Ctrip Flight team replaced polling with Server-Sent Events (SSE) across Web and mobile (Native + React Native), detailing implementation layers, optimizations like pre-connection, pre-request, caching, and Brotli compression, achieving 11.8% faster multi-batch loading and enabling heterogeneous batching for API granularity conflicts.
Ctrip engineers detail DevMaster, an end-to-end engineering system that connects structured requirements, application-level splitting, spec-driven development with Harness quality gates, and TestPilot test generation to close the loop between AI-generated code and production-grade delivery.
Ctrip built a multimodal understanding platform for hotel images and videos that turns model outputs into versioned, reusable assets, enabling image search, cover selection, quality inspection, and video understanding through a unified data foundation, vector retrieval, model orchestration, and evaluation-gray-release loop.
This article provides a comprehensive tutorial on Apollo configuration center, covering its core concepts (application, environment, cluster, namespace), client architecture with long-polling and local caching, overall system design with Config/Admin services and Eureka, availability scenarios, hands-on SpringBoot integration with Maven setup, dynamic configuration testing (updates, rollbacks, offline fallback), multi-environment/cluster/namespace usage, and Kubernetes deployment via Docker and YAML manifests.
The article explains how Ctrip leveraged mature large‑language models to overcome traditional data‑product challenges—such as inconsistent metrics and manual attribution—by designing a ChatBI system that combines a multi‑agent framework, memory management, Harness‑driven tool orchestration, and Skill‑based data access, while also evaluating alternatives like Claude SDK and Ali Agent Scope.
This guide details the end‑to‑end process for developers to integrate the Ctrip WenDao (WorkBuddy) skill, covering environment prerequisites, API token acquisition, configuration methods, command‑line usage, response parsing, common scenarios, security considerations, and troubleshooting steps.
This article details Ctrip's end‑to‑end observability platform—covering metrics, logging, and tracing—its architecture, data governance, AIOps capabilities, and practical case studies, while addressing challenges like data volume, alert noise, and metric explosion in a massive micro‑service environment.
This article traces the evolution of Ctrip’s travel product log platform—from a single‑table DB approach to a platform‑wide ES + HBase solution—detailing the challenges of massive data volume, the architectural decisions, RowKey design, write and query flows, and the subsequent extensions that enabled billion‑scale log storage and fast retrieval.
This article presents a comprehensive AI‑driven system for automatically detecting anomalies in over 1,000 Ctrip train‑ticket business metrics and pinpointing their root causes, detailing the background, unsupervised algorithms, detection and attribution pipelines, practical results, and future improvements.
This article details the architecture, component design, DAG execution engine, optimization techniques, and real‑world performance gains of Ctrip's Eagle algorithm strategy platform, illustrating how modular, visualized, and automated workflow management improves development efficiency, stability, and resource utilization for large‑scale recommendation services.
The article details the motivation, design, architecture, key features, deployment workflow, and future roadmap of Ctrip's PGClowcode low‑code platform, which enables rapid visual development of backend pages through a front‑end React UI and a Spring WebFlux backend, addressing common challenges in large‑scale internal tool development.
This article presents an AI‑driven system that automatically detects anomalies in over 1,000 Ctrip train‑ticket business metrics using six unsupervised algorithms and locates their root causes through a hard‑voting ensemble of four specialized methods, demonstrating practical results and future enhancements.
This article presents Ctrip's practical experience of constructing an offline‑online data warehouse, detailing business pain points, objectives, system architecture, component design, data quality measures, and future directions to achieve scalable, real‑time data processing and management.
The article shares a fresh graduate's experience interviewing for offline data‑warehouse roles at companies like Ctrip, Meituan and Alibaba, outlines the common interview pattern, and lists detailed Spark, Hadoop, and data‑warehouse questions used by these firms.
Ctrip’s automated error warning solution for its WeChat mini‑programs provides a comprehensive pipeline that injects build IDs, collects runtime errors via SDK, maps them with source maps, aggregates data in an APM MySQL store, and delivers real‑time alerts across development, testing, and production stages.
The article presents Taro‑CRN, a low‑cost, Taro‑compatible framework built by Ctrip to enable seamless cross‑platform development of mini‑programs and native apps, detailing its architecture, plugins, integration steps, code examples, and suitable use cases for frontend teams.
