T Level Digital core content area 7: hardware, software, networks, virtual and cloud environments, and resilience.
Read the lesson, study the diagrams, watch the video, then open the matching workbook section.
Number systems Scheme of work · weeks 2 and 6
Binary, denary, octal and hexadecimal
By the end you can: convert between binary, denary and hexadecimal, and add, subtract and multiply in binary.
Every piece of data a computer handles, from a photo to a network address, is stored as binary: patterns of 0s and 1s. A single 0 or 1 is a bit; 4 bits are a nibble; 8 bits are a byte. People working in digital jobs read binary and hexadecimal every day: in IP and MAC addresses, colour codes, error messages and packet captures.
A number base is how many different digits the system uses. The value 173 written in four bases.Each hexadecimal digit stands for exactly one nibble (4 bits), so a byte is always two hex digits: 00 to FF.
Worked example: Denary to binary (173)
Write the place values: 128, 64, 32, 16, 8, 4, 2, 1.
173 − 128 = 45, so put 1 under 128. 64 does not fit into 45: 0. 45 − 32 = 13: 1. 16 does not fit: 0.
By the end you can: describe the features and uses of personal computers, mobile devices, servers and embedded devices.
Personal computers, mobile devices, servers and embedded devices are built for different jobs.
Physical computers compared
Type
Key features
Typical uses
Personal computer (desktop, laptop)
general purpose; keyboard, mouse, screen; upgradable (desktop); portable (laptop)
office work, programming, design, gaming, study
Mobile device (smartphone, tablet)
small, battery-powered, touchscreen, many sensors (GPS, accelerometer, camera), mobile processor, wireless connectivity
communication, apps, photos, navigation, mobile working
Server
powerful processors, lots of RAM and storage, redundant parts, runs 24/7, managed remotely
web, email, file, database, print and authentication services for many users
Embedded device
a small computer built into another product; one fixed job; low cost and power; often real-time
washing machines, cars, medical devices, smart meters, IoT sensors
Inside a smartphone. A System on a Chip (SoC) puts the CPU, GPU and other processors on one chip to save space and battery.A small server rack. Servers use redundant (duplicated) parts because many people depend on them at once.An embedded system reads sensors, runs one fixed program on a microcontroller, and controls outputs. Washing machines, cars, pacemakers and IoT devices all contain them.
The Internet of Things (IoT)
IoT devices are embedded systems with network connections: smart speakers, doorbells, thermostats, fitness trackers, factory sensors. They collect data and can be controlled remotely, which brings convenience and efficiency but also security risks, because many are cheap, rarely updated and joined to the same network as important systems.
Key terms
Embedded system
a computer inside a larger product, dedicated to one task.
Microcontroller
a single chip with a processor, memory and inputs/outputs.
Server
a computer that provides services to other computers over a network.
IoT
Internet of Things: everyday objects with sensors and network connections.
Exam focus
For ‘describe the features’ questions, give a feature and say what it allows, e.g. “servers have redundant power supplies, so the service keeps running if one fails.”
Think it through
Why does a server usually have more RAM than a personal computer?
It runs services for many users at the same time, each needing memory for their requests, data and connections.
Give two features of an embedded device in a washing machine.
It runs one fixed program (the wash cycles), reads sensors such as water level and temperature, and must be cheap, reliable and low-power.
Watch
A short revision video from the MrBrownCS series.
Watch and think:
List three embedded systems in your home.
Why are embedded systems usually cheap and low power?
By the end you can: explain the features and use of input and output devices, processors, memory, storage, motherboards, GPUs, network interfaces, cooling and sensors.
Getting data in and results out
Input devices
Output devices
Sensors (automatic input)
keyboard, mouse, touchscreen, microphone, scanner, webcam, barcode reader, game controller
monitor, printer, speakers, headphones, projector, 3D printer, actuators (motors)
A multi-core CPU. Each core runs its own instructions; small private caches are fastest, and a larger L3 cache is shared.
