Tags: , ,
Edit

Client-server architecture

GISC 4530K
Dr. Huidae Cho
Institute for Environmental and Spatial Analysis...University of North Georgia
Edit

1   What is the client-server architecture?

Client-server-model.svg

Image source
  • Distributed application framework
  • Divides tasks between servers and clients
  • Servers and clients communicate through computer networks such as
    • a Local Area Network (LAN),
    • a Wide Area Network (WAN), or
    • the Internet
  • Typically uses
    • the Transmission Control Protocol (TCP) or
    • the User Datagram Protocol (UDP)
Edit

2   What is a client?

Is it a computer? Well, we need a computer to run any computer programs for client-server computing.

A client is a process running on a computer.

It makes requests to the server and presents responses from the server to the user.

The most typical client is a web browser such as Firefox. We typically use a web browser to run web client applications written in JavaScript.

Edit

3   What is a server?

Again, a server is a process running on a computer.

Depending on the application design, both the client and server can run on a single computer.

The server receives and responds to requests from the client.

The most typical server is a web server such as Apache. Server applications running on a web server can be written in many different languages including Python, Node.js, and PHP.

Edit

4   Four main categories of client-server computing

Edit

4.1   One-tier architecture

  • A program running on a single computer
  • Standalone application
  • Does not require any network traffic
Edit

4.2   Two-tier architecture

  • Client and server tiers
  • Requires a network protocol to link both tiers
  • User Interface (UI) code on the client tier
  • Domain logic and data on the server tier
Edit

4.3   Three-tier architecture

  • Presentation, application, and data tiers
  • Requires a network protocol to link all tiers
  • UI code on the presentation tier
  • Service and data processing on the application tier
  • Data Input/Output (IO) on the data tier
Edit

4.4   N-tier (multitier) architecture

  • Divides an application into multiple logical layers and physical tiers with different responsibilities
  • Improves scalability
  • Adds latency because of increased network traffic
  • Supports different network topologies
Edit

5   Relational database management system

A Relational Database Management System (RDBMS) is used to input, manage, and output data.

It sits behind the server tier and provides data to the server.

It uses a standardized programming language for RDBMS called the Structured Query Language (SQL).

A spatial RDBMS provides spatial capabilities on top of its back-end RDBMS.

Edit

6   TCP vs. UDP

Transmission Control Protocol (TCP)

  • Connection-oriented
  • Built-in error checking
  • Larger overhead
  • Reliable and ideal for sensitive data transmissions

User Datagram Protocol (UDP)

  • Connectionless
  • No error checking
  • No overhead
  • Ideal for time-sensitive real-time communications

Most web applications using the Hypertext Transfer Protocol (HTTP) transport data over TCP. HTTP/3 uses QUIC + UDP.

Edit

7   How are data transmitted between the client and server?

  1. The Internet Protocol (IP) address of the target server is identified by the Domain Name System (DNS).
    • Let’s try nslookup.exe.
  2. Request data packets are routed from the client to the target server through multiple intermediate servers on Content Delivery Networks (CDNs).
    • Let’s try tracert.exe.
  3. The server sends response data packets to the client in the same way, but it already knows the IP address of the requesting client because it is a part of the request data packet.
Edit

8   A minimal HTTPS client using Python

import http.client
conn = http.client.HTTPSConnection("here.isnew.info")
conn.request("GET", "/")
res = conn.getresponse()
data = res.read()
conn.close()
print(data)
Edit

9   Major software for this class

Edit

10   References

Edit