This article explains why using LIMIT with large offsets slows MySQL queries, analyzes the execution flow, and presents four practical optimization techniques—including subqueries, INNER JOIN, bookmark (tag‑record) method, and BETWEEN range scans—backed by real‑world performance data and code examples.
This article provides a comprehensive overview of DDR memory, explaining its basic architecture, the dynamic read/write process, essential commands, timing parameters, performance considerations, and practical tips for debugging and optimization in modern systems.
This article examines Builder.io's performance optimization case, reviews the evolution of lazy‑loading techniques in front‑end development, and explains how Qwik implements ultra‑fine-grained lazy loading and prefetching to eliminate costly hydration and improve page interactivity.
This article demonstrates how the Taichi library can speed up a Python prime‑counting program by roughly two hundred times compared to the unoptimized version and five times faster than using Numba, providing clear code examples and performance results.
The article explores lesser‑used Java Stream APIs—reduce, the matching trio (allMatch, anyMatch, noneMatch) and flatMap—showing practical examples, performance impacts of operation ordering, readability trade‑offs, and when to prefer streams or traditional loops, even mentioning parallelStream considerations.
The first release candidate of PHP 8.2.0 (RC1) has been published, introducing trait constants, a new Random extension, enhanced JIT support, updated MIME types for the built‑in web server, memory optimizations, new socket and Curl options, additional ZipArchive methods, and several performance and cross‑platform improvements.
This guide explains why regular MySQL health checks are crucial and introduces four open‑source performance tools—mysqltuner.pl, tuning-primer.sh, pt‑variable‑advisor, and pt‑query‑digest—detailing how to download, run, and interpret their reports to optimize configuration, detect bottlenecks, and improve security.
This article provides a comprehensive guide to implementing and testing Go Redis Set APIs—including SAdd, SCard, SIsMember, SMembers, SRem, SPop, and SPopN—complete with code examples, a unified test suite, and a performance benchmark that reports QPS and execution time.
This article summarizes a series of technical Q&A from an OSChina event covering Linux networking fundamentals such as port limits, network namespaces, TCP connection handling, C10K problem, packet loss troubleshooting, TCP memory usage, high CPU causes, useful monitoring tools, kernel parameters, and practical socket examples.
The article explains how the Intersection Observer API offloads visibility detection from scroll events to improve performance, details its creation, options, callback structure, and entry data, and describes a polyfill that mimics native behavior using event listeners, mutation observers, and geometric calculations for broader browser support.
Cloudflare has swapped Nginx for its home‑grown Rust proxy Pingora, claiming trillion‑request daily capacity, up to 70% lower CPU and memory usage, faster TTFB, dramatically higher connection reuse, and improved security, while powering its CDN, Workers, and other services.
This article examines the open‑source memory cache Dragonfly, compares its performance and architecture against Redis through detailed benchmark results, discusses Redis’s response and design principles, and provides reproducible test configurations and command lines for both systems.
This article explores how PureComponent and shouldComponentUpdate address unnecessary re‑renders in class components, compares them with functional components and hooks, and provides practical techniques—including React.memo, useCallback, setState updater functions, and refs—to optimize rendering performance in modern React applications.
This article details how Meituan's Kafka platform, serving over 15,000 machines and handling petabytes of daily traffic, faces read/write latency, slow nodes, and large‑scale cluster management challenges, and describes a series of application‑layer, system‑layer, and operational optimizations—including disk balancing, migration pipelines, fetcher isolation, consumer async, SSD caching, isolation strategies, full‑link monitoring, lifecycle management, and TOR disaster recovery—to improve performance and reliability.
Xiaohongshu refactored Android’s AsyncLayoutInflater by introducing a configurable thread‑pool, safe object pools, and a singleton ViewCache, eliminating ArrayMap and LayoutInflater lock issues, which together yielded over 20% faster cold‑starts and page loads, demonstrating significant performance and business benefits.
When a Linux server’s load spikes, you can quickly pinpoint CPU, memory, disk I/O, and network bottlenecks by running ten concise commands—uptime, dmesg, vmstat, mpstat, pidstat, iostat, free, sar, and top—each providing specific metrics for rapid troubleshooting.