What affects processor performance
Feature
What it means
Effect
Number of cores
independent processing units on one chip
more tasks at once (if software uses them)
Clock speed (GHz)
cycles per second
higher = more instructions per second, but more heat
Cache size (MB)
very fast memory on the CPU
fewer slow trips to RAM
Mobile processors
designed for low power, often a system on a chip
longer battery life, less heat, slightly less peak speed
Memory and storage
The memory hierarchy. Each level is slower but larger and cheaper than the one above, so computers copy the data they need upwards.
RAM and ROM
RAM
ROM
Volatile?
yes: loses data without power
no: keeps data
Holds
programs and data in use now
start-up instructions (BIOS/UEFI firmware)
Can be changed?
constantly read and written
read-only (or rarely updated)
An HDD has a seek time because the head and platter must move. An SSD reads flash memory electronically, so it is faster, quieter and shock-resistant.
Secondary storage types
Secondary storage
How it stores data
Strengths
Weaknesses
Magnetic (HDD, tape)
magnetised spinning platters or tape
cheap per GB, huge capacity
moving parts; slower; damaged by knocks
Solid state (SSD, USB, SD)
electrical charges in flash memory
fast, silent, tough, low power
more expensive per GB; limited write cycles
Optical (CD, DVD, Blu-ray)
laser reads pits and lands
cheap, portable, good for distribution
small capacity; scratches; slow
The motherboard, GPU, network interfaces and cooling
A desktop motherboard (simplified). Parts plug into standard slots, which is why desktops are easy to upgrade and repair.
The motherboard connects every component and carries power and data between them. A graphics processing unit (GPU) has thousands of small cores for parallel work: graphics, video and AI. A network interface card connects the computer to a network; it may plug into a PCI/PCIe slot on the motherboard or into a USB port (for example a USB Wi-Fi adapter).
Processors must be cooled or they slow down (throttle) or fail. Air cooling is cheap and simple; liquid cooling handles more heat.
Key terms
Volatile
loses its contents when the power is off.
Clock speed
how many cycles a processor completes per second (GHz).
Cache
small, very fast memory on the processor.
GPU
graphics processing unit: many cores for parallel processing.
PCIe
the motherboard slot standard for expansion cards.
Thermal throttling
slowing a processor down to stop it overheating.
Practical: identify the parts
Use the Motherboard ID task: label the CPU socket, RAM slots, PCIe slots, chipset, SATA ports, power connectors and CMOS battery.
Match connectors to their names: SATA, Molex, 24-pin, 4/8-pin CPU and 6/8-pin PCIe power.
Look up a real motherboard specification online and find its RAM type, maximum RAM and number of M.2 slots.
Your teacher runs the PC disassemble and reassemble practical in class with anti-static precautions.
Exam focus
When asked to justify hardware for a scenario, link each feature to the user’s need: “a video editor needs 32 GB RAM because large project files must stay in memory while editing.”
Think it through
Why does adding more cores not always make a program faster?
The program must be written to split its work into parallel threads; a program that runs step by step can only use one core.
A laptop feels slow when many browser tabs are open. Which upgrade is most likely to help, and why?
More RAM: the tabs need more memory than is available, so the OS keeps swapping data to storage, which is much slower.
Watch
A short revision video from the MrBrownCS series.
Watch and think:
Why doesn’t doubling the cores always double the speed?
By the end you can: explain the features and uses of batch, multitasking, real-time, network and mobile operating systems.
The operating system sits between programs and hardware. Programs ask the OS for resources rather than controlling the hardware directly.
Operating system types
OS type
Key features
Examples of use
Batch
non-interactive; processes high volumes of similar jobs together; runs on a schedule
payroll, printing bills and statements, overnight data processing
Multitasking
runs many tasks concurrently by giving each a time slice; uses interrupts to respond to events
Windows, macOS, Linux desktops
Real-time (RTOS)
guarantees a response within a deadline; used for monitoring and control, and transaction processing
car engine management, airbags, medical monitors, booking seats or card payments
Network OS
shares resources (files, printers), manages users and permissions, supports communication across a network
Windows Server, Linux servers
Mobile OS
for smartphones and tablets; lower processing requirements; manages power for longer battery life; touch interface; app store
Android, iOS, iPadOS
Batch systems favour throughput with no interaction; real-time systems guarantee a response within a fixed time.Multitasking: the scheduler gives each program a short time slice in turn, so fast that they all seem to run at once.
