This article details the background, architecture, and step‑by‑step performance optimizations—including Packager caching, request reduction, Service‑Worker caching, and Webpack‑style externals—that reduced a CodeSandbox sandbox build from around two minutes to roughly one second.
ZRTC, the real‑time audio‑video platform behind 作业帮, has been refined for over three years to support massive, multi‑cloud, multi‑protocol live streaming, employing a unified SDK, intelligent scheduling, custom SFU services, and extensive performance tuning to achieve high concurrency, low latency, and robust high‑availability.
The article presents a complete Java solution for generating, reading, and analyzing a 10 GB file of age values (18‑70) on a 4 GB RAM, 2‑core PC, comparing single‑threaded counting with a producer‑consumer multithreaded design that dramatically improves CPU utilization and reduces total processing time.
In this interview, 58.com front‑end architect Jiang Hongwei discusses his journey into React Native, the evolution and benefits of the new RN architecture, challenges of upgrading, comparisons with other cross‑platform frameworks, and offers career advice for technology professionals.
This newsletter highlights the formation of a JavaScript runtime interoperability group, new Chrome 102 DevTools features, Google’s performance‑boosting private prefetch proxy, a comparison of Vue and React performance, Node.js advantages for microservices, and several recent articles on frontend tooling, image optimization, and modern JavaScript language features.
The article presents MCube, a cross‑platform native dynamic UI framework used in JD.com’s mobile app to improve performance, reduce development effort, and enable rapid, stable feature delivery through a modular architecture, XML DSL, Yoga layout engine, and comprehensive management and tooling support.
Tulong is Meituan’s industrial graph neural network framework and learning platform that combines a compact MTGraph engine, a modular operator‑based GNN library, and visual workflow tools to enable heterogeneous, billion‑edge graph training on a single machine with up to 60 % memory savings and 2–4× speedups, streamlining search, recommendation, advertising and delivery pipelines.
To eliminate frame drops caused by animated PopupWindow and Snackbar conflicts, the article shows how moving costly view inflation and show/dismiss operations to a dedicated HandlerThread—while still invoking show() on the UI thread—creates a separate UI thread that respects Android’s thread‑checking rules and keeps animations smooth.
A Kafka cluster at vivo suffered a total traffic drop across a resource group when a broker’s disk failed, because the default producer partitioner still hashed keys to the failed partition, exhausting client buffers and blocking all healthy partitions, prompting recommendations to avoid keys or use custom partitioners.
This article presents JD Retail's OLAP platform construction and practical scenarios, covering control‑plane design, architecture, business management, operational safeguards, real‑time data updates, materialized view acceleration, join optimization techniques, high‑concurrency queries, and large‑scale write throughput for e‑commerce peak periods.
This article details how a Java Spring Boot microservice's response time was reduced from 5‑6 seconds to under one second by applying gzip compression, parallel data fetching, short‑lived caching, MySQL index tuning, and JVM G1 garbage‑collector adjustments.
This article walks through the challenges of inserting massive data sets with MyBatis, explains the inner workings of batch execution, highlights common mistakes with commit, clearCache and flushStatements, and presents progressively refined code examples—including Oracle-specific optimizations—to dramatically improve performance.
This article details Bilibili's end‑to‑end Presto on Hadoop architecture, covering the multi‑engine SQL stack, dispatcher routing, cluster scale, stability enhancements like coordinator HA and real‑time punish, query limits, Hive UDF compatibility, insert‑overwrite support, Alluxio caching, multi‑datacenter routing, query result caching, Raptorx local cache, JDK upgrades, dynamic filtering, and future roadmap, illustrating how these innovations boosted query throughput and reduced latency.
This article demonstrates how to efficiently update up to 50,000 MySQL records in a Spring Boot‑MyBatis application by comparing simple loops, manual transaction control, multithreaded execution with CountDownLatch coordination, and union‑based batch updates, while analyzing performance and connection‑pool considerations.