This article explains why Redis is used as a cache, outlines common consistency problems such as cache miss, penetration, and avalanche, compares cache‑aside, read‑through, write‑through and write‑behind patterns, and presents practical solutions like delayed double‑delete, retry mechanisms and binlog‑based asynchronous deletion to maintain eventual consistency between cache and database.
This article examines the challenges of achieving view‑level background blur on Android, compares four implementation strategies—including open‑source libraries and custom AOSP modifications—analyzes their performance, compatibility, and visual quality, and provides code examples and practical recommendations for developers.
StarRocks is a high‑performance MPP database whose simplified FE/BE architecture, fully vectorized engine, and CBO optimizer enable fast multi‑table joins, while its partition‑bucket‑tablet storage model supports real‑time metric services and dashboard migrations, accompanied by practical troubleshooting guidance and upcoming enhancements.
This article details Baidu's comprehensive approach to improving Android app launch performance by evaluating existing tracing tools, selecting Perfetto, developing a Gradle-based automatic instrumentation plugin, handling trace collection and analysis with Trace Processor, and implementing automated detection of regressions, lock contention, and method-level CPU and wall‑time degradations.
The nightly service latency spikes were traced to major page faults triggered by the category client’s lazy mmap of a 1.7 GB off‑heap store during package switches, not to GC or CPU throttling, and were resolved by upgrading to a version that pre‑loads the store in 64 MB chunks with brief pauses.
This article examines why encrypted data hinders fuzzy queries, categorizes three implementation approaches—from naive in‑memory decryption to conventional database tricks and advanced algorithmic solutions—evaluates their security, performance, and storage impacts, and provides practical references for real‑world systems.
The article outlines a simple, cost‑effective backend architecture for handling up to 20 million daily health‑status queries across a billion users, detailing data storage, sharding by ID, memory requirements, load handling, and redundancy strategies, while noting practical limitations and promotional notes.
This article investigates MySQL's recommendation against using UUIDs as primary keys by creating three tables (auto‑increment, UUID, and random snowflake IDs), benchmarking insert speeds with Spring Boot/JdbcTemplate, analyzing index structures, and discussing the advantages and drawbacks of each key strategy.
This article outlines Alibaba Cloud Redis development guidelines, covering key naming conventions, value design, command usage, client configuration, lifecycle management, and related tools, with practical examples and code snippets to help avoid common pitfalls and improve performance.
This article compares Java object serialization approaches—including default Serializable, explicit SelfDescribingMarshallable, and trivially copyable techniques—by benchmarking them with JMH and Chronicle Queue, showing that explicit serialization is roughly twice as fast as default and trivially copyable is over ten times faster.
This article examines the puzzling 'prevent gc' comment and Thread.sleep(0) call in RocketMQ’s source, explains how it leverages JVM safepoint mechanics to trigger garbage collection, discusses counted vs uncounted loops, and demonstrates practical code modifications to improve performance.
This article explains why enabling method breakpoints in IntelliJ IDEA dramatically slows Java application startup in debug mode, explores the JVM and JPDA mechanisms behind the slowdown, and provides practical steps and configuration tweaks to avoid or mitigate the performance hit.
This article explains how the Cube rendering engine and its FocusEngine enable small, fast, and resource‑efficient large‑screen mini‑programs on low‑end OTT devices, covering hardware constraints, core focus components, implementation details, performance optimizations, development experience, and future directions.
This article explains the architecture, module composition, thread model, and performance optimizations of the Cube mini‑program rendering engine, highlighting its lightweight size, cross‑platform support, advanced CSS capabilities, QuickJS integration, and real‑world use cases on TV and POS devices.
This article describes a comprehensive LLDB‑plugin based monitoring framework that intercepts Xcode’s Script Bridge API to measure and report precise debugging latency for Swift projects, outlines the implementation details, timing scenarios, data reporting, and includes a recruitment notice for the ByteDance client infrastructure team.
This article explains Redis 6.0's shift from a single‑threaded network model to threaded I/O, outlines the performance reasons behind the change, describes the new execution flow, and provides step‑by‑step instructions for configuring multithreading in redis.conf.
The article investigates frequent HttpClient connection‑pool exhaustion caused by excessive CLOSE_WAIT sockets in a high‑traffic Java service, explains the underlying TCP and HttpClient mechanisms, presents diagnostic commands and code snippets, and offers concrete configuration and architectural recommendations to prevent similar issues.
