web systems.
The article describes how the limitations of an in‑memory NoSQL HOLD platform led to the creation of TDSQL, a distributed MySQL‑based SQL system featuring ZooKeeper‑coordinated Scheduler, Agent, and Gateway modules, automatic cross‑IDC disaster recovery, transparent horizontal scaling, strong synchronous replication, and future integration with container technologies.
The article chronicles how 58.com’s web architecture evolved from a tiny, single‑machine setup to a multi‑layer, Java‑based, highly available service‑oriented platform, detailing the technical decisions, scaling challenges, and automation practices adopted at each traffic milestone.
This article explains what system load capacity is, how to measure it using requests per second, identifies key factors such as bandwidth, hardware, OS and server configurations, and provides practical tuning steps for Linux, TCP, Nginx, Tomcat and databases to improve scalability.
This article details how 58.com’s Bangbang system evolved from a traditional IM platform to a massive, multi‑layered architecture handling over 40 billion daily requests, covering its four‑layer design, third‑party integration challenges, client lightweighting, and the push technologies that power its mobile O2O services.
This article examines the core challenges of massive web platforms—huge user bases, high concurrency, and massive data—and walks through a step‑by‑step evolution of architecture from a single‑server setup to distributed caching, clustering, read/write separation, CDN, and micro‑service strategies for achieving high availability, performance, and security.
This article chronicles the step‑by‑step evolution of Weibo's recommendation system architecture—from the early independent 1.0 stage through the layered 2.0 phase to the current platform‑centric 3.0 design—detailing the driving business and technical factors, architectural components, advantages, shortcomings, and key outcomes at each stage.
This article outlines fifteen common hidden requirement types—from detail tweaks and cross‑platform adaptations to scalability, security, and content‑operation needs—and offers practical strategies for developers and product teams to anticipate, design for flexibility, and reduce costly changes during development.
The article analyzes the inherent drawbacks of Redis Cluster—such as its decentralized P2P design, gossip overhead, upgrade difficulty, lack of hot‑cold data separation, client protocol challenges, and implementation limits—and proposes architectural enhancements like proxy, dashboard, and agent components to improve scalability, manageability, and performance.
The article chronicles Tencent's technological evolution across four stages—from massive user growth to flexible platforms—highlighting the strategic use of efficient architectures, dynamic operations, and cloud services that enabled it to support billions of users while maintaining performance and reliability.
This article details JD.com's transition from static to dynamic, high‑performance product detail page architecture, describing the multi‑layer system design, data heterogeneity, caching strategies, scaling techniques, operational challenges, and the solutions implemented to meet rapid business demands during massive traffic events.
An in‑depth look at 360’s long‑connection push platform built with Go, covering system architecture, key performance indicators, scalability challenges, optimization strategies, operational practices, and practical Q&A for building reliable, high‑throughput messaging services.
The article explains how distributed systems split a task into multiple subtasks across separate servers to reduce execution time, while clusters replicate the same service on many servers to increase throughput, highlighting their complementary roles in scalability and high availability.
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.
The article explains the historical background, core characteristics such as scalability, cost‑efficiency, fault tolerance, and elasticity, and outlines essential design principles for distributed systems, emphasizing low hardware requirements, horizontal scaling, avoidance of single points of failure, reduced communication overhead, and stateless services.
This article compares relational databases with distributed document-oriented NoSQL databases, explaining their scaling models, data modeling differences, and offering practical advice on when and how to adopt document databases such as SequoiaDB for flexible, high‑performance applications in the big‑data era.
This article describes how MySQL architectures evolve from a single instance to vertical splitting, master‑slave replication, horizontal sharding, and cloud‑based SaaS solutions to handle increasing traffic, data size, and operational requirements while maintaining scalability and availability.
This article summarizes publicly available information and personal experience to outline Google’s product portfolio, design principles, workload categories, and the distinction between its giant and medium‑sized data centers, providing a speculative view of the company’s overall architecture.
The article details how Qihoo 360 adopted MongoDB since 2011, scaling to over 100 applications, 1,500 instances and 20 billion daily queries, and shares their architectural choices, backup strategies, best‑practice recommendations, and advice for teams considering MongoDB in large‑scale, cloud‑native environments.
The article reveals how operating‑tool systems, often underestimated, demand high technical rigor to ensure automation success rates and absolute reliability for emergency actions, requiring sophisticated failure handling, capacity awareness, and scalable design—challenges comparable to core online services.
The article explains MySQL Cluster as a real‑time, ACID‑compliant, in‑memory distributed database, describes its node architecture, transparent sharding for linear scalability, various SQL/NoSQL access methods, and presents a benchmark achieving 200 million queries per second on 32 data nodes.
The article examines how Facebook’s massive data centers operate at scale, detailing server serviceability, automated fault‑diagnosis systems like CYBORG, staffing ratios, and the engineering practices that enable reliable service for billions of users worldwide.
NGINX achieves high performance and scalability by using a multi‑process architecture with a single master process, multiple single‑threaded non‑blocking worker processes, and dedicated cache loader/manager processes, limiting worker count per CPU core and handling many connections via an event‑driven state machine.
This article introduces MySQL Cluster, an in‑memory, real‑time, ACID‑compliant distributed database, explains its multi‑master architecture, data, application and management nodes, transparent partitioning, NoSQL APIs, and presents a benchmark achieving 200 million queries per second on up to 32 nodes.
This article explains how to achieve ultra‑high HTTP request rates by using load‑generating tools such as Apache Bench, httperf, JMeter and Tsung, discusses Linux socket limits, provides step‑by‑step installation and configuration examples for Tsung on CentOS, and shows how to run tests and interpret the results for building a high‑performance web cluster.
This article details how Dianping’s sub‑40‑person operations team structures its groups, designs a dual‑datacenter architecture, and creates comprehensive monitoring, automation, configuration, and analysis systems—including Zabbix, Cat, workflow, Button, and a custom radar platform—to achieve high‑availability, self‑service, and continuous improvement.
The article outlines how narrowing operational permissions, automating processes, and adopting data‑center operating systems enhance security, scalability, visibility, and fault handling while reducing manual effort and enabling rapid, low‑risk changes across large infrastructures.
