This article analyzes the trade‑offs between software and hardware video decoding on Android, presents performance measurements for different MediaCodec modes, identifies compatibility and latency challenges, and proposes a decoder‑monitoring system and seamless switch architecture to maximize playback efficiency.
This article details Meituan's internal, heavily customized TensorFlow 1.x implementation that addresses large‑scale sparse parameter support, distributed training challenges, communication bottlenecks, and pipeline optimizations, achieving over ten‑fold scalability improvements and significant per‑node performance gains in recommendation system workloads.
This article details the design and implementation of an advertising data platform at eBay, explaining the business scenario, why ClickHouse was chosen over alternatives, the technical challenges faced, and the solutions involving lambda architecture, table engine choices, compression techniques, data ingestion pipelines, consistency guarantees, and deployment practices.
The paper presents a holistic algorithm‑system‑hardware co‑design for advertising deep‑learning inference, combining model pruning, approximate computing, kernel fusion, scheduling and PCIe transfer optimizations with GPU and NPU upgrades, achieving up to five‑fold speed‑up and significantly higher latency‑bounded QPS for large‑scale ad services.
This article explains why H5 pages still need performance attention in modern mobile‑first products and provides a detailed, Baidu‑cloud‑CDN based solution—including background, problem analysis, configuration steps, precautions, and measurable speed‑up results—for accelerating static assets such as CSS, JS, and images.
Over three years Gaode Map continuously optimized its full‑link performance, cutting core‑link latency by half through a systematic three‑step approach—identifying bottlenecks, reverse‑specialized problem solving, and long‑term forward control—resulting in dramatically improved user experience across diverse device tiers.
The article guides readers from grasping hardware, OS, and software layers—including Intel Xeon, virtualization, VPC, and containers—to gathering performance data with tools like JProfiler and perf, then applying data‑driven optimization across business logic, architecture, code, caching, databases, runtime and hardware, illustrated by real‑world case studies and emphasizing continual learning.
This article details Vivo's AI container platform, explaining the design, deployment options, performance tuning, stability measures, and future enhancements of the ingress‑nginx gateway that powers AI services within their Kubernetes clusters.
This article explores practical Swift performance enhancements, covering compilation‑time reductions through smarter type inference, memory‑usage improvements via structs, enums, and collection strategies, and runtime speed gains by understanding function dispatch and applying final, private, and lazy sequence techniques.
This case study details JD.com's 11.11 home‑appliance campaign, covering project background, brand goals, consumer‑centric design thinking, visual creativity with interactive scenes, front‑end performance optimizations, AB‑testing results, and insights on turning design into measurable commercial value.
This article explores how caching accelerates read/write performance and reduces backend load, analyzes its benefits and costs, and presents practical strategies for cache updates, granularity control, penetration, avalanche, and hot‑key issues with concrete examples and diagrams.
This article details Douyin's systematic approach to improving creation experience by defining measurable goals, building a comprehensive metric system, performing relevance analysis, and implementing concrete Android and iOS performance optimizations—including album loading, component architecture, and small‑screen video quality—while outlining monitoring, tooling, and internal platform support that together deliver significant user‑facing gains.
Cloud rendering leverages cloud computing to process and return rendered results to users, enabling Kuaishou to deliver advanced special effects by combining rendering engines with AI capabilities for improved performance and functionality.
This article examines the benefits, costs, and design patterns of caching—including update policies, granularity control, penetration, bottom‑hole, avalanche, and hot‑key issues—while offering concrete optimization techniques and practical recommendations for building robust Redis‑MySQL cache layers.
This article details how Haro Mobility built a search‑driven ride‑matching platform using ElasticSearch and Flink, covering the business background, architectural evolution, data‑sync challenges, performance tuning, stability measures, and the resulting improvements in order completion and user engagement.
To align Hema’s app short‑video experience with leading platforms like Douyin, the team introduced a proxy‑based caching layer, pre‑loading mechanisms, m3u8 handling, bitrate and resolution optimizations, dual‑player strategies, and audio controls, resulting in significantly reduced playback latency and smoother user interaction across iOS devices.
The Youku Playback SDK kernel provides a cross‑platform, high‑reliability framework that decouples data acquisition, decoding, and rendering into independent AVSource, AVDecoder, and AVRender modules, enabling efficient thread usage, configurable builds for partners, adaptive buffering, health monitoring, and comprehensive error handling for optimal playback performance.