This article introduces a Spark offline development framework that separates configuration from code, supports SQL and Java applications, and provides fast, automated data backtracking with reduced environment preparation time, lower failure rates, and significant performance gains for large‑scale data warehouses.
This article explains how ByteDance's DanceCC team improved debugging speed for large Swift iOS projects by customizing the LLDB toolchain, disabling costly Swift type‑resolution features, fixing static‑framework loading, optimizing external module searches, and introducing shared symbol caches, achieving up to ten‑fold reductions in variable‑view and expression‑evaluation times.
This article details eBay's evolution of its massive HDFS storage—from a single‑cluster design to ViewFS Federation, then to Router‑Based Federation—highlighting the performance bottlenecks, optimization techniques, FastCopy integration, and future plans for further scaling and automation.
The article explains how technical leaders can apply quantitative analysis and systematic measurement to optimize both software system performance and organizational workflows, using a picture‑recognition service example and a real‑world DevOps incident to illustrate the need for end‑to‑end process mapping, bottleneck identification, and continuous improvement.
The Youku iOS team investigated why first‑breakpoint pauses could exceed three minutes after a massive Swift migration, analyzed LLDB internals and external module dependencies, tried and discarded several refactoring and compiler‑flag approaches, and finally achieved a dramatic speedup by tweaking LLDB settings and adding a small Swift helper for missing property prints.
This article walks through the challenges and solutions for building a high‑performance long‑connection service with Netty, covering Netty basics, non‑blocking I/O, sample Java NIO and Netty code, Linux kernel tuning, QPS optimization, data‑structure tweaks, GC adjustments, and the final performance results.
This article explains how to build a scalable long‑connection server using Netty, covering the fundamentals of non‑blocking I/O, practical code examples for achieving millions of concurrent connections, common Linux kernel limits, CPU and GC bottlenecks, and a series of tuning techniques to dramatically improve QPS and stability.
This article provides a comprehensive introduction to Elasticsearch, covering data types, Lucene fundamentals, cluster architecture, node roles, shard and replica mechanisms, mapping, basic usage, health monitoring, indexing workflow, storage strategies, and practical performance tuning techniques.
This article details how the iFanFan team transformed a monolithic Vue front‑end into a micro‑frontend system using a unified CLI tool, covering routing, packaging, independent development, version management, isolation, communication, CI/CD integration, multi‑product support, and performance optimizations.
This article shares Ctrip's practical experiences in scaling a hybrid private‑cloud platform to over ten thousand nodes, covering Kubernetes control‑plane stability, host monitoring, network observability, image management, and capacity planning to ensure high availability for massive online services.
Baidu’s Search Architecture team evolved compute management from static physical‑machine deployment to cloud‑native mixed deployment, employing a governance system that combines elastic containers, Service Mesh, and committee‑driven operations, while leveraging tidal scaling, container tiering, performance‑curve‑based VPA/HPA, and fine‑grained traffic scheduling to deliver cost‑effective, flexible, and performance‑driven resource allocation across both temporal and task‑type dimensions.
When a Java application launches, initial requests often feel sluggish because the JVM relies on an interpreter, but as frequently executed code becomes hot, JIT compilation translates it into optimized native machine code, dramatically improving response times—a process explained along with practical strategies like Dragonwell’s JWarmUp and request pre‑warming to accelerate performance.
The article introduces a token‑level pipeline parallelism strategy that splits the sequence‑length dimension of Transformer‑based language models, explains why this approach is feasible, presents a dynamic‑programming formulation for optimal slicing, discusses engineering challenges, and evaluates its performance on large GPT models.
In a shared Elasticsearch cluster, the team used slow‑log analysis to pinpoint costly queries caused by unnecessary fuzzy matches and integer‑mapped low‑cardinality fields, then optimized them by converting matches to filters and remapping those fields to keyword, re‑indexing, which cut latency from over 100 ms to under 10 ms and eliminated slow‑query alerts.
