Design and Architecture of a Cloud Shopping Cart System

The shopping cart serves three main purposes: enabling users to select multiple items for checkout like a traditional store, acting as a temporary favorites list, and providing a prime location for merchants to promote products.

Early development involved ERP splitting, service‑oriented business decomposition, and WCS separation.

Module overview diagrams illustrate the overall functional components of the shopping cart system.

Layered design consists of an interaction layer (shopping page, checkout page), a business assembly layer (standard and non‑standard cart processes), and a foundational service layer that encapsulates data distribution and core functionalities.

Cluster design includes a high‑traffic, sensitive‑data shopping‑cart cluster and a less‑sensitive settlement‑cart cluster handling payment configurations and auxiliary information.

The technical architecture adopts a distributed design to achieve stability (24/7 reliability), high performance under concurrency, elastic scaling via VM or container technologies, elimination of single points of failure, and automated fault masking.

A three‑tier cache strategy is employed to improve data access efficiency.

Asynchronous checks are implemented to monitor system health without blocking main flows.

Storage heterogeneity is discussed, noting its simplicity of workflow but challenges under high‑traffic, high‑concurrency scenarios.

Payment solutions for the shopping cart are presented, followed by a heterogeneous middleware approach for payment processing.

An Nginx+Lua front‑end aggregates business interfaces to streamline request handling.

Anti‑scalper mechanisms and multi‑dimensional user feature identification techniques are introduced to detect and mitigate fraudulent activities.

Author: PetterLiu (original article: https://www.cnblogs.com/wintersun/p/6683957.html).