This article explains JD's design of a billion‑scale vehicle‑part matching cache, detailing the data model, bitmap storage structure, SDS pre‑allocation, key‑value compression, cluster routing, and Lua‑based high‑performance search interface that achieves sub‑13 ms latency.
This article explains how Redis uses Linux epoll and a single‑threaded event‑driven architecture to handle tens of thousands of connections efficiently, detailing the initialization of the server, the event‑loop mechanics, and the processing of client commands from accept to reply.
This article provides a comprehensive guide to Netty, covering its definition, advantages, key features, application scenarios, performance characteristics, architectural components, threading models, code examples, handling of packet framing, zero‑copy techniques, connection management, object pooling, and supported serialization protocols.
This article explains how RocketMQ tackles registration latency, synchronous coupling, and traffic spikes by introducing an intermediate queue, designing a persistent high‑availability broker, leveraging Linux page cache and memory‑mapped files, and employing topics, tags, and sharding to enable efficient asynchronous processing and scalable consumption.
The article explains how RocketMQ tackles growing business traffic by introducing an asynchronous broker layer, using commitlog and consumeQueue files, page‑cache, mmap, topic/tag routing, and a nameserver to achieve high‑throughput, low‑latency, and highly available message delivery.
MIPS Tech has shifted from its legacy CPU architecture to RISC‑V, unveiling the high‑performance, scalable eVocore P8700 and power‑efficient I8500 multi‑processor IP cores aimed at high‑performance, real‑time applications such as networking, data centers, automotive, and edge computing.
The XQUIC project, Alibaba’s high‑performance open‑source QUIC transport library that powers mobile Taobao and large‑scale gateway services with up to 20 % latency reduction, offers a standards‑compliant stack (QUIC‑Transport, QUIC‑TLS, HTTP/3, qlog) and multi‑path XLINK, and invites students in the Open Source Summer program to extend the code, improve diagnostics, and earn mentorship and rewards.
This article provides a comprehensive introduction to Redis, covering its high‑performance in‑memory architecture, core features, supported data types, caching usage in Spring Boot, common cache pitfalls and mitigation, persistence options, master‑slave replication, and Sentinel high‑availability mechanisms.
This article introduces the high‑performance Disruptor library, explains its background, core concepts such as RingBuffer, Sequence, Sequencer, and WaitStrategy, and provides a complete Spring Boot‑based Java demo with code snippets illustrating producers, consumers, and configuration.
This document describes the background, problems, requirements, analysis, and detailed design of a high‑performance, highly available, and scalable log‑analysis pipeline for MySQL that leverages Kafka, ClickHouse, MySQL, and custom services to aggregate, enrich, and visualize massive query logs.
This article, the second in a Go high‑performance series, details concurrency optimizations including lock‑free data structures versus locked lists, sharding and RWMutex to cut lock contention, controlling goroutine creation with pooling, using sync.Once for cheap one‑time initialization, and employing sync.Cond for efficient goroutine notification.
This article introduces a collection of high‑performance open‑source components from Tencent—including a key‑value store, API framework, networking library, AI inference engine, Android hot‑patch system, coroutine library, plugin framework, cross‑platform UI framework, mini‑program tool, and an IoT real‑time OS—detailing their core principles, features, and GitHub links.
This article explains the limitations of traditional blocking I/O, introduces Java NIO with selectors, compares the two models using a kindergarten analogy, critiques raw JDK NIO, and demonstrates how Netty simplifies and optimizes network development with concise server and client code examples while promoting a comprehensive Netty learning book.
XQUIC is Alibaba's open‑source, standards‑compliant QUIC implementation that delivers lightweight binaries, high‑performance transport, and cross‑platform support, enabling faster, more reliable data transfer for mobile apps and cloud services while fostering community collaboration through open source.
This article explains how to replace Java's LinkedBlockingQueue with the high‑throughput LMAX Disruptor library, covering dependency setup, event definition, Disruptor creation, producer and consumer implementations, handler configuration, startup/shutdown procedures, and provides both Java and Groovy demo code.
DPVS v1.9.0, the latest open‑source four‑layer network load balancer from iQIYI, now fully supports DPDK‑20.11 LTS, offers streamlined compilation, universal flow management, dynamic mbuf fields, enhanced per‑core scheduling and UDP health checks, and has been running stably in production for three months.
The article presents Maxwell, a C/assembly‑based high‑performance accounting database designed for Alipay, detailing its architecture, hotspot account challenges, performance optimizations such as Raft and a custom storage engine, and real‑world results showing thousand‑fold throughput gains and robust operation during massive promotional events.
