Design and Architecture of a Cloud‑Based Shopping Cart System

Introduction

Shopping carts serve as multi‑item selection, temporary storage, and a key promotion channel for merchants.

Early Design

Initial decomposition includes ERP splitting, business service decomposition, and WCS splitting.

Module Overview

The functional modules of the cart are presented, illustrating the overall capabilities.

Layered Design

From the application layer perspective, three layers are defined: the Interaction layer (shopping page for adding items and checkout page for order submission), the Business Assembly layer (providing standard cart workflow), and the Base Service layer (encapsulating external system data and core utilities).

Cluster Design

Two vertical clusters are designed: the Cart cluster, which handles high traffic and stores sensitive user information, and the Settlement cluster, which manages non‑sensitive configuration data for order settlement.

Technical Architecture Goals

The distributed design aims to achieve stability (7×24 service), high performance under concurrent load, elastic scaling via virtualization (VM/LXC), no single point of failure, and automated fault masking for network, application, and database failures.

Three‑Level Cache

A three‑level caching architecture is introduced to improve data access speed and reduce backend load.

Asynchronous Checks

Asynchronous validation mechanisms are employed to ensure data consistency and improve system responsiveness.

Payment Scheme

The payment solution includes a heterogeneous middle‑platform design, Nginx+Lua aggregation of business interfaces, and detailed flow diagrams for cart payment processing.

Anti‑Bot and Multi‑Dimensional User Identification

Strategies for anti‑bot protection and multi‑dimensional user feature recognition are described to prevent fraudulent activities.

Source: PetterLiu, originally published on https://www.cnblogs.com/wintersun/.