Alibaba Cloud's RDS team detailed how cloud‑native transformations, multi‑active deployments, ARM localization, and advanced kernel optimizations enabled the 2021 Double 11 shopping festival to handle massive traffic, cut costs by over 30%, and achieve greener, highly available database services.
This article details Laiye Technology’s AI R&D team’s end‑to‑end approach to designing and optimizing a custom entity‑relationship annotation tool, covering data‑labeling challenges, shortcomings of Excel and off‑the‑shelf solutions, architectural requirements, line‑breaking and mark‑position algorithms, performance improvements, and real‑world results.
Meituan heavily customized TensorFlow 1.x for large‑scale sparse training, replacing variable embeddings with hash tables, improving load balancing, using RDMA communication, pipeline‑embedding graphs, high‑performance hash tables, and operator merges, achieving over ten‑fold scalability, up to 51% operator speedups, and enabling billions‑parameter models on CPU clusters with future GPU expansion.
This article describes how a BCRM sales‑assistant front‑end was built in ten working days by parallelising development, adopting Vue 3, Vite and the NutUI3 component library, and applying performance tricks such as static‑hoisting, Promise.all, async/await optimisation and Vuex‑Map caching to achieve a fast, maintainable product.
This article details how Ele.me's Lriver mini‑program container was optimized over a year to dramatically reduce first‑screen load time, covering architectural improvements, pre‑loading strategies, rendering and worker enhancements, and the monitoring tools that validated the performance gains.
This article reviews the background, architecture, and a series of performance‑optimizing techniques—including consumption, batch, storage, and execution‑engine tweaks—applied to a real‑time pipeline that processes hundreds of billions of records daily, and presents the resulting resource savings and latency improvements.
This article details a step‑by‑step performance optimization of a Python backend service, covering requirement analysis, architecture redesign with caching and Redis queues, load‑testing results, TCP TIME_WAIT issues, and kernel parameter tweaks that ultimately raised throughput to 50,000 QPS with zero errors.
To reduce response time when invoking multiple independent APIs, this article demonstrates how to refactor sequential Java calls into asynchronous executions using thread pools and CompletableFuture, providing complete code examples and performance comparisons.
Zhihu’s Moneta service stores over a trillion rows and faces massive write and read loads; this article explains why TiDB was chosen, how its architecture and features such as HTAP, Raft, Titan and table partitioning enable millisecond‑level query latency, high availability, and seamless scaling.
Adopt micro frontends when large, parallel development teams need independent releases, cross‑technology migration, or high‑frequency updates, and design them with a central app shell loading shared dependencies, choosing between iframe, web component, or lifecycle‑hook bundles, managing single‑ or multi‑instance routing, shared state or event‑bus communication, and optimizing performance through lazy rendering and shared library bundling.
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 describes the design, challenges, architecture, key features, algorithm optimizations, and future roadmap of Kuaishou's Starry vector retrieval platform, which delivers high‑performance, high‑reliability, and easy‑to‑use large‑scale ANN search for diverse business scenarios.
This article details a two‑week effort to diagnose and resolve high CPU usage, server load, Hystrix circuit‑breaker inefficiencies, and Spring data‑binding exceptions in a Java backend service, resulting in doubled QPS capacity, stable circuit breaking under heavy traffic, and significant performance gains.
This article explains how Flink’s new Sort‑Shuffle mechanism improves stability and performance for massive batch jobs by reducing file counts, optimizing I/O, minimizing memory usage, and providing detailed implementation, test results, tuning tips, and future enhancements.
This article details how a Go‑based Dubbo registration center suffered high CPU usage, how profiling identified the AssembleUrlWeight function as the bottleneck, and how custom URL‑parameter extraction and weight‑insertion code cut CPU consumption by over ten percent and improved performance by up to twenty times.
This article examines the signs that a database table has reached its limits, explains why vertical and horizontal sharding are needed, offers concrete sizing formulas, compares hash, range and consistent‑hash partitioning, and shares large‑scale case studies from Suning, JD, Meituan, Ant Financial and Taobao.
This article analyzes why Alipay’s cold start on our device is slower than on competing phones and presents a series of optimizations—including CPU boosting, Dex2oat strategy adjustments, GC frequency reduction, and I/O improvements—that collectively close the performance gap.