This article presents twenty practical MySQL optimization techniques—from selecting specific columns and avoiding OR in WHERE clauses to proper indexing, join strategies, and using EXPLAIN—helping developers improve query speed, reduce resource consumption, and write more maintainable SQL code.
This article investigates why upgrading the Feishu Android project's CI build environment from JDK 8 to JDK 11 caused a 50% increase in build time, analyzes JVM, Gradle, and Docker interactions, and presents concrete JVM flag and environment‑variable solutions that restored parallelism and reduced build duration back to pre‑upgrade levels.
An in‑depth comparison shows that while Dragonfly claims to be the world’s fastest memory cache, Redis 7.0 often delivers higher throughput and lower latency across multiple benchmark scenarios, highlighting architectural trade‑offs and the potential for future feature adoption from Dragonfly into Redis.
After a recent deployment a backend update operation became extremely slow, revealing RPC retry, thread‑pool misconfiguration, and row‑lock contention between services, which were analyzed and resolved through thread‑pool tuning and transaction separation.
An in‑depth guide explains why using MyBatis’s foreach for bulk inserts can cause severe performance degradation, especially with large tables, and demonstrates how switching to ExecutorType.BATCH or limiting batch sizes to 20‑50 rows dramatically reduces insertion time from minutes to seconds.
This article details a five‑month performance project that reduced the heart‑beat app’s average launch time from over eight seconds to around four seconds by dissecting cold‑start phases, optimizing Application initialization, streamlining the launch chain, and applying home‑page, lock, and monitoring strategies.
This article dissects why a well‑tuned Kafka cluster can process up to 20 million messages per second, examining producer batching and custom protocols, broker page‑cache, file layout and zero‑copy techniques, as well as consumer group strategies that together unlock its high throughput.
This article compares the open‑source memory cache Dragonfly, created by a former Google and Amazon engineer, with Redis 7.0, analyzing benchmark results, architectural differences, and community reactions to determine which system delivers superior performance and scalability.
MapStruct dramatically outperforms Spring's BeanUtils for Java object mapping by generating compile-time code that avoids reflection, achieving roughly ten-to-thirty times faster conversions—even across dozens of fields—while requiring only simple PO/Entity definitions and optional @Mapping annotations, making it a superior, low-learning-curve replacement.
This article details a step‑by‑step optimization of a Java backend that imports massive Excel files into MySQL, covering library switches, data caching, batch INSERT statements, and parallel stream insertion to cut processing time from minutes to under two minutes.
This article explains why Kafka can handle up to 20 million messages per second and 600 MB/s throughput by detailing producer batching and custom protocols, broker page‑cache, file layout and zero‑copy techniques, as well as consumer group strategies for efficient message consumption.
This guide surveys more than a dozen Linux command‑line utilities—including nethogs, nload, iftop, slurm, netstat, and many others—explaining their unique features, typical use cases, and installation commands so administrators can effectively monitor bandwidth, connections, and overall network health.
This article compares the MySQL binlog parsing tools my2sql and binlog2sql, detailing their parameters, performance benchmarks on 2 GB binlog files with varying thread counts, feature differences, usage constraints, and provides practical guidance for selecting the appropriate tool.
When trying to change a table’s primary key from int to bigint, a 500‑million‑row MySQL table required data archiving and a bulk MODIFY, but parallel DELETE statements caused lock‑wait timeouts; experiments reveal that under REPEATABLE READ the range scan locks the boundary row, leading to contention.
This article defines software performance as executor + (process + IO + computation), explains when performance should be considered, and presents a comprehensive framework of executor, I/O‑type, and generic optimizations—including concurrency, sharding, caching, compression, async, batching, and algorithmic improvements—supported by cross‑domain case studies.
The jQuery 3.6.1 maintenance update introduces infrastructure improvements, a focus‑related bug fix, significant performance gains for jQuery.trim, and removes outdated links, while also preparing the library for native String#trim support in older browsers.
This article explains the four key techniques—page‑cache usage, sequential disk writes, zero‑copy transfers, and partitioned segment indexing—that enable Kafka to achieve exceptionally high write performance, detailing how each mechanism reduces latency and maximizes throughput.