The article explains why enterprises need real-time data warehouses, compares Lambda and Kappa architectures, describes Ctrip Hotel's Lambda‑plus‑OLAP variant built with Flink and StarRocks, and details practical solutions for ordering, wide‑table generation, and data validation that enable billion‑row, low‑latency analytics.
This article presents Ctrip's backend engineering practices for the Direct Connection Platform, detailing how a sandbox testing environment, automated acceptance, rate‑limiting, and circuit‑breaking mechanisms were introduced to boost supplier onboarding speed and enhance overall system stability.
This article explains the principles of Dynamo‑style leaderless replication, strict and loose quorum, version repair, and how Ctrip Hotel applied these concepts in its multi‑media storage systems Hare and InfoKeeper to achieve high availability and performance.
This article details Ctrip's Data Replicate Center (DRC) architecture, challenges, and solutions for migrating MySQL databases to the cloud, covering read‑only and read‑write scenarios, cost analysis, proxy modules, transaction‑table replication, data isolation, filtering, deployment workflow, and achieved performance and cost benefits.
This article presents Ctrip's Taro cross‑platform solution for mini‑programs, detailing the motivation behind multi‑platform development, the technical selection of Taro 3, the modular plugin architecture, repository management with git submodules, CI/CD pipelines, and the overall benefits for code reuse and development efficiency.
This article details how Ctrip transformed its technical support team into a public TS organization, describing the evolution of its support models, the architecture of its operation system, AI‑driven service accounts, wiki automation, crawler tools, tagging strategies, monitoring dashboards, and future plans to enhance efficiency and user satisfaction.
The article introduces DynamicForm, a React‑driven visual dynamic form configuration system designed to eliminate repetitive form coding, offering drag‑and‑drop layout, rich control types, validation, data binding, and a modular architecture that streamlines component property management on Ctrip’s Lego platform.
This article details Ctrip’s mini‑program collaborative development process, covering modular repository management, automated CI pipelines using GitLab and miniprogram‑ci, and continuous delivery via the MCD platform, illustrating how these practices streamline code integration, testing, packaging, and release for dozens of business lines.
The article outlines Ctrip's multi‑stage architectural evolution, detailing its three‑tier composition of operations, framework, and applications, and examines concrete case studies such as the release system, configuration management, SOA, and a large‑scale User Profile big‑data project.
The article outlines Ctrip's data governance framework, detailing background challenges, metadata construction, cost and quality optimization techniques, data flow improvements, platform modules, health metrics, and concludes with a summary of achievements and future directions.
This article details Ctrip’s large‑scale intelligent workforce scheduling project, covering the business background, constraint‑rich problem modeling, heuristic algorithm selection and benchmarking, performance optimizations through multithreading and distributed computing, and the design of a scheduling platform that delivers high‑quality rosters within minutes.
This article examines Ctrip’s Hermes messaging system, tracing its evolution from a simple Mongo‑based queue to a broker‑centric, MySQL/Kafka hybrid architecture, and explains the design choices, performance optimizations, cluster management via lease‑based meta‑server, and lessons learned for building high‑throughput, low‑latency MQ solutions.
This article describes how Ctrip improved hotel recommendation by iterating from logistic regression to GBDT and deep neural networks, designing continuous and discrete features, adopting multi‑task learning with click and conversion signals, and building a large‑scale distributed DNN training and unified feature‑processing framework to boost model accuracy and engineering efficiency.
This article presents the background, research, design choices, core features, extensions, challenges, and outcomes of Mars, a proxy‑based lightweight frontend mock platform developed by Ctrip's IBU ticket team to improve development efficiency and testing flexibility across devices and environments.
This article shares the Ctrip flight big‑data platform’s journey of evaluating and migrating data storage from Hive, MySQL and Redis to CrateDB, covering performance requirements, query patterns, maintenance challenges, containerization, and production results that reduced interface latency and resource consumption.
This article outlines Ctrip's approach to time series forecasting, covering background, common tools such as factor‑based models, traditional statistical methods like ARIMA, and machine‑learning techniques including tree and neural networks, and shares practical experiences on data splitting, feature engineering, model stability, and evaluation.
This article presents a comprehensive overview of Ctrip's multilingual machine translation platform, detailing demand analysis, system architecture, data pipeline, algorithmic innovations such as task‑space fusion and term‑translation interventions, as well as extensive performance optimizations for low‑resource languages.
