Meituan’s Meipai service, serving over 150 million monthly users with a hybrid private‑public cloud architecture, spent three years building a comprehensive, three‑dimensional monitoring platform that unifies client‑to‑server metrics, alerts and reporting to ensure resilient, scalable operations and rapid business growth.
The Hermes project at Meituan‑Dianping re‑engineered its overseas vacation front‑end by separating front‑ and back‑ends, modularizing code with ES6, building a Vue‑based reusable component library, automating builds and testing, and optimizing page‑load performance through CDN caching, monitoring, and offline strategies, dramatically boosting developer productivity and user experience.
This guide walks you through analyzing the Meituan Waimai app's comment section, uncovering the AJAX requests that load data, constructing the proper URL pattern, looping through pages to fetch JSON, and storing the results in files or a database for further analysis.
Meituan Dianping’s technical salon showcased its automation journey—detailing database automation platforms, service‑tree management, Puppet‑based web control, and a CMDB case study from Shanghai Zhaogang—highlighting rapid iteration, standardization challenges, and the evolution of automation from tools to core operational practice.
Meituan‑Dianping migrated its 2,000‑node HDFS cluster to Federation by fixing ViewFs compatibility, simplifying mount points, leveraging FastCopy for massive data moves, improving token handling, and automating split‑workflow steps, thereby overcoming single‑NameNode bottlenecks and providing a practical blueprint for large‑scale Hadoop deployments.
The Meituan Android team traced a surge of OutOfMemoryError crashes to duplicated drawable caches created by the AppCompat vector‑from‑resources flag, reproduced the condition on OPPO N1 devices, used heap‑dump analysis to pinpoint oversized byte‑array allocations, and eliminated the issue by disabling the flag, dramatically cutting OOM incidents.
This article details Meituan Cloud's journey from virtual machines to Docker containers, explaining why the shift was needed, the technical challenges faced, the implementation steps, and the measurable performance and cost benefits achieved through containerization.
Meituan‑Dianping leverages deep‑learning models—including ClickNet for semantic search matching, an AlexNet‑based image‑quality ranker, and a Faster‑RCNN/FCN‑driven OCR pipeline—to personalize results, select attractive POI images, and extract text, achieving higher click‑through rates, conversions, and operational efficiency across its O2O services.
The article reviews Meituan‑Dianping’s tech blog evolution since 2013, highlights the ten most‑read posts of 2016 across backend, big data, mobile, and AI, and outlines emerging focus areas such as testing, cloud storage, and algorithmic innovations for the coming year.
Meituan’s group‑buy order system was refactored from a monolithic, single‑point‑failure design into a sharded, micro‑service architecture with connection‑pool middleware, dual‑cluster databases, and robust messaging, delivering stable, high‑performance handling of millions of daily orders even during peak events.
Meituan’s food‑delivery order platform evolved from a simple monolithic prototype in 2013 to a distributed, highly available service that now handles tens of millions of orders daily, using asynchronous processing, sharding, multi‑data‑center deployment, and automated operations to achieve sub‑150 ms transaction latency and 99.9999% uptime.
The talk details Meituan‑Dianping’s end‑to‑end delivery big‑data system, highlighting mobile‑ and local‑centric usage, a two‑stage forecasting pipeline that combines autoregressive baselines with a boosting‑based multiplier model, layered log‑collect‑process‑serve architecture, sophisticated feature engineering, real‑time inference, and strategies for logistics constraints and cold‑start merchants.
The Meituan‑Dianping tech salon detailed how food‑delivery big data drives full‑chain marketing, using RFM‑based user segmentation, rich demographic and behavior profiles, churn‑prediction and survival models, and scenario‑driven expansion tactics to acquire, retain, and grow customers across the order lifecycle.
This article presents Meituan Dianping's data‑driven precise marketing architecture, detailing a layered pyramid system, user profiling, budget monitoring, and two real‑world cases—potential user mining and a smart coupon engine—demonstrating how big‑data techniques improve marketing efficiency and ROI.
In this interview, Meituan’s technology committee chair Xia Huaxia explains how the company’s architecture evolved from a simple LAMP stack to a multi‑layer, cloud‑native system, detailing the separation of infrastructure, business, and front‑end components, the use of open‑source tools, and the systematic simplification, standardization, automation, and quantification of business processes to support rapid O2O growth.
This article describes how Meituan migrated from Hive‑SQL and MapReduce to Spark on YARN, built an interactive Zeppelin‑based development platform, created reusable ETL templates, constructed a Spark‑driven feature and data‑mining platform, and applied Spark to interactive user‑behavior analysis and large‑scale SEM services, highlighting performance gains and operational benefits.
The article chronicles Meituan CEO Wang Xing’s entrepreneurial path—from early failures and campus social networks to the rise of Meituan—highlighting his iterative product strategy, relentless trial‑and‑error, and the leadership traits that turned repeated setbacks into lasting success.
This article details Meituan's 2014 presentation on building a performance analysis framework and monitoring platform, describing the bottlenecks they faced, the metrics they collected, the insights gained, and the tangible improvements achieved across front‑end and back‑end processing.
This article explains Meituan's end‑to‑end recommendation system architecture, data processing pipeline, candidate generation strategies, model training and online ranking techniques, illustrating how data, algorithms, and real‑time signals are combined to improve relevance and conversion.
This article details Meituan's end‑to‑end recommendation system, covering its four‑layer architecture, data sources, candidate‑generation strategies, fusion methods, and both linear and non‑linear re‑ranking models, while highlighting practical optimizations like AB testing and online learning.
This article outlines Meituan’s end‑to‑end recommendation system, describing its data layer, candidate‑generation triggers, fusion strategies, and machine‑learning‑based ranking models—including collaborative filtering, location‑based, query‑based, graph‑based methods, and both linear and non‑linear models—while highlighting practical optimizations such as AB testing, real‑time behavior handling, and fallback strategies.
The article describes Meituan's supply chain platform evolution from a manual, low‑volume system to a highly automated, multi‑business supporting architecture, detailing its stages, modular product center design, and the platform‑plus‑differentiation approach that enables rapid scaling across various services.
Meituan’s new Docker‑based automated build system isolates each build in lightweight containers, using YAML‑defined images to compile code, preprocess assets, and package artifacts, which speeds builds, eliminates environment conflicts, and delivers pre‑built packages for faster, more reliable releases across the platform.
This article outlines how Meituan applies machine learning to industrial challenges by detailing the full workflow—from problem modeling and data preparation to feature engineering, model training with algorithms like Logistic Regression and GBDT, and optimization techniques that address underfitting and overfitting for large‑scale deployment.
Meituan’s recommendation platform comprises a data layer, a multi‑strategy candidate generation layer, a fusion‑and‑filtering layer, and a ranking layer that uses additive‑grove tree ensembles and online‑updated logistic regression, leveraging extensive user behavior logs, location, query, graph and real‑time signals to deliver personalized deals.
The article argues that improving the utilization of existing resources—through smarter organization, shared‑economy models and internet‑enabled coordination—creates far greater value than merely swapping technologies, using cinema, automotive and Meituan examples to illustrate this insight.
This article examines Meituan’s evolution from a group‑buying platform to an O2O powerhouse, detailing its T‑shaped strategy, the rise of its Cat Eye movie business, competitive pressures, and the broader shift from e‑commerce 1.0 to 2.0 in China.
Meituan’s rapid O2O expansion hinges on a disciplined, data‑driven ground‑sales force that first conquers suburban districts before encircling city centers, leveraging aggressive incentives, real‑time IT tools, and strategic partnerships to wrest market share from rivals like Dianping while building a comprehensive ecosystem.