Each week WecTeam Frontend Weekly highlights top technical reads, offering concise insights into React’s Scheduler mechanics, a side‑by‑side comparison of modern build tools like esbuild, Snowpack, Vite and wmr, and a case study on dramatically boosting H5 page load speed in the Dewu app.
This article examines common risks associated with game art assets, presents real-world external case studies of asset‑related failures, and outlines systematic static and dynamic inspection methods across production, upload, and packaging stages to improve quality, performance, and player experience.
This article explains Redis's original single‑threaded event‑driven design, the problems it causes, how the Lazy Free mechanism and asynchronous deletion were introduced in Redis 4.0, and how Redis 6.0 adds multi‑threaded I/O while discussing its limitations and comparing it with Tair's approach.
This article examines frequent SQL anti‑patterns—such as inefficient LIMIT usage, implicit type conversion, suboptimal UPDATE/DELETE joins, mixed ordering, unnecessary EXISTS clauses, condition push‑down failures, early limiting, and intermediate result handling—and provides rewritten queries that dramatically improve performance.
This article explains Cube's evolution from early native rendering challenges to a sophisticated asynchronous, cross‑platform rendering architecture that uses bitmap caching, platform‑specific layers, and a multi‑stage optimization process to boost mobile UI performance on both Android and iOS.
This article explains the design, mathematical foundation, and performance‑oriented optimization of a lasso tool that enables batch selection, rotation, scaling, and translation of graphics on an HTML Canvas‑based online whiteboard, including matrix‑based transformation, state recording, undo/redo, and experimental results.
This article details how ZuoYeBang's engineering team migrated their Hive‑based offline data warehouse to Delta Lake, tackling latency, scalability, and query‑performance challenges through stream‑to‑batch processing, data‑lake architecture, and optimizations like DPP and Z‑ordering.
ByteDance rebuilt Spark's History Server into a cloud‑native UIService that stores only essential UI metadata, reducing storage usage by over 90%, cutting UI latency by up to 94%, and enabling seamless horizontal scaling for large‑scale analytics workloads.
This article details Bilibili's end‑to‑end Presto‑based SQL‑on‑Hadoop architecture, covering overall system components, query routing, Presto feature set, extensive stability and availability enhancements, performance boosts through caching and multi‑datacenter deployment, and outlines future development plans.
This comprehensive article explains what Elasticsearch is, how it builds on Lucene to provide distributed real‑time search and analytics, covering data types, cluster components, shard routing, indexing pipelines, storage formats, segment merging, and practical performance‑tuning tips for production deployments.
This article provides a theoretical overview of seven major Java performance optimization techniques—including code reuse, computation parallelism, asynchronous processing, lazy loading, result set reduction, resource conflict handling, and JVM tuning—offering readers a comprehensive framework to understand and improve application efficiency.
openLooKeng 1.6.0 introduces major enhancements such as faster data persistence with Kryo serialization, HDFS spill support, improved connector capabilities, low‑latency features like Star Tree indexing and cube updates, and a refined task recovery mechanism that reduces failure detection time, delivering a more reliable and high‑performance big‑data engine.
This article recounts a hiring manager’s interview with a candidate lacking high‑concurrency experience, then outlines the candidate’s self‑driven three‑stage practice—single‑machine optimization, load‑balanced testing, and middleware integration—providing concrete steps for engineers to acquire high‑concurrency expertise.
This article explains the fundamentals of Android app startup, categorizes cold, hot, and warm launches, identifies performance bottlenecks using tools like ADB, Systrace, and Traceview, and presents concrete optimization strategies for Application creation and splash-screen rendering that reduce launch time by up to 60 percent.
This article explains OpenCL fundamentals, the Qualcomm Adreno GPU architecture, compatibility considerations, and practical optimization techniques—including profiling, bottleneck identification, and CPU‑to‑GPU conversion tips—to help developers write high‑performance mobile OpenCL code.