This page lists past Ctrip Tech Salon sessions covering backend architecture, API gateway design, serverless services, front‑end engineering, hot‑fix techniques, Electron usage, and data‑collection SDKs, providing links to detailed technical reviews of each practice.
The article details Ctrip’s Flight Assistant redesign journey, outlining the transition from the 1.0 era’s separate flight‑dynamic and flight‑assistant lines through the integrated 2.0 version to the intelligent, user‑centric 3.0 version, and explains the design strategies, GUCDR model, and three key “secrets” employed to improve usability and information overload.
This article presents a comprehensive overview of Ctrip's hotel recommendation system, covering its technical architecture, data processing pipelines, various ranking and embedding models—including FM, Wide&Deep, DeepFM, and FTRL—deployment methods such as PMML and TensorFlow Serving, offline and online evaluation results, and challenges like cold‑start and diversity.
This article shares the practical experience and thinking behind Ctrip's vacation data governance project, covering team efficiency optimization, demand sorting, data domain definition, warehouse layering, unified dimension modeling, metric standardization, and the overall benefits of a centralized data governance framework.
This article describes how Ctrip’s senior test engineer designed and executed SIP‑based load‑testing scenarios with SIPp to evaluate the call‑center system’s capacity, concurrency limits, IVR/PBX overflow handling, and geographic routing accuracy during peak travel periods.
This article describes how Ctrip's ticket order post‑service team identified scaling bottlenecks, selected Redis, designed a two‑level cache architecture, implemented active and passive caching strategies, tackled consistency and failure issues, and applied various optimizations to achieve over 90% cache hit rate and dramatically reduce database load.
This article describes the concept, tag taxonomy, data flow architecture, and Lambda‑based query service design of Ctrip Business Travel's user profiling system, highlighting how batch and real‑time processing with Spark, Flink, Hive, MongoDB and Redis enable precise marketing, risk control and personalized services.
This article describes the motivation, advantages, challenges, overall architecture, and natural‑language‑understanding algorithms of Ctrip’s one‑stop chatbot that consolidates multi‑service line support into a single entry, enabling seamless travel‑related queries, recommendations, and bookings.
This article describes Ctrip's SRE team's journey in adopting chaos engineering, outlining the motivations, roadmap, concrete experiments, platform maturity, and future automation goals to improve system resilience and operational reliability in a large‑scale microservice environment.
This article details Ctrip's practical experience of refactoring its train‑ticket RN module with the Rematch state‑management framework, outlining the background problems of Redux, the three‑phase migration plan, specific challenges such as store compatibility, component decoupling, reuse, and the resulting code‑size reduction and readability improvements.
The article details Ctrip's challenges with multiple AB testing splitters, presents performance gains after migrating to a new splitter, and explains the comprehensive redesign covering overall architecture, interface consolidation, SDK slimming, and a custom distributed cache backend to achieve higher throughput and lower latency.
This article details the evolution of Ctrip's flight ticket data warehouse, describing its historical tech stack, current architecture—including Hive, Presto, ClickHouse, CrateDB, and Flink—data synchronization methods, layer design, quality monitoring, and a real‑time price‑monitoring use case.
This article describes how Ctrip built a unified hotel data intelligence platform, evaluated various database solutions, selected ClickHouse as the primary engine, and implemented performance, high‑availability, and monitoring strategies to handle billions of records and thousands of concurrent queries.
This article outlines Ctrip's flight ticket data warehouse evolution, current big‑data technology stack, data synchronization methods, layered architecture, quality monitoring system, and a real‑time price anomaly detection case, providing practical insights for building scalable, reliable data warehousing solutions.
The article details Ctrip’s Marco Polo content platform, describing its data, algorithm, and functional layers, and explains how AI techniques such as NLP, semantic matching, named‑entity recognition, and image classification are applied to automate product‑centric content mining, article generation, quality rating, and first‑image selection.
This article examines how Ctrip Hotel implements DevOps testing—from business and technical drivers, through continuous delivery challenges, to the design of the Moss platform, its toolchain, data center, and future automation goals—highlighting practical solutions for rapid, high‑quality software delivery.
