← Back to Hub ← 返回主页

Week 2: Introduction (Part 2) 第 2 周：引言（第二部分）

Core Concepts 核心概念

1. Business Information & Connection Types (1.4)

Enterprises process distinct data types (voice, data, image, video). To support these, networks employ two primary line connection configurations:

  • Point-to-Point Connection: Provides a dedicated link between two devices (e.g., PC directly cabled to a Switch, or microwave link between two towers).
  • Multipoint (Shared) Connection: More than two devices share a single physical transmission link (e.g., legacy bus Ethernet or Wi-Fi channel).

2. Data Quality & Information Transmission (1.5)

Business decisions rely on high data quality. Key data quality dimensions include:

  • Timeliness: Data must arrive when needed. Delayed network traffic makes real-time data useless.
  • Accuracy: Information must be free of errors and correctly represent the raw source.
  • Completeness: All necessary details must be delivered without packets missing.

During transmission, systems use **DNS** (resolving host domain names to numeric IP addresses) and **HTTP** to request and fetch web pages.

3. Centralized vs. Distributed Data Processing (1.6)

Centralized Data Processing: All database storage and compute processes run on a single central host mainframe. Clients act as dumb terminals.

Distributed Data Processing (DDP): Disperses computing work and data storage across multiple nodes connected via a network. DDP offers high local responsiveness and system reliability.

1. 商业信息与连接类型 (1.4)

企业处理不同的数据类型（语音、数据、图像、视频）。为了支持这些，网络使用两种主要的线路连接配置：

  • 点对点连接 (Point-to-Point)：在两个设备之间提供一条专用链路（例如 PC 直接连接到交换机，或两座塔之间的微波链路）。
  • 多点（共享）连接 (Multipoint)：两个以上的设备共享单个物理传输链路（例如传统总线以太网或 Wi-Fi 信道）。

2. 数据质量与信息传输 (1.5)

商业决策依赖于高数据质量。关键数据质量维度包括：

  • 时效性 (Timeliness)：数据必须在需要时到达。延迟的网络流量会使实时数据失去价值。
  • 准确性 (Accuracy)：信息必须无差错，且正确表示原始来源。
  • 完整性 (Completeness)：必须交付所有必要的细节，且无数据包丢失。

在传输期间，系统使用 **DNS**（将主机域名解析为数字 IP 地址）和 **HTTP** 来请求和获取网页。

3. 集中式与分布式数据处理 (1.6)

集中式数据处理：所有数据库存储和计算流程都在单个中央大型机主机上运行。客户端充当哑终端。

分布式数据处理 (DDP)：将计算工作和数据存储分散在通过网络连接的多个节点上。DDP 提供高度的本地响应能力和系统可靠性。

Concept Visualization: Network Architecture 概念可视化：网络架构

Centralized network processing relies on a single core host, whereas distributed architectures leverage multiple peer nodes sharing processing logic.

集中式网络处理依赖于单个核心主机，而分布式架构则利用多个共享处理逻辑的对等节点。

Centralized Architecture Main Host Distributed Architecture

Tips 提示

  • DDP Pros & Cons: While Distributed Data Processing increases local responsiveness and system reliability (if one node fails, others continue), it requires database replication models to prevent data inconsistency.
  • DNS Caching: DNS servers cache IP mappings locally. If you update a domain's IP, propagation delay can cause users to access the old IP until the TTL (Time to Live) expires.
  • DDP 优缺点：虽然分布式数据处理提高了本地响应能力和系统可靠性（如果一个节点发生故障，其他节点仍能继续），但它需要数据库复制模型以防止数据不一致。
  • DNS 缓存：DNS 服务器在本地缓存 IP 映射。如果你更新了域名的 IP，传播延迟可能会导致用户访问旧 IP，直到 TTL（生存时间）过期。

Practical Exercises & Step-by-Step Guide 实践练习与分步指南

Exercise 1: Draw Centralized vs Distributed Flows 练习 1：绘制集中式与分布式流程图

Steps to complete:

  1. Go to Draw.io.
  2. Centralized Architecture: Place one large Mainframe/Host shape in the center. Surround it with four Terminal/PC shapes. Draw solid unidirectional arrows pointing from the terminals to the central mainframe, representing synchronous queries.
  3. Distributed Architecture: Place three Database Server shapes in a triangle pattern. Link them using bidirectional dashed arrows labeled "Replication / Sync Engine" to illustrate asynchronous distributed database updates.

完成步骤：

  1. 访问 Draw.io 网页版。
  2. 集中式架构：在中心放置一个大的 Mainframe/Host（主机）图形。在其周围放置四个 Terminal/PC（终端）图形。绘制指向中央主机的单向实线箭头，表示同步查询。
  3. 分布式架构：三角形排布放置三个 Database Server（数据库服务器）图形。使用带有“Replication / Sync Engine”（数据复制/同步引擎）标签的双向虚线箭头将它们连接起来，表示异步分布式数据库更新。

Exercise 2: Capture and Analyze DNS/HTTP Payload 练习 2：捕获并分析 DNS/HTTP 有效载荷

Steps to complete:

  1. Start a Wireshark capture on your active network interface.
  2. Open a browser, clear your cache, and visit a plaintext site (e.g. example.com).
  3. Stop the capture. In the Wireshark filter, type dns and find the query for example.com and the response providing the IPv4 address.
  4. Clear the filter and type http. Find the HTTP GET packet. Expand the Hypertext Transfer Protocol section in the detail pane to view the HTTP Host and User-Agent headers.

完成步骤：

  1. 在活动的网络接口上启动 Wireshark 捕获。
  2. 打开网页浏览器，清除缓存，并访问一个纯文本网站（例如 example.com）。
  3. 停止捕获。在 Wireshark 过滤器栏输入 dns，找到针对 example.com 的查询以及提供其对应 IPv4 地址的响应数据包。
  4. 清除过滤器并输入 http。找到 HTTP GET 数据包。在详细信息窗格中展开 Hypertext Transfer Protocol 部分，以查看 HTTP 主机 (Host) 和浏览器代理 (User-Agent) 标头。

Self-Check Practice Questions 自测练习题

1. Identify the term for raw, unprocessed facts in business systems. 在商业系统中，识别代表未加工的原始事实的术语。

2. Describe the main characteristic of processed business information. 描述处理后的商业信息的主要特征。

3. Explain why companies require structured data communication networks. 解释为什么企业需要结构化的数据通信网络。

4. Identify the term that represents data that has been structured and analyzed. 识别代表经过结构化和分析的数据的术语。

5. Describe the business benefit of high-speed data communications. 描述高速数据通信对企业的益处。

6. Explain how data networks support transactional business systems. 解释数据网络如何支持事务型商业系统。

Vocabulary Matching Game 词汇配对游戏

Match the term on the left with its definition on the right. (Click a term, then click its matching definition). 将左侧的术语与右侧的定义进行配对。（点击术语，然后点击匹配的定义）。

DDP
Centralized Processing
Point-to-Point
Multipoint Connection
Timeliness
All database storage and logic run on a single main mainframe 所有数据库存储与处理逻辑在单个核心主机上运行
Dedicated physical transmission link connecting exactly two devices 连接且仅连接两台设备的专用物理传输链路
Dispersion of compute work and databases across networked nodes 计算工作和数据库在网络连接的多个节点间分散
Data quality dimension requiring information to arrive when needed 要求数据必须在需要时及时到达的数据质量维度
Shared transmission link where more than two devices communicate 两个以上设备共同进行通信的共享式传输链路