OpenKruise v1.1 extends Kubernetes with ordered container startup, in‑place upgrade sequencing, StatefulSet PVC auto‑deletion policies, a refactored Advanced DaemonSet with lifecycle hooks, and a DisableDeepCopy performance tweak, all while remaining compatible with clusters running Kubernetes 1.16 or newer.
After months of using MongoDB for frontend monitoring logs, the data grew to billions of records causing severe query slowdown, prompting a migration to ClickHouse where columnar storage, partitioning, and OLAP capabilities dramatically improved performance and storage efficiency.
This article describes how the Same‑City team migrated H5 mini‑games to a native Swift/SpriteKit solution, detailing the mixed Objective‑C/Swift environment, core game architecture, use of Swift functional and generic features, and the resulting improvements in package size, startup time, FPS, and memory consumption.
This article explains how Apache Doris adopts CPU‑level vectorization and columnar storage to boost query performance, details the design and current status of its vectorized engine, and outlines future work such as JOIN acceleration, storage‑layer vectorization, import optimization, and extensive SQL function support.
The article delivers a comprehensive, code‑level examination of Linux’s schedutil (sugov) CPU‑frequency governor, detailing its architecture, key data structures, registration with the cpufreq core, utilization‑driven frequency calculation (including iowait boost and limit handling), and the start‑stop lifecycle, while noting its default status and remaining mobile‑platform challenges.
This article details how ZhaiZhai tackled severe order database performance and capacity issues by adopting ShardingSphere for sharding, describing the challenges, mitigation steps, shard key design, migration process, and the significant performance improvements achieved.
This comprehensive guide explains Elasticsearch fundamentals, including data types, Lucene and inverted indexes, cluster and node roles, shard and replica mechanisms, mapping, installation steps, health monitoring, write and storage processes, segment merging, and practical performance tuning tips for large‑scale search deployments.
NextRPC is a multi‑stage RPC model that delivers partial responses over multiple network channels, improving latency and conversion in Taobao’s transaction and recommendation services; it employs a hybrid asynchronous streaming and parallel execution architecture with client‑side orchestration and server‑side sub‑request handling, achieving over 5% UV lift and up to 25% recommendation uplift.
This article shares practical lessons from optimizing the performance of legacy web applications, covering how a hidden DB‑connection leak caused severe latency, why average latency is misleading, and the essential tools, testing, and team practices needed to keep services fast and reliable.
This guide explains Linux performance optimization by defining key metrics such as throughput and latency, clarifying average load, detailing CPU context switches, describing common performance analysis tools, and providing practical methods for diagnosing and improving CPU, memory, and I/O bottlenecks in production environments.
ByteDance’s Spark History Server was re‑engineered into a cloud‑native UIService that reduces storage usage by over 90%, cuts UI latency by up to 94%, and enables horizontal scaling, dramatically improving the user experience for large‑scale Spark jobs.
Alibaba’s MNN, a lightweight high‑performance deep‑learning inference engine, earned top honors in China’s 2022 “Science & Innovation China” awards, and delivers impressive gains such as 350% speedup on X86 CPUs, 2.1‑2.3× acceleration on ARM with sparse models, plus integrated OpenCV/Numpy functionality for edge AI deployment.
This article provides a comprehensive overview of Elasticsearch, covering its data types, Lucene-based inverted index, cluster architecture, sharding and replication mechanisms, mapping definitions, basic usage, health monitoring, storage internals, and practical performance tuning tips for large‑scale search deployments.
This article recounts the author's journey from using MongoDB for front‑end monitoring logs to migrating to ClickHouse, detailing the challenges with large‑scale data, optimization attempts, the fundamental differences between row‑ and column‑oriented databases, and the resulting performance and storage improvements.