Alibaba Sports’ AI-powered smart workout system transforms a simple smartphone and a few square meters of space into an interactive home fitness solution, using MNN‑based pose estimation to recognize and correct dozens of exercises, while addressing challenges like accuracy, performance, and automated testing.
This article examines a user‑side logging system where encrypted and compressed logs are uploaded, analyzes why decryption often stalls or fails, explores Node.js multi‑process and inter‑process communication methods, addresses sticky‑packet and exception handling issues, and presents solutions for performance, memory leaks, and reliable npm package publishing.
This article details the architecture of the Qingzhou Business Gateway, identifies its granular control, data‑loss, and performance issues, and explains a series of optimizations—including FFI usage, table‑pool reuse, coroutine caching, radixtree routing, and connection‑pool tuning—that raise single‑node QPS to 80 k while preserving functional capabilities.
This article explores service governance fundamentals, examines why etcd’s Raft‑based consensus and BoltDB storage make it ideal for large‑scale systems, compares it with ZooKeeper and Consul, and shares Baidu’s practical architecture, performance tricks, and operational metrics for high‑availability, high‑performance service management.
The article presents a Guard abstraction that temporarily borrows and returns Protobuf field pointers via set_allocated/release, eliminating costly CopyFrom operations in recommendation pipelines, enabling zero‑copy field sharing across central‑control and recall stages, improving CPU usage and latency while handling safety and rollback.
The article explains how Alibaba’s QianNiu multi‑platform workbench reduced open‑plugin related user complaints by half through defining open nodes, optimizing long plugin startup chains, building permission‑request loops, establishing comprehensive data‑driven metrics, and creating an open‑experience dashboard that monitors performance, reliability, and user‑perceived issues across mobile and desktop.
This article details the migration from Webpack to Vite, highlighting performance improvements and addressing common issues like SVG component support and third-party package bugs.
This article revisits a Python combinatorial challenge—selecting five numbers from a random fifteen‑element list that include consecutive values—presenting an optimized solution that leverages NumPy arrays, set operations, and intersection logic to cut execution time from twelve seconds to about one and a half seconds, while explaining the underlying reasoning.
This article explains how isolating resources such as threads, processes, clusters, and data centers improves efficiency, prevents cascading failures, and enhances fault tolerance in distributed backend systems, providing practical examples with Netty, Dubbo, Tomcat, and Hystrix.
This article shares practical Java performance‑optimization techniques, covering UI and non‑UI latency metrics, baseline data collection, bottleneck discovery with tools like Arthas, chronic issue handling, and a comprehensive set of governance measures ranging from network‑level caching to code‑level refactoring, asynchronous processing, and service splitting to achieve stable sub‑200 ms response times.
This technical sharing session explores the Memory Graph capabilities within Volcengine's MARS-APMPlus monitoring suite, detailing how the Feishu iOS development team leveraged this non-intrusive tool to effectively diagnose and resolve Out-Of-Memory crashes, significantly enhancing mobile application stability and performance.
This article compiles frequently asked HBase questions, troubleshooting steps, performance optimization techniques, configuration guidance, and an overview of new HBase 2.0 features such as off‑heap memory, Procedure v2, In‑Memory Compaction, and MOB support, providing practical solutions for administrators and developers.
This article explains how computer systems combine hardware and system software, describes the memory hierarchy, operating‑system abstractions, Amdahl's law, and various forms of parallelism, and shows why mastering these fundamentals can dramatically improve program efficiency and reliability.
When a Dubbo interface handling 18 billion daily requests began failing due to thread‑pool exhaustion, a systematic investigation uncovered traffic spikes, Redis request overload, and mis‑configured connection‑pool parameters, leading to a series of targeted fixes—including timeout adjustments, Redis scaling, and pool‑parameter tuning—that restored stability and dramatically reduced error rates.
This article investigates the Android dex2oat compilation restrictions introduced for targetSdkVersion ≥ 29, analyzes their impact on app performance, explores SELinux enforcement, and presents a practical solution using system commands and binder calls to trigger dex2oat compilation for both primary and secondary APKs, improving load and runtime speed.
This article explains the design and implementation of a flexible danmaku SDK that uses AI‑based image segmentation to let comments flow around people in videos, covering architecture, algorithmic processing, server deployment, client rendering options, performance challenges, and future extensions.
By upgrading Alibaba Cloud CDN from TLS 1.2 to TLS 1.3 and adding a Qiniu fallback, the DeWu app cut image and video loading latency by roughly 10‑15% on iOS and Android, eliminating the handshake bottleneck and enhancing security and user experience.
