Top Banner
Computing Resource Paradigms CS3353
19

Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Jan 14, 2016

Download

Documents

Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Computing Resource Paradigms

CS3353

Page 2: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Computing Resource Paradigms

• Centralized Computing

• Distributed Computing

Page 3: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Early Centralized Computing Paradigm

• Characterized by one central computer.

• Housed in a limited access, climate controlled room.

• A shared resource – users connect through dumb-terminals.

• Limited interactive access.

• Jobs are submitted through batch job-queues.

Page 4: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.
Page 5: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Centralized Computing

• There remain highly specialized computer systems of this type:

• Supercomputers designed for special resource intensive problem solving.

Page 6: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Cray T3E/512-PE Supercomputer Mainframe

Page 7: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Centralized Computing Advantages

• Ideal for maintaining a high level of security.

• Climate controlled environment is equipment friendly.

• Reduces maintenance costs.

Page 8: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Distributed Paradigm

• Demand driven by departments wanting access to more computing resources beyond the confines of centrally controlled systems.

• Unix becomes a popular operating system for this paradigm.

• HP, Sun, SGI, and others ride this wave during the 1980’s.

Page 9: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

DEC Micro-VAX

Page 10: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Client-Server

• The spread of networks and low cost network hardware led to the client-server computing paradigm. This successfully merges the two central and distributed paradigms.

• Computers are interconnected by a network.

• The computers are classified as either clients or servers in the network.

Page 11: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Server

• A powerful computer configured to provide services to other computers within a computing system environment.

• Example Services:– Disk– Printer– Application– E-mail– etc

Page 12: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Server

• The server is typically a powerful computer configured with:

• Fast processors (1 or more)

• Fast and reliable disk storage

• Memory rich

Page 13: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Server

• Central location for the server allows ease of maintenance and high levels of security.

• Simplifying the maintenance of the system helps control costs.

• Small operations (businesses) can put a powerful central computing system in a small room. There it can be relatively secure and easy to maintain.

Page 14: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Server Rack

Page 15: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Behind the Server Rack

Page 16: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

The Client

• A client is configured to access specific services within the boundary of the defined computing system environment.

• Client classifications:– Fat client: fully functional workstation– Thin client: a dumb terminal or graphical

display– Hybrid client: between thin and fat.

Page 17: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Client

• Fat clients are fully functional computers with local storage and computing capability.

• Thin clients rely on a server to provide storage and computing resources. Characterized as diskless and without a processor.

• Hybrid clients provide some computing resources independent of a server; storage is provided by a server.

Page 18: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

Client

• Thin/Hybrid clients can cost less than fat-clients to purchase and maintain when carefully chosen.

• Work well for:– public access areas– high security environments– When full function workstations are unnecessary.– Reduces maintenance for cost conscious businesses.

Page 19: Computing Resource Paradigms CS3353. Computing Resource Paradigms Centralized Computing Distributed Computing.

$300 client