This comprehensive guide explains caching fundamentals, motivations, and hardware/software perspectives, details client, network, and server cache types, addresses common problems like breakdown, avalanche, and penetration with solutions, and covers strategies, consistency, expiration policies, distributed architectures, design best practices, and implementation examples.
The article details Kwai’s OneService platform—a low‑code, self‑service data platform that streamlines API creation, deployment, and operation, covering its background, architecture, key technologies such as API matrix and configuration‑as‑code, high‑availability and performance strategies, achieved results, and future roadmap.
This article details the systematic approach Kuaishou’s video editing app took to define, reproduce, analyze, and resolve iOS out‑of‑memory (OOM) crashes, leveraging edit‑context capture, action tracking, memory profiling, and targeted optimizations that reduced OOM incidents by over 80%.
This article details the engineering optimizations applied to the image cropping stage of Kuaishou's YKit AI SDK, covering instruction-level fixes, SIMD acceleration, I/O cache improvements, algorithmic refinements, parallel processing, and device‑tier strategies to achieve up to 4.6× speedup on mobile devices.
The Go community’s arena proposal introduces a contiguous‑memory allocation method that reduces allocation overhead and enables bulk freeing, offering up to 15% CPU and memory savings for large applications while ensuring safety through dynamic checks.
The author describes a high‑QPS Java web service, analyzes its GC behavior using jstat and gceasy, applies JVM parameter adjustments such as enlarging the young generation, tweaking MaxTenuringThreshold and disabling biased locking, and achieves roughly a 15% overall performance gain with improved GC throughput.
This article details the motivation, technology selection, architecture design, and multi‑stage performance optimization of a server‑side rendering (SSR) solution built with Vue and Nuxt for the 360 Search Encyclopedia mobile project, covering challenges, experiments, deployment strategies, and final conclusions.
NextRPC introduces a multi‑segment payload RPC model that streams responses through successive "Next" stages, enabling parallel processing and decoupling of core and non‑core logic, which improved Alibaba’s Double‑11 transaction conversion rates by over 5% overall and more than 25% for optimal‑product selection scenarios.
This article details the 2021 overhaul of iOS WeChat's full‑text search, covering engine selection, segment‑merge optimization, a new VerbatimTokenizer, multi‑level separator support, table schema choices, asynchronous index updates, and extensive performance gains across chat, contacts, and favorites.
This article explores user experience from a front‑end perspective, defining UX, examining the many factors that influence it—including user state, environment, and product features—and outlining practical design considerations such as performance, interaction details, and accessibility to create truly engaging digital products.
This article explains Kafka’s fundamental architecture—including topics, partitions, replication, consumer groups, cluster coordination, and performance techniques such as sequential writes, zero‑copy, and log segmentation—to help readers improve their design and coding skills for large‑scale messaging systems.
Alinode PGO leverages profile‑guided optimization to generate a cache of hot startup code for Node.js functions, cutting cold‑start latency by up to 61% (150% speed‑up) and allowing further reductions by launching solely from the PGO cache, as demonstrated with real‑world benchmarks.
The article shows how OpenSumi’s UI lag, caused by global‑state re‑renders, redundant props, and repeated menu and tree creation, can be eliminated by memoizing components, splitting views, caching menus, removing unneeded width props, batching icon CSS insertion, and debouncing ResizeObserver events, cutting latency from seconds to smooth responsiveness on both M1 and Intel machines.
This article details Alibaba's mobile engineering team's original approach to full‑link observability, describing the challenges of the Taobao app architecture, the evolution of monitoring to observability, the Falco OpenTracing model, and practical performance optimizations that improve issue‑resolution efficiency and user experience.
This article introduces the basics of WeChat mini programs and the mpvue and uni‑app frameworks, analyzes their architectures and performance differences, and presents a detailed migration plan from a large mpvue project to uni‑app, including code adjustments, configuration changes, and optimization techniques.