The article explains how JD’s Activity Cockpit tackles mega‑sale challenges by monitoring the consumer golden‑link, applying Optimize Table, BitMap, and materialized view techniques to reduce data volume, accelerate queries, and enable precise real‑time marketing for brands.
This article explains how NetEase Cloud RTC implements H.265 video encoding, covering capability negotiation design, cross‑platform performance tests on Android, iOS, macOS and Windows, and engineering strategies such as whitelist policies, CPU overuse handling, QP threshold tuning, and the resulting quality and efficiency benefits.
KonvaJS is a TypeScript‑based Canvas framework offering concise APIs and React/Vue bindings, built on a hierarchical Stage‑Layer‑Group‑Shape tree that manages DOM containers, batch rendering via requestAnimationFrame, off‑screen caching, color‑key hit detection, drag handling, filters, selectors, serialization, yet lacks dirty‑rectangle optimization for large‑scale shape updates.
Baidu’s massive information‑flow recommendation platform employs an intelligent delivery pipeline—spanning micro‑service R&D, automated white‑box testing, performance monitoring, and optimized deployment—that supports nearly a hundred daily releases, cuts QA effort, delivers over half of requests within a day, and enables near‑zero‑touch, high‑frequency rollouts.
This article demonstrates three progressive techniques—direct list insertion, grouped batch insertion, and multithreaded batch insertion—using MyBatis to dramatically reduce MySQL bulk‑import time from over a minute to under ten seconds while handling packet size limits.
This article introduces Impala's overall architecture, storage options, key features, concurrency mechanisms, CBO‑based join optimization techniques, storage‑layer principles and data‑filtering strategies, and summarizes practical performance‑tuning experiences from Tencent's financial big‑data platform.
This article explains the browser rendering pipeline, the concepts of reflow and repaint, how they are triggered, their performance impact, and provides practical techniques such as minimizing layout thrashing, using GPU‑accelerated properties, and leveraging requestAnimationFrame to optimize front‑end performance.
The iQIYI deep‑learning platform identified two TensorFlow Serving memory‑leak problems—a string‑accumulating executor map caused by unordered input maps and an uncontrolled gRPC thread surge under heavy load—submitted upstream patches that sort inputs and cap thread counts, eliminating OOM crashes and stabilizing production.
This article shares practical experiences with Doris, covering materialized view concepts, typical use cases, advantages, creation syntax, prefix index principles, performance‑boosting scenarios such as order analysis, PV/UV counting, detail queries, and operational tips for high‑throughput and low‑latency workloads.
This article analyzes performance and concurrency problems in a code‑diff service, compares the original JGit‑based approach with a GitLab‑API solution, addresses new accuracy issues, and presents a compensation strategy using java‑diff‑utils to achieve stable, efficient backend diff processing.
This article examines the fundamental characteristics of Redis, explains why network latency often dominates performance, and presents practical optimization strategies such as using Unix IPC, multi-key commands, transactions, scripts, pipelining, avoiding slow commands, tuning persistence, adjusting OS settings, and adopting distributed architectures to achieve higher throughput and lower latency.
NetBank’s database deployment has evolved from a simple two‑site three‑center disaster‑recovery model to a sophisticated three‑site five‑center architecture, incorporating distributed databases, multi‑tenant isolation, transaction consistency, latency optimization, and containerized deployment to achieve high availability, scalability, and cost efficiency.
Tencent Cloud's VP Li Cong explains how advanced traffic scheduling, resource orchestration, dynamic routing, and personalized "thousand‑faces" strategies enable CDN systems to locate the fastest nodes worldwide, boosting performance by up to 10% while enhancing security and preparing for next‑generation edge computing.
Getty is a Go-based, layered network framework that provides data interaction, business control, and network layers with extensible monitoring interfaces, featuring custom SDKs, goroutine pool optimizations, lazy reconnection, sync.Pool tuning, and benchmark results demonstrating high throughput and low resource usage.
This article introduces the Jamstack architecture, compares it with traditional web stacks, and details how Douyin's e‑commerce 818 main‑hall performance optimization project applied Jamstack principles, CDN caching, API prefetching, and client‑side techniques to achieve faster, more scalable page delivery.
The article explains MySQL vertical partitioning techniques—including instance‑level database splitting, schema‑level splitting, and column‑level table splitting—detailing their advantages, drawbacks, and step‑by‑step command‑line procedures with code examples for implementing each method in production environments.