The article thoroughly examines Redis’s high‑performance architecture, detailing the evolution from blocking to non‑blocking I/O, the Reactor pattern’s single‑ and multi‑reactor models, Redis’s I/O multiplexing thread design, and how its hybrid single‑thread core with auxiliary I/O threads mitigates bottlenecks under heavy traffic.
The article examines the evolution of Tencent Cloud’s F-Stack—a full user‑space networking framework built on DPDK, FreeBSD, and micro‑threading—detailing its architecture, supported versions, POSIX‑compatible API, real‑world deployments such as HttpDNS and DNSPod, performance gains over kernel stacks, and the roadmap for future enhancements.
This article summarizes the design, architecture, and performance optimizations of KGraph, Kuaishou’s high‑throughput distributed graph database, covering its graph model, storage engine built on PMem, the KRPC network framework, scalability, key challenges, and future directions.
This article details the evolution of Vivo's e‑commerce coupon platform, covering its business drivers, architectural redesign, sharding strategy, multi‑mode issuance, high‑performance safeguards, and a step‑by‑step zero‑downtime migration to a standalone service.
The 2021 Trusted Cloud Conference in Beijing unveiled ten key cloud computing trends—including cloud native, high‑performance computing, chaos engineering, hybrid cloud, edge computing, zero‑trust, optimized governance, digital government, low‑carbon cloud, and enterprise digital transformation—highlighting the strategic direction of China's cloud industry.
This article details Meitu's adoption of the Presto ad‑hoc ROLAP engine, comparing it with Hive on Spark and Impala, describing enhancements for coordinator high‑availability, and explaining a cross‑cluster scheduling strategy that leverages idle Presto resources to improve overall big‑data workload efficiency.
This whitepaper explains the concept, history, core attributes, typical industry use cases, major products, and selection guidelines for in‑memory databases, highlighting their performance advantages, technical challenges, and future trends driven by evolving memory technologies.
Tinyid is a Java‑based distributed ID generator that uses a database segment algorithm, supports multiple master databases, offers both HTTP and Java client interfaces, and provides high throughput and availability for billions of IDs daily.
The article explains how agile teams can thrive amid VUCA conditions by aligning clear goals, fostering intrinsic motivation, and developing diverse capabilities, offering practical management principles and four key practices—goal transparency, capability building, self‑organization, and motivation extraction—to achieve high‑performance outcomes.
This article explains the fundamentals of IO events and IO multiplexing, compares thread‑based and event‑driven models, details the Reactor pattern variants, and discusses synchronous versus asynchronous IO, providing a practical guide for building high‑performance network frameworks in Linux.
At the 2021 Trusted Cloud Conference, the China Academy of Information and Communications Technology released the top ten cloud computing keywords—cloud native, high performance, chaos engineering, hybrid cloud, edge computing, zero trust, optimized governance, digital government, low‑carbon cloud, and enterprise digital transformation—highlighting their significance and future development directions.
This article explains Apache RocketMQ's high‑performance distributed messaging architecture, detailing its core components—Producer, Broker, Consumer, and Namesrv—along with routing metadata, message ordering, priority handling, parallelism, and essential concepts such as topics, groups, tags, and reliability mechanisms.
This article examines the challenges of cloud‑native storage, compares common solutions like Rook‑Ceph and OpenEBS, and explains how NeonIO’s containerized architecture, high performance, ease of use, and strong high‑availability features make it a compelling choice for Kubernetes workloads.
In a May 15 meetup, ByteDance engineer Ma Haoxiang discussed his background, the culture at ByteDance, recommended resources, and detailed how distributed databases differ from traditional relational databases, highlighting their massive capacity, low cost, high performance, and the specific performance and disaster‑recovery challenges faced during Douyin’s Spring Festival red‑envelope activity.
This article explains how Netty, an asynchronous event‑driven Java NIO framework, simplifies building high‑performance, maintainable network servers and clients by handling I/O, threading, and protocol concerns, and it covers its thread models, zero‑copy techniques, common interview questions, and real‑world use cases such as Dubbo and RocketMQ.
This article explains Nginx's high performance by detailing its master‑worker process architecture, asynchronous event‑driven I/O model, modular design, and how it handles HTTP connections, requests, and concurrency compared to traditional servers like Apache.
This article explains the fundamentals of IO events and multiplexing, compares thread‑based and event‑driven architectures, details the Reactor pattern variants, and clarifies synchronous versus asynchronous IO, providing a practical guide for building high‑performance network frameworks in Linux environments.