Alibaba's Big‑Taobao team describes how shifting data‑assembly logic to front‑end Function‑as‑a‑Service transformed engineers into business partners, cut integration costs, boosted development speed, and required careful cost‑benefit analysis, incremental migration, traffic safety measures, and a low‑code orchestration platform to ensure stable, scalable adoption.
This article explains the mmap system call’s core concepts, mapping process, kernel mechanisms, differences from traditional file I/O, performance benefits, and practical usage details, helping readers understand how memory‑mapped files work in Linux and why they’re advantageous for high‑performance applications.
The article explains why MyBatis foreach‑based batch inserts can cause extreme CPU usage and long execution times, demonstrates how combining multiple VALUES into a single INSERT or using ExecutorType.BATCH dramatically improves performance, and provides practical code examples and sizing recommendations.
This guide explains what CPU context switches are, how to monitor them with vmstat and pidstat, analyzes a sysbench‑generated load case, examines interrupt sources, and provides practical thresholds and troubleshooting steps to improve Linux system performance.
This article explores the fundamentals and practical benefits of buffering in Java applications, covering file I/O streams, Logback asynchronous logging, and Kafka producer batching, while providing code examples, configuration tips, optimization strategies, and cautions to avoid data loss.
A comprehensive AWS benchmark reveals that RedisJSON dramatically outperforms MongoDB and ElasticSearch across isolated writes, isolated reads, mixed workloads, ops‑per‑second, and latency, maintaining sub‑millisecond response even under high update rates, while the other databases suffer significant slowdowns.
This article examines how modern SSDs demand higher data rates and IOPS, compares the SATA, SAS, and PCIe storage interfaces, and discusses performance trade‑offs, compatibility considerations, and latency factors to help choose the most suitable interface for a given system.
This article examines how excessive logging, AsyncAppender queue saturation, JVM reflection optimizations, and Lambda class loading can cause thread blocking in Log4j2, analyzes root causes with code examples, and provides practical guidelines and best‑practice configurations to prevent performance degradation in high‑throughput Java services.
This article examines why encrypted data is unfriendly to fuzzy queries, categorizes three implementation approaches—naïve, conventional, and advanced—and evaluates their security, performance, and storage trade‑offs while providing practical code examples and reference resources.
This article introduces a curated list of Linux command-line utilities—including nethogs, nload, iftop, slurm, netstat, and more—explaining their unique features, typical use cases, and installation commands to help system administrators and developers effectively monitor network traffic and bandwidth.
This article presents practical Python performance optimization techniques—including using local variables, minimizing function calls, pre‑computing loop conditions, preferring direct iteration, employing generator expressions, compiling code objects, and structuring modules—to improve speed and reduce memory usage.
The article systematically analyzes why backend APIs become slow—covering database slow queries, complex business logic, thread‑pool misconfiguration, lock design flaws, and machine problems—and presents practical Java‑based optimization techniques such as pagination fixes, index tuning, multithreading, proper thread‑pool sizing, lock refinement, caching, and asynchronous callbacks.
The article examines how design patterns serve as intermediate code conventions, discusses their performance impact, class‑explosion risk, team skill gaps, project environment, and agile constraints, and offers practical advice for balancing code quality with business realities.
This article explains how adding a method breakpoint in IntelliJ IDEA can increase Java application startup time by thousands of percent, explores the JVM and JPDA reasons behind the slowdown, and provides practical steps to remove or mitigate the performance impact.
Facing heavy GC overhead and slow model inference in a Java compute service, the authors explore using Java Native Interface to offload file loading and regression calculations to C++, detailing environment setup, Maven integration, native method implementation, common pitfalls, performance testing with JMH, and achieving up to 80% latency reduction.
The article shares a senior architect's experience improving a legacy multi‑service web system's performance by 3000% by fixing a DB connection leak, adopting tail‑latency metrics, and investing in load testing, monitoring, logging, and dedicated SRE resources.
This article explains how to collect comprehensive MySQL performance metrics—including connections, buffer cache, locks, statement counts, throughput, server variables, and slow‑query analysis—using only MySQL's native SHOW commands, providing a fast, low‑overhead monitoring solution.
This article shares a real‑world case study of how a hidden database‑connection leak in a pod’s health‑check caused severe latency, and outlines four key lessons on performance metrics, testing, legacy system maintenance, and the critical impact of every line of code.