This article explains how to use Java 17 features such as pattern matching and switch expressions in IntelliJ IDEA 2021.2.1, demonstrates performance‑boosting tips by disabling unnecessary plugins, and provides practical code examples to streamline your Java development workflow.
This article details how the Fluid CNCF project combined with JindoRuntime was used to overcome storage‑compute separation bottlenecks in an autonomous‑driving machine‑learning platform, achieving up to 300% faster training, reduced OSS bandwidth pressure, and higher GPU utilization through distributed caching on Kubernetes.
This article explains how functional components, React hooks, and memoization techniques such as React.memo, useCallback, and useMemo can be combined to reduce unnecessary re‑renders, simplify side‑effect management, and improve overall UI performance in modern React applications.
This article explores TiDB performance optimization by examining layered approaches—hardware, software, and SQL—and scenario‑specific tactics such as read/write hotspot handling, OLAP challenges, and distributed transaction tuning, offering practical insights and real‑world case studies from industry experts.
This article details ByteDance's implementation of an exabyte‑scale data lake using Apache Hudi, covering scenario requirements, engine selection, functional support, schema management, extensive performance tuning, and future directions, while also noting recruitment opportunities within the team.
This article details Baidu's internal adoption of a service mesh built on Istio and Envoy, covering the motivations, architectural design, low‑invasion integration methods, extreme performance tuning, stability and traffic governance capabilities, surrounding ecosystem tools, and the resulting operational benefits.
Users uninstall apps when friction—such as privacy worries, useless features, crashes, slow performance, or intrusive ads—overrides desire, so applying a five‑level APP experience pyramid (Security, Functionality, Stability, Usability, Delight) with privacy checklists, exploratory testing, full‑stack monitoring, and real‑user usability sessions can markedly boost retention.
This article provides a thorough overview of Apache Spark, covering its origins, comparison with MapReduce, core concepts such as RDD, DAG, Jobs, Stages, and Tasks, the submission process, Web UI, and detailed performance tuning techniques including data skew mitigation.
This article shares practical lessons from three iterations of a data platform, focusing on Elasticsearch and Lucene optimizations that enable cross‑month queries, year‑long data export, and sub‑second query responses for tables handling billions of rows per day.
This article examines the emerging Storage Class Memory (SCM) market, outlines its various technologies, evaluates performance and cost trade‑offs, explores three concrete use cases—AI training acceleration, instant data recovery, and greener data‑center operation—and discusses the latency and workload‑model challenges that must be solved for widespread adoption.
This article introduces ClickHouse, a high‑performance open‑source columnar database, explains its architecture versus row‑based systems, details migration from MySQL, showcases installation, performance benchmarks, data‑sync strategies, common pitfalls, and summarizes its benefits for large‑scale analytical workloads.
This article documents a Python service performance optimization journey, covering requirement analysis, environment setup, load testing, database bottleneck mitigation, caching, TCP time‑wait tuning, and kernel parameter adjustments that ultimately raised QPS to 50,000 with stable latency and zero errors.
This article analyzes the latency of video publishing in a mobile app, proposes a parallel video‑track transcoding solution, details implementation steps, performance testing, and optimization techniques to reduce encoding time by up to 30% on Android devices.
This article presents practical strategies for building high‑performance PyTorch training pipelines, covering bottleneck identification, efficient data loading, RAM‑based datasets, profiling tools, multi‑GPU training with DataParallel and DistributedDataParallel, custom loss implementation, and hardware‑vs‑software trade‑offs to accelerate deep‑learning workloads.
This article demonstrates how to dramatically speed up importing tens of thousands of records into MySQL by evolving a naïve list‑to‑MySQL approach into grouped batch inserts and finally a multi‑threaded MyBatis solution, while also addressing packet size limits and configuration tweaks.
This article introduces ClickHouse as a column‑oriented OLAP database, compares it with traditional row‑based databases, and details a real‑world migration from MySQL that reduced query time from minutes to under one second, along with performance testing, sync strategies, and common pitfalls.
This talk shares practical high‑availability strategies learned from Alibaba’s Taobao platform and Alibaba Cloud, covering traditional IDC stability, cache and disaster‑recovery designs, cloud‑native fault‑tolerance, performance‑capacity trade‑offs, traffic shaping, multi‑region replication, and lessons from real‑world incidents like GitLab failures.
