Network Architect Slides Network Architect
Client-Server Fundamentals
V. 1.0 // UNIT 1.1
The Phone Smash Test
If you smash your phone right now, do you lose all your Instagram photos?
Local Storage
The physical memory on your device.
The Network
Data stored on someone else's computer.
Who's Who?
The Client (Requester)
The device or software that asks for information. It starts the conversation.
Example: Your phone browser, Spotify app, or Netflix app.
The Server (Provider)
A powerful computer that waits for requests and serves up data.
Example: Google's data centers, Amazon's cloud, Discord's databases.
The Communication Loop
Client
REQUEST
RESPONSE
Server
Scenario: The Scroll
01
The Trigger
You open your feed. The Client sends an HTTP Request asking for the latest posts.
02
The Fetch
The Server looks up your ID, pulls photos from storage, and packages them into a Response.
03
The Render
The Client receives the data and displays it on your screen. You start scrolling!
Request Response Blueprint Worksheet Request Response Blueprint
UNIT 1.1 // NETWORK ARCHITECTURES
Name:
Date:
PART A
Defining the Architecture
Match the component to its correct technical description by drawing a line.
THE CLIENT
THE SERVER
THE REQUEST
THE RESPONSE
The data (HTML, image, video) sent back to be displayed.
The powerful computer system that hosts and manages data.
The user's device or software that initiates the communication.
The message sent from the device asking for specific information.
PART B
The Connection Schematic
In the blueprint area below, diagram a user checking their email on a smartphone. Labels must include: Client , Server , Internet Path , Request , and Response .
CLIENT AREA
SERVER AREA
PART C
Technical Analysis
1. Imagine you are typing a URL into a browser. Which device is "listening" and which is "talking"? Explain using the client-server model.
2. If a server goes offline (crashes), can the client still display the website? Why or why not?
3. List three examples of servers you interact with every day that you never actually "see" in person.
Power of Peers Slides Power of Peers
Distributed Networking & P2P
V. 1.0 // UNIT 1.2
The Classroom Experiment
Model A: Central
One student at the whiteboard. Everyone waits for them to draw a piece of the picture.
Bottleneck Risk!
Model B: P2P
Everyone has a small piece of paper. You pass them around and trade until everyone has the whole picture.
Resilient & Fast!
No Bosses, Just Peers
Decentralization
There is no central server. Every device on the network is both a client and a server at the same time.
Nodes
Each computer in the network is called a "Node." Nodes talk directly to each other without a middleman.
Where is P2P used?
BitTorrent
File sharing where you download small pieces of a file from hundreds of other users simultaneously instead of one server.
Very fast for large files
Resilient if one user leaves
Blockchain
A distributed ledger where every "node" keeps a copy of the transaction history. No bank (central server) required.
Highly Secure
Fully Transparent
P2P: The Trade-off
Advantages
No Central Failure: If one node goes down, the network stays alive.Scalability: More users = more bandwidth and resources.
Challenges
Security: Harder to control who is on the network and what they share.Consistency: Hard to keep data synced across thousands of nodes. Node Network Game Guide Node Network Game
TEACHER FACILITATION GUIDE // P2P VS CENTRALIZED
30 MINS
Objective
Students will physically simulate the difference between a centralized server-based network and a distributed peer-to-peer network to understand bottlenecks, speed, and fault tolerance.
Materials Needed
• 20+ Index Cards
• Markers / Pens
• Stopwatch
• "The Secret Image" (A simple grid drawing)
Phase 1: Centralized Model
The Roles
1 Server: (The fastest writer)8+ Clients: (The rest of the group)
The Task
The Server has the "Secret Image" (e.g., a 4x4 grid of shapes). Each Client needs a copy. The Server must draw the image onto each client's card one at a time. Clients cannot help.
"Teacher Note: Start the timer. Clients will get bored quickly—this is a simulation of high latency and server bottlenecks."
Phase 2: Peer-to-Peer (P2P) Model
The Roles
9+ Peers: (Everyone is equal)
The Task
The "Secret Image" is divided into 4 pieces. Give different pieces to different students. Students must move around the room and "download" (copy) pieces from each other until their card is complete.
"Teacher Note: Observe the chaos! Even though it looks messy, it's often significantly faster than the centralized model."
Injecting a Fault
Mid-way through each round, announce a "Server Crash" or "Node Failure" .
Centralized Model
"The Server must put their pen down for 30 seconds."
Result: The entire network stops. Latency spikes to infinity.
P2P Model
"Student X and Student Y must leave the network (sit down)."
Result: The network slows slightly, but other nodes still have the data.
Debrief Questions
1. Which model was faster when the number of students (users) increased?
2. Why is P2P harder for a teacher (central authority) to control?
3. In the P2P model, what happens if everyone only wants to "leech" (download) and nobody wants to share?
Cloud Layers Slides Cloud Layers
IaaS, PaaS, & SaaS Models
V. 1.0 // UNIT 1.3
The Amazon Pivot
In 1995, Amazon sold its first book.