This article showcases three thoughtful UI design techniques used in Ctrip's flight ticket pages—highlighting how tomorrow's tickets are prioritized at night, price trends replace traditional calendars, and a dynamic bottom filter bar enhances both selection clarity and immersive browsing.
This article details the design goals, architecture, core functionalities, and optimization strategies of Ctrip's user precise marketing system, which leverages RESTful integration, flexible rule-based and machine‑learning models, real‑time monitoring, and AB testing to improve traffic utilization and conversion rates.
This article details Ctrip's hotel data platform challenges with billions of daily updates and near‑million queries, evaluates various storage options, explains why ClickHouse was chosen, and describes the full‑load and incremental pipelines, monitoring, server clustering, and practical tips that enable sub‑second query performance at massive scale.
This article details Ctrip's Order Index (OI) service evolution from version 1.0 to 2.0, describing the original architecture, its limitations, the redesign of data synchronization, job scheduling, and query engines, and the resulting platformization that enables faster, more flexible order aggregation across dozens of business lines.
This article examines the limitations of traditional learning‑to‑rank for Ctrip hotel sorting, introduces reinforcement learning as a remedy, outlines three progressive implementation plans (A, B, C) with algorithm choices and engineering trade‑offs, and presents early experimental findings that demonstrate RL's potential to improve conversion rates.
This article presents an overview of Ctrip's travel recommendation system, detailing its architecture, user‑behavior analysis, product catalog handling, various recall strategies, ranking methods—including machine‑learning models like XGBoost—and future directions toward deeper AI and NLP integration.
This article introduces Ctrip Hotel's automated comparison platform, detailing its motivation, overall architecture, and the four modules—interface, database, tracing, and cache comparison—along with configuration, execution methods, and results, highlighting its impact on testing efficiency and scalability.
This article presents an overview of computer‑vision based OCR in Ctrip's operations, detailing deep‑learning text detection methods for controlled and uncontrolled scenarios, sequence‑based recognition models, training strategies with synthetic data, and performance results, while discussing current challenges and future improvements.
The article outlines Ctrip’s three‑phase technology evolution—from a simple call‑center architecture to layered internet and mobile platforms, and finally to a cloud‑based big‑data and AI‑driven ecosystem—highlighting architectural changes, operational challenges, and strategic lessons for fast‑growing internet companies.
At Ctrip, the CRN framework extends React Native with custom tooling, performance optimizations, modular bundling, lazy loading, preloading, and comprehensive release and monitoring systems, enabling widespread production use across dozens of RN bundles, reducing first‑screen load times by up to 45% and improving stability.
The article presents Ctrip’s Blockchain Service Platform (CBaaS), detailing its motivation, architecture, and the advantages of using Hyperledger Fabric’s modular design—including contract execution, consensus, permission control, and multi‑chain capabilities—to simplify enterprise blockchain adoption and enable secure, data‑rich on‑chain applications.
This article introduces Ctrip's AIOps journey, explaining the AI‑driven operations concept, showcasing typical use cases such as anomaly detection, intelligent fault diagnosis, and resource utilization improvement, and detailing the underlying statistical and machine‑learning algorithms that enable these capabilities.
The article introduces Ctrip’s Model Engine Platform, a comprehensive system that streamlines feature preparation, engineering, model management, and product orchestration to enable fast, reliable real‑time AI model serving across various business scenarios, while addressing common challenges such as manual data handling, offline‑only prediction, and long development cycles.
The article details Ctrip's mature real‑time anti‑fraud platform, describing its big‑data parallel processing, AI‑driven models, device‑fingerprinting, CDNA service, and evolving architecture that together achieve sub‑150 ms decision latency while handling billions of daily transactions.
The article describes the evolution of Ctrip's wireless APM platform from the early UBT-based monitoring to a globally‑oriented, metric‑rich system that processes over 100 billion data points daily using Storm and Elasticsearch, detailing its design, key performance dimensions, data‑volume trade‑offs, and implementation choices.
In 2017 Ctrip reorganized its wireless engineering to adopt a lifecycle‑driven, platform‑based approach, introducing the Mobile Tech Platform (MTP) and Mobile Continuous Delivery (MCD) platforms that unified component services, development frameworks, and automated build‑release pipelines for over 20+ apps, dramatically improving efficiency and quality.
