Sharing insights and experiences from Dada Group's R&D department on product refinement and technology advancement, connecting with fellow geeks to exchange ideas and grow together.
Overwatch is an open‑source distributed monitoring platform built by Dada‑Jingdong Home that collects, aggregates, and visualizes RPC traffic across thousands of micro‑services in real time, enabling engineers to quickly pinpoint the root cause of system failures using directed‑graph visualizations and CQRS‑based data queries.
The article describes how JD Daojia's inventory system serves millions of stores by using health‑monitoring platforms, pre‑allocation policies, idempotent order processing, distributed locks, and rollback mechanisms to ensure stability, prevent overselling, and handle high‑concurrency scenarios such as flash‑sales.
This article explains the concept of continuous integration, its benefits, and details JD Daojia's end‑to‑end CI implementation—including Git commits, static code analysis, unit testing, automated packaging, deployment, UI/JSF/APP testing, code coverage, and future automation plans—illustrated with architecture diagrams and configuration snippets.
This article explains how JD Daojia leverages Redis for order list storage, ensuring data consistency, implementing distributed locks, and preventing cache penetration and avalanche through lock granularity and MQ-based cache updates, with detailed Java code examples.
To prevent loss of asynchronous tasks and data corruption during application restarts, this article explains how to safely shut down the JVM using shutdown hooks, custom signal handling, and proper hook implementation, while highlighting risks of forced termination and best‑practice considerations.
This article introduces reinforcement learning and deep reinforcement learning, explains key algorithms such as DQN and policy gradients, discusses black‑box optimization and the cross‑entropy method, and provides resources and code examples for further study.
This article explores how to optimize Babel usage to reduce bundle size and improve performance by addressing helper duplication, third-party module handling, and import optimization through transform-runtime, include patterns, and import transformation plugins.
The article outlines the complex business logic of JD Daojia's transaction system and presents architectural decisions such as service setization, database sharding, frontend‑backend separation, parallel execution, asynchronous processing, dependency governance, and future enhancements to improve scalability, reliability, and performance.
This article details how New Dada transformed its financial account system from a simple courier‑focused service into a platformized, standardized solution that serves multiple business entities, improves automation, and enhances development efficiency through unified APIs and self‑service tools.