This article provides a comprehensive guide to optimizing SpringBoot services, covering metric exposure with Prometheus, custom business monitoring, Java flame‑graph profiling, SkyWalking distributed tracing, HTTP and Tomcat tuning, layer‑wise code improvements, and practical code examples for real‑world performance gains.
The article discusses the importance of caching in high‑concurrency distributed systems, outlining its benefits and costs, various update policies such as LRU/LFU/FIFO, expiration, active refresh, and advanced optimizations like penetration protection, hole mitigation, avalanche prevention, and hot‑key rebuild techniques.
This talk explains how Ant Group’s frontend engineers built edge‑AI services that run directly in browsers, boosting real‑time performance, preserving privacy, and cutting cloud costs, while showcasing two real‑world cases—pet identification and screen‑break insurance—and detailing the WebGL‑based engine optimizations that lifted device coverage from 30% to 93%.
To sustain Vivo Push Platform’s massive real‑time traffic, engineers re‑architected two Redis clusters, trimmed capacity by 58 %, split clusters, randomized hotspot‑prone keys, and introduced three‑level caching, cutting peak CPU load by 15 %, halving response time and improving overall Redis efficiency during peak loads.
This article shares hard‑won lessons from optimizing fifteen high‑load web applications, highlighting how a tiny DB‑connection leak in a pod probe caused severe slowdown and how fixing it, along with proper load testing, monitoring, and investment in tools and people, can dramatically improve system performance.
The article evaluates software versus hardware video decoding for Baidu’s Android app, presents benchmark data showing surface‑mode hardware decoding’s superior efficiency, identifies compatibility and first‑frame latency challenges, and proposes a monitoring module plus seamless soft‑to‑hard decoder switching to achieve high hardware‑decode usage while maintaining fast startup and low error rates.
This article provides a comprehensive introduction to Elasticsearch, covering data types, Lucene fundamentals, inverted indexes, cluster components, node roles, shard and replica mechanisms, mapping, installation, health monitoring, write path, storage strategies, segment management, refresh and translog processes, as well as practical performance and JVM tuning tips.
This article explains the structure of modern CPU caches, the principles of cache lines, associativity, and coherence protocols, and demonstrates how these hardware details affect program performance through multiple code examples covering loop stride, matrix traversal, multithreading, and false sharing.
This article explains why Redis achieves high query performance by leveraging in‑memory storage, specialized data structures such as SDS, linked lists, dictionaries, ziplists and skiplists, a single‑threaded event loop with I/O multiplexing, and various optimization techniques that avoid common bottlenecks.
This article details China Unicom Software Research Institute's multi‑year journey of adopting Kubernetes‑based service mesh, outlining the evolution from SDK‑driven microservices to sidecar‑based architectures, migration strategies with Baidu, performance optimizations, observability enhancements, and future product roadmaps.
This article examines a real‑world Elasticsearch cluster handling hundreds of terabytes, explains why high‑cardinality aggregations can be slower, and shows how setting execution_hint=map and tuning doc_values dramatically improves aggregation performance for ultra‑high‑concurrency workloads.
Lazada's H5 container team tackled severe first‑screen delays on high‑traffic sales events by building a monitoring system, introducing pre‑heat and pre‑render techniques, and adopting server‑side rendering, ultimately cutting load times by up to 50% and dramatically improving user conversion across Southeast Asia.
Lazada’s Southeast Asian e‑commerce app achieved major performance gains by reorganizing startup tasks, merging launch pages, implementing full‑link dynamic prefetch, incremental cart updates, checkout pre‑rendering, and request merging, resulting in up to 250 ms faster launches and significant reductions in network latency across Android and iOS.
This article details how ByteDance’s ClickHouse UBA edition improves dictionary encoding for low‑cardinality columns by redesigning the Part‑merge process, introducing a single‑dictionary merge, and implementing an automatic fallback for high‑cardinality columns, resulting in significant storage savings and query‑performance gains across large‑scale applications.