This article recounts how Ctrip’s Service Butler project achieved a one‑month launch by establishing clear division of labor, intensive cross‑team communication, Scrum‑of‑Scrums, War‑Room sessions, and a strong team‑spirit culture, illustrating practical agile management for large‑scale development.
This article details Ctrip's end‑to‑end personalized recommendation system for travel, covering data collection, candidate recall methods, ranking models, feature engineering practices, and future directions, illustrating how millions of users receive tailored travel suggestions.
The article details Ctrip's large‑scale personalized recommendation system, covering data sources, recall strategies, ranking models, feature engineering techniques, and practical insights from industry research to improve user experience in travel services.
This article describes how Ctrip’s data science team applied machine learning models to predict response times and outbound call effectiveness, transforming the hotel order confirmation process into an intelligent, automated workflow that reduces unnecessary calls, improves resource allocation, and boosts overall call‑center efficiency.
This article describes how Ctrip built a container‑cloud platform by dividing infrastructure into "under‑water" (foundations) and "over‑water" (developer‑facing practices), evolving through three architectural stages, and addressing networking, monitoring, resource fragmentation, JVM tuning, Dockerfile management, and a plugin‑based extension model to support DevOps and micro‑service deployments.
This article explores Ctrip’s continuously evolving architecture, detailing its three-layer composition of operations, frameworks, and applications, and examines real-world case studies of its release system, configuration management, SOA, and a massive User Profile big‑data project, highlighting key innovations and lessons learned.
The article recounts Ctrip’s 2017 EGO event where senior engineers shared practical experiences on scaling React Native in their app, detailed the company’s TCP‑based wireless network service architecture, and presented Android plugin refactoring techniques, highlighting technical challenges, solutions, and future directions.
This article explains why Ctrip creates user profiles, describes the product and technical architectures, and details the data collection, computation, storage, high‑availability querying, and monitoring components that power its personalized travel recommendations and services.
This article chronicles Ctrip's multi‑year architectural evolution—covering operations, framework, application layers, publishing system, configuration management, SOA, and the large‑scale UserProfile project—highlighting the motivations, challenges, and solutions that shaped its high‑availability, high‑performance platform.
This article reviews Ctrip's two‑year architecture transformation, describing how the company replaced hardware load balancers with a software‑defined SLB, introduced application‑level grouping, multi‑update mechanisms, health‑check sharing, monitoring, and the TARS release platform to achieve faster, more reliable deployments.
This article describes Ctrip's automated web vulnerability detection system, detailing the shift from active to passive scanning, the distributed architecture using traffic mirroring, message queues, Redis, and MySQL, and the processes for data collection, de‑duplication, scanning, and vulnerability management.
This article describes how Ctrip's information security team redesigned its data platform from a simple RabbitMQ‑MySQL pipeline to a scalable, real‑time and offline big‑data architecture using Kafka, Storm, Hadoop, Spark, and a custom count server, dramatically improving processing capacity and supporting risk‑control operations.
This article presents a comprehensive overview of Ctrip's evolving system architecture, detailing its operational strategies, framework components such as SOA and release systems, and the large‑scale UserProfile big‑data platform, illustrating how each iteration addressed prior challenges while introducing new capabilities.
The article introduces Zhao Xingu's presentation on how Ctrip's wireless R&D team designed and implemented the Mobile Delivery Continuous Platform (MCD) to support hundreds of engineers across multiple product lines, covering integration, testing, release, and operation for efficient mobile app delivery.
The article details how Ctrip’s CTO Gan Quan has leveraged big‑data platforms, deep‑learning algorithms, cross‑screen user tracking, and rapid AB testing to create a real‑time, personalized recommendation engine that shortens travel decision cycles and drives significant revenue growth.
The article details Ctrip CTO Gan Quan’s insights on how the travel platform leverages a comprehensive big‑data infrastructure, AI‑driven algorithms, and real‑time user behavior tracking to deliver personalized travel recommendations, improve conversion rates, and shorten user decision cycles across multiple business lines.
Ctrip built a private container cloud to handle massive seasonal traffic spikes, enabling rapid, automated scaling and shrinking of resources, improving deployment speed, resource utilization, and operational intelligence across more than 20 business units.
This article details Ctrip's engineering practice of building a mobile continuous delivery platform that automates integration, testing, release, and operation of hundreds of native, hybrid, and React‑Native bundles, addressing scalability, efficiency, and quality challenges through message‑driven architecture, bundle splitting, automated testing, and gray‑release strategies.