This article explains how MySQL processes LIMIT with an offset, why large offsets cause slower queries and deep‑pagination problems, and presents practical optimization techniques such as id‑based pagination, sub‑queries, and index‑aware strategies to improve performance.
A backend engineer shares a step‑by‑step case study of diagnosing a 20‑second batch‑score query API, then applying three optimizations—index tuning, multithreaded execution, and request‑size limiting—to bring the response time down to under 500 ms.
From a front‑end developer’s viewpoint, this article examines Rust’s learning curve, pros and cons compared with JavaScript/TypeScript, performance benchmarks, multithreading, testing, and real‑world application scenarios such as tooling, WebAssembly, and desktop clients, while also providing learning resources and reference links.
This article details how a hidden 100 ms latency in a Spring Boot channel service was identified and resolved by tracing request flow, using the Arthas Java diagnostic tool, and ultimately fixing a Tomcat‑embed bug caused by Swagger‑ui resources.
This article provides a detailed, line‑by‑line analysis of a chart comparing FPGA and ASIC across dimensions such as upfront costs, unit cost, time‑to‑market, performance, power consumption, field updates, density, design flow, granularity, verification needs, upgrade paths, and additional features, helping engineers decide which technology best fits their high‑performance AI workloads.
An in‑depth analysis of a production incident where HttpClient connections accumulated in CLOSE_WAIT, covering symptom identification, TCP state inspection, root‑cause debugging, HttpClient pool internals, and practical configuration tweaks to prevent similar outages.
Facing massive AP‑heavy analytics workloads, 58 Group evaluated TiFlash, ClickHouse and StarRocks, chose StarRocks for its superior write/read performance and ease of operation, built internal tools for topology, cluster, Kafka import and slow‑SQL management, and migrated to cloud‑native Docker deployments, achieving up to 90% query speedup and massive data‑volume reductions.
This article introduces Arthas, Alibaba's open‑source Java diagnostic tool, explains its installation via wget and java‑jar, and provides detailed usage of common commands such as stack, jad, sc, watch, trace, jobs, logger, dashboard, and redefine for real‑time JVM monitoring and debugging.
This comprehensive guide explains why TiDB upgrades are essential, compares offline, forced, and smooth upgrade methods, outlines detailed step‑by‑step procedures, highlights advantages and risks, and shares practical check‑points and real‑world case studies to help DBAs ensure a safe, performant migration.
This article explains why oversized logs hurt system performance, presents three practical log‑shrinking strategies—printing only necessary logs, merging duplicate entries, and simplifying content—illustrates them with Java code snippets, and evaluates their impact through a real‑world case that cuts daily log volume from 5 GB to under 1 GB.
This article examines a MySQL InnoDB pitfall where a unique index fails to block duplicate rows—especially when indexed columns contain NULL values or when logical deletion is used—and presents practical solutions such as adjusting index columns, adding timestamps, delete status counters, hash fields, and proper bulk‑insert strategies.
This article outlines the major new features of PHP 8 and PHP 7, explains why PHP 7 outperforms PHP 5, describes the execution flow, web‑request lifecycle, underlying architecture, data structures, garbage‑collection mechanism, and the PHP‑FPM master‑worker model with optimization tips.
This article explains Java object pooling concepts, introduces the Commons Pool 2 library and its usage with Redis's Jedis client, compares it with the high‑performance HikariCP database connection pool, and presents JMH benchmark results demonstrating significant throughput gains, while also covering configuration parameters and interview questions.
Fury is a JIT‑based native multi‑language serialization framework that automatically handles shared and cyclic references, offers zero‑copy support, and delivers 20‑200× speed improvements over existing solutions, making it a high‑performance drop‑in replacement for Java, Python, Go, and C++ serialization needs.
This article chronicles the architectural evolution of China Telecom's Tianyi Account gateway from its initial Zuul‑based 1.0 version through successive upgrades to a Kong‑powered 2.0 and 3.0 system, highlighting performance bottlenecks, technology selections, plugin development, CP/DP separation, cloud‑native deployment, and the resulting high‑concurrency capabilities.