Shopee Games chose the Egret engine for its lightweight 2D performance on low‑end Southeast Asian devices, built a custom toolchain with texture‑atlas and draw‑call optimizations, trimmed unused modules, migrated the project to a modern npm‑based workflow with Webpack, linting, testing and CI/CD, and created scaffolding utilities that now let developers treat H5 game development like any frontend project, speeding iteration and improving code quality.
This article provides a comprehensive overview of Tencent's self‑developed MySQL kernel TXSQL, covering its evolution, overall architecture, columnar storage engine, instant DDL capabilities, enterprise‑grade features, high‑availability mechanisms, performance optimizations, and the rigorous development and testing processes behind the product.
This article presents a comprehensive technical overview of Alluxio, covering its role as a distributed data orchestration layer for AI workloads, core features such as caching and unified namespace, performance challenges in large‑scale machine‑learning pipelines, and the extensive optimizations and testing performed at Tencent to achieve high throughput and scalability.
During 2021, the service‑mesh ecosystem—centered on Istio—matured with performance optimizations, protocol extensions, diverse deployment models, eBPF integration, and a growing community of tutorials, conferences, and open‑source projects, solidifying its role as essential cloud‑native networking infrastructure.
This article details Industrial and Commercial Bank of China's comprehensive MySQL governance framework, covering risk identification, prevention strategies, a four‑step methodology, automated quality gates, production‑level monitoring, SRE management, and future visions for rapid incident detection and self‑healing.
This article explains why MySQL queries become slow, how to enable and analyze slow‑query logs, and provides practical steps for index, SQL, schema, optimizer, and architecture optimizations to keep performance strong even with millions of rows.
The paper details a hundred‑billion‑scale real‑time monitoring system, outlining a layered architecture from collection to alerting, comparing Oceanus + Elastic Stack and Zabbix + Prometheus + Grafana solutions, and showing how targeted optimizations in stream processing and Elasticsearch achieve scalability, low latency, and significant cost savings.
This article describes Ctrip's cloud‑native Service Mesh deployment, detailing its multi‑IDC high‑availability design, fault‑scenario analysis, xDS push metrics, event‑handling optimizations, cold‑start improvements, and progressive canary release strategies to ensure reliable, scalable service traffic management.
This article details how a company migrated its reporting architecture to TiDB, leveraged TiFlash for HTAP workloads, upgraded to version 4.0 with new tools like DM and TiUP, achieved up to four‑fold query speedups, and shares the pitfalls and optimization tips learned along the way.
This article provides a comprehensive, step‑by‑step guide on why and how to split monolithic applications, prepare for migration, execute database vertical and horizontal sharding, handle schema changes, ensure consistency, and maintain stability during and after the split, enriched with real‑world pitfalls and solutions.
This article details the challenges Ctrip faced with Istio control‑plane performance at massive scale and explains the systematic methodology, concrete optimizations—including O(n²) to O(n) patching, Merkle‑tree based success‑rate metrics, and startup improvements—that reduced xDS push latency from minutes to seconds while enhancing reliability.
ByteDance’s internal Presto platform, supporting nearly one million daily queries across ad‑hoc, BI visualization, and near‑real‑time analytics, achieves high performance and stability through SparkSQL compatibility, multi‑Coordinator architecture, dynamic routing, adaptive query cancellation, History Server, materialized views, and a dedicated Hudi connector.
This article introduces the Project Matrix initiative that re‑examines and restructures Spark‑ML linear models, detailing the background of Spark‑ML usage at JD, the performance‑focused optimizations such as blockification and virtual centering, and outlines upcoming work to further improve scalability and accuracy.
This article shares practical methods for continuously improving technical abilities and boosting work performance despite long‑hour 996 jobs, covering curiosity‑driven problem hunting, deep research, hands‑on verification, applying results, and disciplined time‑management techniques.