The article details MOO Music’s experience with Flutter memory management, explaining the Dart heap, stack, and external memory structures, common leak sources such as missing listener deregistration, unbounded widget lists, retained timers, static references, and third‑party or engine bugs, and offers practical detection and optimization techniques.
This article summarizes the CovalenceConf 2019 talk on VS Code’s startup performance, detailing measurement practices, code‑size reductions, lifecycle re‑ordering, V8 code caching, requestIdleCallback usage, and perceived‑performance tricks that together shrink launch time from seconds to under one second.
This article details how to dramatically speed up the import of massive Excel fee‑record files in a Java backend by switching to EasyExcel, caching database lookups, using batch INSERT statements, and applying parallel streams while eliminating excessive logging.
This article provides a comprehensive overview of Spark SQL, covering its architecture, DataSet/DataFrame APIs, code examples for creating and querying datasets, join strategy selection, handling Hive tables, small‑file issues, inefficient NOT‑IN subqueries, Cartesian products, and a catalog of useful built‑in functions.
This article analyzes how landing‑page load speed impacts user experience and conversion, presents data‑driven measurements and key‑path breakdowns, and details practical frontend optimization methods—including container initialization, parallel billing, resource pre‑loading, and server‑ and client‑side pre‑rendering—to significantly improve performance across commercial ad flows.
This article explains how iOS app launch works, presents methods for measuring startup time using system calls and Xcode Instruments, shares a detailed case study of the Maimai app, and offers concrete pre‑main and post‑main optimization techniques—including binary reordering, dynamic‑library reduction, concurrency control, and unconventional tricks—to achieve sub‑second launch performance.
This article explains the concepts of debounce and throttle in JavaScript, illustrates their principles and typical use cases such as button submissions, search suggestions, window resize, and drag‑and‑drop, and provides practical implementations with code examples and options for immediate execution, timestamps, and timers.
This article explains system design fundamentals—including architecture, modules, components, interfaces, and data—outlines the main tasks and step‑by‑step process, covers performance challenges, RAIL performance model, key web metrics, and Facebook's design values, providing a comprehensive guide for building robust applications.
By illustrating how to replace simple HashMap caching with Redis‑based distributed caches and locks—using SETNX, Lua scripts, and Redisson—the article shows Spring Boot developers how to prevent cache breakdown, ensure data consistency, and dramatically improve throughput in high‑concurrency web applications.
This article reviews Ceph's architecture, enumerates common benchmarking tools, analyzes its advantages and challenges, and presents a comprehensive set of performance‑optimization methods covering storage‑engine tuning, network communication, data placement, configuration parameters, hardware‑specific adaptations, and future research directions.
This article systematically presents the cold‑start performance problems of the Beike iOS app, defines testing standards, explains the essence of optimization, lists common pitfalls, and details a comprehensive set of practical solutions ranging from lifecycle‑aware task scheduling and minimal‑set launchers to dynamic‑library lazy loading, compile‑time I/O elimination, static‑initializer handling, dead‑code removal, and monitoring standards.
This article summarizes the key techniques presented at CovalenceConf 2019 for optimizing Visual Studio Code's startup performance, covering measurement, bundling, V8 code caching, lifecycle phase ordering, requestIdleCallback, and perceived‑performance tricks that together reduce launch time to under one second.
This article documents a step‑by‑step performance optimization of a Python web module, covering requirement analysis, environment setup, load‑testing results, database and TCP bottleneck identification, caching strategies, kernel tuning, and the final achievement of 50,000 QPS with low latency.
Tencent’s MongoDB team solved the severe performance degradation caused by millions of collections by redesigning WiredTiger to use a shared table‑space with prefix‑mapped keys, eliminating data‑handle explosion, cutting memory use and lock contention, reducing startup from hours to minutes, and boosting read/write throughput by up to two orders of magnitude while preserving full MongoDB compatibility.
The article outlines how Baidu iFanFan reduces mobile web page load times from 2‑3 seconds to under one second by bundling static resources, pre‑rendering and pre‑initializing WebViews, employing a micro‑frontend master‑slave architecture, using offline packages, native request routing, and early network pre‑loading.
This article outlines the career path toward software architecture, detailing the market expectations, essential competencies such as project management, technical depth, business insight, communication, decision‑making, and optimization techniques for systems, backend services, and databases, while emphasizing continuous learning and community involvement.