Today, they host 33% of the world's cloud infrastructure .
Why did a bookstore end up owning the "pipes" of the internet?
Netflix
Reddit
Fortnite
Zoom
All powered by AWS (Amazon Web Services)
Cloud Computing "As A Service"
IaaS
Infrastructure
"I build the building."
You rent the servers, storage, and networking. You install the OS and apps yourself.
AWS / Azure
PaaS
Platform
"I build the tools."
A space for developers to create apps without worrying about the underlying servers.
Heroku / Firebase
SaaS
Software
"I build the app."
Everything is managed. You just log in and use it in your browser.
Google Docs / Slack
Who Manages What?
Applications
Data
Runtime
Operating System
Virtualization
Servers / Networking
On-Premises YOU MANAGE EVERYTHING
IaaS THEY MANAGE HARDWARE
SaaS THEY MANAGE EVERYTHING
Why Go Cloud?
Speed
Launching a server takes 2 minutes in the cloud, compared to 2 weeks for physical shipping.
Cost
"Pay as you go." No need to buy expensive hardware upfront. Only pay for what you use.
Scale
Add 1,000 servers instantly when traffic spikes (like a movie release or sale).
Reliability
Data is copied across multiple data centers globally. If one fails, another takes over.
Cloud Service Sorter Worksheet Cloud Service Sorter
UNIT 1.3 // IaaS, PaaS, SaaS
Name:
Category Match
Review the common tech services below. Identify if they are Infrastructure (IaaS) , Platform (PaaS) , or Software (SaaS) . Then, provide a 1-sentence reasoning.
Service Name Category Reasoning Gmail / Outlook AWS EC2 (Virtual Servers) Firebase (App Tools) Spotify Azure Virtual Machines Salesforce CRM
Critical Thinking
The Startup Scenario:
You are launching a new social media app for high schoolers. You have a team of two developers and very little money. You don't have time to build physical servers or manage operating system updates.
Which cloud model (IaaS, PaaS, or SaaS) would you choose to build your app on? Why?
If you chose IaaS, what is one task your team would have to do that you wouldn't have to do with SaaS?
Cloud Risks
If Amazon Web Services (AWS) has a major power outage in their Virginia data center, what happens to the companies using SaaS vs those using IaaS with local backups?
Heavy Traffic Slides Heavy Traffic
Scaling & Load Balancing
V. 1.0 // UNIT 1.4
Black Friday Meltdown
When millions of people click "Buy" at 12:00 AM on Black Friday, why doesn't Amazon crash?
In 2013, Amazon went down for just 40 minutes . They lost $4.8 Million in sales.
To a giant network, time is literally money.
Server 503
Service Unavailable
How to Handle More Users?
Vertical Scaling
"The Bigger Hammer"
Upgrading a single server with more RAM and a faster CPU.
LIMIT: There is a physical limit to how fast one chip can go.
Horizontal Scaling
"The Strength in Numbers"
Adding more identical servers to share the workload.
BENEFIT: Virtually limitless. Just keep adding more nodes!
The Traffic Cop: Load Balancing
Load Balancer
Decides which server is the least busy and sends the request there.
Server A
Server B
Server C
Proximity Matters: CDNs
Content Delivery Network
If you are in London and want to watch a video hosted in Los Angeles, the data has to travel 5,000 miles.
A CDN keeps a copy of that video in a server in London. Speed increases because the distance decreases.
LOCAL COPY
LOCAL COPY
Connected World Slides Connected World
IoT & Edge Computing
V. 1.0 // UNIT 1.5
The Smart Toaster Attack
In 2016, the Mirai Botnet took down huge chunks of the internet (Twitter, Netflix, Airbnb).
They didn't use supercomputers. They used 100,000 hacked smart cameras, fridges, and toasters.
The Internet of Things (IoT) is convenient, but it's also a massive new "attack surface."
Botnet Detected
Internet of Things (IoT)
The Concept
Embedding sensors, software, and network connectivity into everyday physical objects, allowing them to collect and exchange data.
Big Data
Billions of devices generate more data than humans ever could. This data is used for "Smart Cities" and "Smart Homes."
Low Power
IoT devices often run on batteries for years, requiring special protocols (like Zigbee or LoRa) that use very little energy.
The Problem with Distance
A self-driving car generates 4TB of data per day . If it has to send that to a cloud server 1,000 miles away to decide if it should hit the brakes...
Latency is too high!
Edge Computing
Processing data locally (on the device or a nearby mini-server) instead of sending it all to the central cloud.
Real-time decisions
Saves bandwidth
The Next Frontier
5G / 6G
Ultra-fast speeds and near-zero latency will make Edge computing seamless.
Local AI
AI chips inside your devices (phones, glasses) will process your voice/images privately.
Zero Trust
Moving toward security where no device is trusted until verified, even inside your own home.
"Everything that can be connected, will be connected."