This article explains how to design a flash‑sale system that handles massive concurrent requests by focusing on high performance through dynamic‑static separation, hotspot optimization, and code‑level tuning, while ensuring strong consistency for inventory and maintaining high availability via traffic shaping, fault‑tolerance, and operational best practices.
This article explains the core concepts of high‑performance Linux network programming, covering IO events, IO multiplexing, thread and event‑driven architectures, the Reactor pattern and its variants, as well as synchronous versus asynchronous IO, providing clear examples and practical guidance for building efficient server frameworks.
This guide explains IO events and multiplexing, compares thread‑based and event‑driven models, details the Reactor pattern variants, and clarifies synchronous vs asynchronous IO, providing a practical roadmap for building high‑performance Linux network frameworks.
This article details the design, implementation, and performance testing of a Go‑based backend that simulates a Spring Festival red‑packet system capable of handling up to 6 × 10⁴ QPS and one million concurrent connections, demonstrating how to achieve 10 billion request processing in a controlled environment.
CURVE is NetEase's open‑source, high‑performance, highly available distributed storage platform that supports block, object, and cloud‑native database use cases, emphasizing simple solutions, hardware‑software integration, and extensive open‑source component reuse.
This article introduces JD Cloud’s self‑built time‑series database HoraeDB, explaining its core concepts, typical use cases, architectural layers, high‑performance features, down‑sampling strategies, compression techniques, and stability measures for handling massive, 24‑hour monitoring data at scale.
Tinyid is a Java‑based distributed ID generation system that extends the leaf‑segment algorithm with multi‑DB support, offering both HTTP and Java client interfaces, high performance through local generation, and robust availability via dual segment caching and configurable database redundancy.
This guide explains Linux socket programming concepts such as non‑blocking I/O and epoll, introduces the OpenResty web platform built on Nginx and Lua, walks through environment setup, configuration, Lua scripting, and common troubleshooting steps for building high‑concurrency backend services.
This article explains how a friend loaded over 2 billion rows into MySQL using XeLabs TokuDB, detailing configuration, bulk‑load commands, benchmark results of roughly 574 k rows per second, and comparisons with InnoDB performance on a cloud server.
NetEase Shufan has open‑sourced Curve, a high‑performance distributed storage system that claims 1.84× the IOPS of Ceph, offering low latency, high availability, and autonomous operation for cloud‑native workloads.
This article introduces ip2region, an offline IP‑address location library offering 99.9% accuracy, sub‑millisecond query speed, tiny database size, multiple search algorithms, and bindings for many languages, along with installation, testing, concurrency tips, and database generation instructions.
This article explains Nginx's multi‑process, event‑driven design, its master‑worker model, modular architecture, I/O multiplexing mechanisms such as epoll, and how these factors together enable the server to achieve high performance and support massive concurrent connections.
This article explains Nginx’s multi‑process architecture, event‑driven model, module system, HTTP connection workflow, and I/O multiplexing mechanisms such as select, poll, and epoll, illustrating why the server achieves high performance and scalability.
This article explains Nginx's high performance by detailing its multi‑process architecture, asynchronous event‑driven I/O, request handling flow, module types, comparison with Apache, maximum connection calculations, HTTP request/response structure, and the differences among select, poll, and epoll.
Netty, a Java‑based asynchronous event‑driven framework created in 2008, offers a unified API for multiple transports, a powerful reactor thread model, built‑in codecs, extensive protocol support, high throughput, low latency, robust error handling, and a vibrant ecosystem adopted by major tech companies and open‑source projects.
This article explores the fundamentals of high‑performance I/O design in Java, comparing BIO and NIO throughput, analyzing thread‑based and event‑driven architectures, and detailing the Reactor pattern with code examples and performance considerations for scalable web services.
This article explores the design of a high‑performance short‑link system, covering its benefits, redirection mechanics, hash‑based and auto‑increment generation methods, database schema, Bloom filter optimization, and a scalable OpenResty‑based architecture for handling massive traffic.
OPPO's ESA DataFlow is a self‑developed high‑performance data‑flow framework that processes over a trillion events per day, offering flexible routing, scalable sources and sinks, persistent mmap‑based channels, built‑in monitoring, and easy extensibility for diverse data‑collection scenarios.
This article explains why Nginx delivers high performance and concurrency, covering its multi‑process architecture, master‑worker model, event‑driven I/O (select, poll, epoll), module system, HTTP request handling, and comparison with Apache, while providing practical insights into configuration and scalability.
This article provides a comprehensive introduction to Netty, covering its core concepts, architecture, threading model, zero‑copy mechanisms, serialization options, comparison with JDK NIO and Tomcat, and practical usage scenarios for building high‑performance asynchronous network applications in Java.