Go 1.19 has been released, bringing toolchain, runtime, and library enhancements such as refined generics, improved documentation comments, a formally aligned memory model with new atomic types, security changes to os/exec, soft memory‑limit garbage collection, a new unix build constraint, and numerous performance optimizations, all while preserving full Go 1 compatibility.
Go 1.19, released on August 2, 2022, brings extensive toolchain, runtime, and library enhancements—including memory model refinements, LoongArch support, improved doc comments, new atomic types, security changes to os/exec, GC soft memory limits, and performance upgrades—while maintaining full backward compatibility with Go 1.
This article explains why logging is essential for maintainable software, outlines best‑practice guidelines for readable, performant, and secure log output, and details Log4j components, log level hierarchy, and practical macro examples for each level.
ByteDance engineers have contributed a series of Linux kernel patches that enable uncompressed kernels and other optimizations, reducing Kexec reboot latency from roughly 500 ms to just 15 ms, thereby minimizing server downtime for their high‑traffic services like Douyin.
This article summarizes the author's experience optimizing backend interface performance, covering common causes such as slow MySQL queries, complex business logic, thread‑pool and lock design flaws, and offers practical solutions including pagination tricks, indexing, caching, multithreading, and asynchronous callbacks.
When a Linux server suddenly spikes in load, this guide shows how to pinpoint the root cause within a minute by running ten key commands that reveal CPU, memory, disk I/O, and network metrics, enabling rapid performance troubleshooting.
This article outlines seven major areas of Java backend performance optimization—including reuse, computation, result‑set, resource‑conflict, algorithm, efficient implementation, and JVM tuning—explaining concepts, techniques, and practical examples to help developers improve application speed and resource utilization.
This article walks through a real MySQL deadlock scenario, explaining the four necessary deadlock conditions, illustrating how Next‑Key locks on non‑unique indexes cause gap locks, and presenting step‑by‑step analysis, experimental verification, and practical recommendations to avoid such deadlocks in production systems.
Since JDK 8, HotSpot’s garbage collectors have undergone extensive refinements—introducing G1 as default, adding ZGC and Shenandoah, and optimizing Parallel GC—resulting in notable gains in throughput, reduced pause latency, and lower auxiliary memory usage, as demonstrated by SPECjbb2015 benchmarks across JDK 8 to JDK 18.
This guide explains how to monitor Linux CPU usage, scheduler queues, lock contention, network I/O, and disk I/O using graphical tools and command‑line utilities such as gnome‑system‑monitor, vmstat, pidstat, nicstat, and iostat, with syntax, parameters, example outputs, and installation steps.
This article presents ten useful Python techniques—including string joining, list comprehensions, enumerate, zip, itertools, Counter, dictionary creation, generators, multiple return values, and the sorted function—each explained with clear examples and performance comparisons to help developers write cleaner and faster code.
Using thread pools in Java reduces resource consumption, speeds up response, and enhances system stability, while the optimal thread count depends on task type—IO‑bound tasks use roughly twice the CPU cores, CPU‑bound tasks match the core count, and mixed workloads follow a calculated formula.
This article discusses practical methods for reducing log volume in backend applications, including printing only necessary logs, merging log entries, simplifying messages, and applying these techniques in a real-world Java case to shrink daily log size from several gigabytes to under one gigabyte while preserving debugging capability.
This article explains that Spring Boot 3.0 requires Java 17 and Spring Framework 6, outlines the 14 new features of Java 17, and presents benchmark results showing Java 17's performance advantages over Java 11 and Java 16 with different garbage collectors.
The article outlines practical approaches to monitor and resolve performance bottlenecks in backend applications, covering database slow‑query logs, interface latency, message‑queue backlogs, segmentation timing, caching, batch calls, multithreading, and database tuning techniques such as indexing and transaction isolation.
The article explains the performance problems of using OFFSET and LIMIT for large‑scale database pagination, demonstrates their inefficiency with full‑table scans, and presents a faster cursor‑based alternative that relies on storing the last primary key and using LIMIT instead of OFFSET.
To overcome Go’s per‑connection goroutine memory and GC limits at massive scale, the article designs a three‑layer Reactor‑based network library that uses epoll/kqueue, lock‑protected file descriptors, zero‑copy buffers, and load‑balanced sub‑reactors, enabling stable operation with up to a million concurrent connections.