This article presents the motivation, architecture, functional and interface layers, storage design, and performance evaluation of CFL, a MySQL plug‑in storage engine created by Ctrip to provide a high‑throughput, SQL‑based time‑series database while preserving existing MySQL operational practices.
The article recounts the author’s experience at AAAI‑2017 in San Francisco, summarizes key talks, panels and award‑winning papers, and details Ctrip’s hybrid collaborative‑filtering model with a stacked denoising auto‑encoder that improves recommendation performance and addresses data sparsity.
This article explains Ctrip's flight ticket tracking framework (UBT), detailing client‑side and server‑side event collection methods, the purpose and trade‑offs of each tracking type, metric definitions, data association challenges, common pitfalls, and best practices for reliable data‑driven analysis.
The article explains Ctrip's comprehensive User Behavior Tracking (UBT) system for flight tickets, detailing client‑side (ctm, action, trace, pv) and server‑side tracking, their implementation, data flow, key metrics, common challenges, and best practices for reliable analytics.
Ctrip announced that its AI‑driven recommendation system paper, titled “A Hybrid Collaborative Filtering Model with Deep Structure for Recommender Systems,” has been accepted by the prestigious AAAI conference, highlighting the company’s cutting‑edge deep‑learning techniques, large‑scale deployment, and broader AI innovations in travel services.
This article shares Ctrip's experience of adopting React Native for its train ticket business, covering why RN was chosen, its current state, the Ctrip RN framework, practical implementation details, encountered pitfalls, and various optimization techniques to achieve near‑native performance and rapid iteration.
The article chronicles Ctrip's mobile app architecture evolution—from a simple MVC design in 2011, through a tightly coupled early server model, to a modular API‑Gateway‑based backend, intelligent downgrade mechanisms, multi‑Dex loading, and finally a bundle‑oriented dynamic plugin framework—highlighting the technical challenges, solutions, and lessons learned for large‑scale mobile development.
This article reviews the rapid growth of China’s OTA market, the rise of black‑market threats, and how Ctrip’s security team has iteratively redesigned its risk‑defense architecture—from a .NET‑based real‑time system, through an offline risk‑library, to the integrated Ares platform—highlighting each stage’s strengths, shortcomings, and lessons learned.
Facing rapid growth in China's OTA market, Ctrip's security team outlines the evolution of its risk defense architecture—from the early .NET-based 1.0 system, through the 1.5 risk‑library, to the 2.0 Ares platform—detailing technical choices, strengths, shortcomings, and future directions for combating black‑market abuse.
This article explores how Ctrip integrates data analysis with product design, presenting two detailed case studies— the evolution of the homestay channel and the optimization of the Guesthouse app order detail page— to illustrate the role of data, user research, and A/B testing in driving user‑centered product improvements.
The article introduces Ctrip DAL, an open‑source Java/C# database access framework that supports sharding across MySQL and SQL Server, provides a code generator and client library, and helps large enterprises manage multi‑team, multi‑database environments while ensuring scalability, consistency, and security.
At Ctrip, the Netranger system leverages software-defined networking to automate network operations, from server port provisioning and switch onboarding to configuration visualization, security zone management, IP path discovery, and syslog alert analysis, addressing the challenges of a traditional, high‑change data‑center network.
This article presents Hong Guangming's detailed overview of Ctrip's payment engine reconstruction, covering the legacy system's challenges, the new split‑service architecture, channel‑specific services, settlement reconciliation, and the multi‑layered strategies employed to achieve high availability and operational stability.
Ctrip DAL is an open‑source, multi‑language database access framework that supports sharding, code generation, and cross‑team management, enabling large‑scale enterprises to protect existing investments while rapidly scaling their data layer across MySQL and SQL Server.
This article examines how Ctrip integrates data analysis into product design through two detailed case studies—its homestay channel evolution and the Kezutong app order‑detail redesign—highlighting tools, methodologies, and results that illustrate data‑driven decision making in product management.
This presentation details Ctrip's payment engine reconstruction, describing the limitations of the legacy system, the new service‑oriented architecture separating payment planning and execution, channel‑specific services, the reconciliation system, and the operational measures taken to achieve high payment availability.