This article details ByteDance's (今日头条) migration to RocketMQ, explaining the business drivers, the reasons for choosing RocketMQ, the proxy‑based architecture for producers and consumers, performance gains, operational challenges, and the disaster‑recovery designs implemented to ensure high availability in a micro‑service environment.
This talk presents Alibaba's intelligent marketing platform, detailing its three ad systems, the massive data and low‑latency requirements, and the architectural solutions—including cold‑hot data separation, OLDB storage, full‑link tracing, elastic compute, and platform‑level componentization—that achieve high performance, high availability, and cost‑effective operation.
This article analyzes the core challenges of building flash‑sale (秒杀) systems—high concurrency reads and writes, strict consistency, and ultra‑high availability—and presents a layered architectural approach covering dynamic/static separation, hotspot optimization, database tuning, and comprehensive high‑availability strategies.
QNAT is a high‑performance, open‑source x86 NAT solution developed by 360 OPS, featuring flexible address‑pool selection, session limits, traffic pass‑through, detailed session logging, and a CLI management interface, and is deployed in many IDC and office network egress points.
This article explores the architectural design of a flash‑sale (秒杀) system, covering high‑performance techniques, consistency guarantees, high‑availability strategies, hotspot optimization, and practical implementation details such as static‑dynamic separation, caching, data integration, and database tuning.
Plato, the newly open‑sourced high‑performance graph computing framework from Tencent’s TGraph project, delivers industry‑leading speed and memory efficiency for billion‑node social network graphs, achieving minute‑level processing with as few as ten servers, and supports a wide range of graph algorithms and learning tasks.
This article provides a comprehensive technical guide to DPDK and SPDK, covering their background, architecture, user‑space I/O bypass mechanisms, core performance optimizations such as HugePages, SIMD and cache management, and practical ecosystem recommendations for building high‑throughput network and storage services.
FeatureKV is a high‑performance, scalable key‑value storage system built on WeChat’s internal file system and metadata service, designed to handle billion‑scale read and write demands, support batch offline writes, provide version management, and achieve low‑latency online reads for services like Look‑at‑Look, ads, and payments.
DPVS is an open‑source, DPDK‑based user‑space load balancer that achieves line‑rate throughput by assigning each worker to a dedicated CPU core and NIC queue, provides lock‑free per‑CPU data structures and multiple forwarding modes including NAT64, and simplifies maintenance while supporting IPv4/IPv6 in large‑scale production deployments.
This article explains the limitations of single‑connection request handling, demonstrates the performance benefits of multiplexing, and details a .NET Core test that reaches a million requests per second with low latency using Protobuf messages and BeetleX integration.
Netty is a high‑performance, asynchronous, event‑driven NIO framework that supports TCP, UDP and file transfer, offering robust, customizable, and scalable networking solutions widely adopted across internet, big‑data, gaming and telecom industries, with detailed analysis of its architecture, I/O model, threading, serialization, reliability, traffic shaping, graceful shutdown, and security features.
This article provides a curated list of nineteen books that cover high‑concurrency programming techniques, algorithms, and practical implementations across Java, Go, C++, Python, Erlang, and other languages, serving as a valuable resource for developers seeking deeper knowledge in concurrent system design.
The article presents ContainerLB, a software‑defined L4 load‑balancing service built on Intel DPDK and X86 servers for JD.com, detailing its architecture, FULLNAT mode, high‑availability design, performance optimizations, and test results that demonstrate near‑line‑rate packet processing and scalable deployment in a container‑native data‑center.
This article presents a comprehensive technical sharing on Golang high‑performance engineering, covering architecture design, profiling tools, concurrency optimizations, distributed rate limiting, timer mechanisms, GC tuning, and deployment strategies, supplemented with PPT links and extensive code excerpts for practical reference.
This article presents the design, DPDK‑based implementation, performance optimizations, and test results of KDNS—a C‑language DNS server that replaces ContainerDNS's Go server, achieving near line‑rate processing, high availability, and significant latency improvements over traditional bind9 solutions.
This article summarizes Yu Shiwei’s “Building High‑Performance Teams”, outlining team concepts, types, three essential elements, conflict handling, organizational dynamics, and practical incentives to help managers create cohesive, autonomous, and results‑driven teams.
This article explains how the Disruptor framework achieves ultra‑low latency and high throughput in Java by using a ring‑buffer based producer/consumer model, detailing its architecture, sequencing algorithm, wait strategies, and practical code examples for both publishing and processing events.
X‑Paxos is Alibaba’s high‑performance, independently designed Paxos library that extends the classic consensus algorithm with multi‑threaded architecture, pluggable logging, adaptive batching and pipelining, and flexible node roles, delivering strong consistency, high availability, and low latency for global distributed databases and services.
This article examines 360’s explosive KV storage demand and compares three solutions—Redis, Redis Cluster, and Pika—detailing their features, performance, persistence options, eviction policies, architectures, migration tools, and ideal usage scenarios for high‑traffic, large‑scale applications.
This article explains how Alibaba Cloud designed a high‑performance local storage solution—leveraging Virtio‑blk, SPDK, and NVMe SSDs—to meet the massive peak pressure, ultra‑low latency, and high IOPS demands of the Alipay red‑packet business, and how the technology is now available in the public cloud.
This article details Alibaba Cloud's high‑performance local storage solution for the massive Alipay Red Packet service, covering business requirements, existing block storage limitations, architectural redesign, key components like Virtio‑blk, SPDK, and NVMe SSD, and performance benchmarks demonstrating dramatically lower latency and higher IOPS.
This article introduces Vesta, a Java‑based multi‑scenario distributed ID generator, covering its design goals such as global uniqueness, rough ordering, reversibility and manufacturability, its architecture, publishing modes, performance benchmarks, and step‑by‑step deployment instructions for various deployment scenarios.
This article reviews the 2016 Dyn DDoS incident, explains why DDoS attacks remain a critical security challenge, and details Alibaba's internally built high‑performance DDoS mitigation solution AliGuard, covering its development timeline, multi‑user architecture, scalability from 40 Gbps to over 400 Gbps, and deployment models.
Taobao’s Open Platform sustains hundreds of billions of daily API calls and messages by employing a pipeline‑based high‑performance API gateway with multi‑level caching, asynchronous processing, granular traffic control, a highly reliable push‑pull messaging system, and a zero‑loss data‑sync service that dynamically balances resources during massive traffic spikes.
Meituan’s self‑developed MGW replaces LVS with a DPDK‑based four‑layer load balancer that handles tens of gigabits per second and millions of concurrent connections by eliminating interrupts and locks, isolating control and data planes, and providing OSPF‑ECMP clustering with seamless failover and session synchronization.
Google’s internal Maglev load balancer, described in a recent research paper, bypasses the kernel to operate directly on the NIC, uses ECMP routing, consistent hashing, and memory‑shared queues to achieve sub‑microsecond packet processing while scaling horizontally across clusters.
This article explains how to achieve multi‑million‑packet‑per‑second network capture on Linux 3.16 using standard C/C++ code, by distributing interrupts across cores, employing AF_PACKET with FANOUT and RX_RING, and optimizing memory handling to eliminate costly kernel locks and copies.
This step‑by‑step guide shows how to install and heavily tune Nginx on Linux, apply kernel TCP stack optimizations, and use Tsung for load testing, enabling a single server to sustain around 904,000 HTTP requests per second.
HiStore is a columnar database developed by Alibaba's middleware team, designed for massive OLAP workloads with high compression ratios, low storage and maintenance costs, supporting ad‑hoc multi‑dimensional queries, knowledge‑grid optimization, efficient data loading, and offering features such as approximate queries and high‑availability clustering.
The article reviews best‑practice designs for massive O2O e‑commerce storage, summarizing a Jan 30 technical salon where Meituan’s Cheng Zhenqiang detailed KV storage optimization, Alibaba Cloud’s Zhang Zirong explained high‑performance server architecture tuning, and Meituan Cloud’s Li Kai presented a unified distributed storage system with cross‑data‑center availability, erasure coding, and metadata management.
The article details how Snapdeal built a high‑performance, low‑latency ad‑serving platform that handles billions of daily requests by employing horizontal and vertical scaling, AP‑focused CAP choices, in‑memory data structures, and a suite of open‑source backend technologies.
The article presents a comprehensive overview of OpenResty, describing its architecture, suitable scenarios, selection criteria, real‑world adoption at Qihoo 360 and Sina Mobile, onboarding challenges, future roadmap, and community efforts, illustrating why it is a powerful backend solution for high‑throughput services.
This article details the design, implementation, and performance of Lushan, a high‑throughput offline static data storage server built with libevent that supports dynamic library mounting, key‑value indexing, and efficient query handling for large‑scale recommendation workloads.
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.
This guide walks you through installing and configuring Nginx for extreme request rates, optimizing kernel TCP settings, and using Tsung to benchmark a server that can sustain more than 900,000 requests per second, with detailed command examples and configuration snippets.