Time-slicing gives each running task a tiny slice of processor time in turn, so they appear to run at the same time. An interrupt is a signal (a key press, a network packet arriving, a timer) that makes the processor pause, deal with the event, then carry on.
Key terms
Batch processing
grouping jobs and processing them together without user interaction.
Time-slicing
sharing processor time between tasks in small turns.
Interrupt
a signal that makes the processor deal with an event.
RTOS
real-time operating system: responds within a guaranteed time.
Exam focus
Match the OS type to the scenario’s need: deadlines → real-time; high volume and scheduled → batch; many users and shared resources → network OS; battery life and touch → mobile OS.
Think it through
Why is an online seat booking system an example of real-time transaction processing?
Each booking must be processed immediately and the seat locked, so two customers cannot buy the same seat.
Why do mobile operating systems close background apps more aggressively than desktop ones?
To save battery and memory: mobile devices have less power and RAM, so idle apps are paused or closed.
By the end you can: explain the features and uses of file management, defragmentation, compression, package managers, protection and backup software.
Utility software helps look after the computer rather than doing a user’s task. Most operating systems include utilities, and others are installed separately.
Common utilities
Utility
What it does
Why it matters
File management
create, copy, move, rename, delete and search files and folders; manage permissions
keeps data organised and findable
Defragmenter
rearranges the scattered parts of files on a hard disk so each file is stored together
fewer head movements, faster reads on HDDs (not needed for SSDs)
File compression
makes files smaller (lossless ZIP, or lossy for media)
saves storage and speeds up transfers
Package manager
installs, updates and removes software and its dependencies from trusted repositories (apt, winget, npm, pip)
consistent, secure, repeatable installs and quick updates
By the end you can: explain the features of IDEs (code editing, debugging, screen design), compilers and interpreters.
An IDE puts code editing, debugging and screen design tools in one program, speeding up development and making errors easier to find.
Features of an integrated development environment
IDE tool
Features
Benefit
Code editing
syntax highlighting, auto-complete, auto-indent, error underlining, find and replace, refactoring
faster coding and fewer typing errors
Debugging
breakpoints, stepping line by line, watching variable values, call stack
find logic errors by seeing exactly what the program does
Screen design
drag-and-drop layout of forms, buttons and text boxes
build user interfaces quickly and consistently
Build and run
one click to compile/run, integrated terminal, version control
quick testing and teamwork
Both turn high-level code into something the processor can run; they differ in when the translation happens.
Some languages use both: Java and C# compile to an intermediate bytecode, which a virtual machine then runs. Python is usually interpreted, which makes it quick to try things out.
Key terms
IDE
integrated development environment: tools for writing, testing and debugging code in one program.
Breakpoint
a marked line where the program pauses while debugging.
Compiler
translates a whole program into machine code before it runs.
Interpreter
translates and runs a program one line at a time.
Think it through
Why might a games studio compile its final game?
Compiled code runs faster and players do not need the source code or an interpreter; it also protects the code from being easily read.
How does a breakpoint help find a logic error?
The program pauses at that line so you can inspect variable values and step through, seeing where the values stop matching what you expected.
planning and tracking software projects and teamwork
In the real world: a doctor’s surgery
Reception uses a database for patient records and appointments, a spreadsheet to track monthly budgets, word processing for referral letters (often mail-merged from the database), email to communicate with the hospital, and project management software to plan a move to a new records system.
Key terms
Application software
programs that help users carry out tasks.
Relational database
data stored in linked tables.
Mail merge
combining a template with a list of data to create personalised documents.
Gantt chart
a timeline showing tasks, durations and dependencies.
Think it through
Why would a business use a database rather than a spreadsheet for 50,000 customer records?
Databases handle large volumes, link related tables, enforce validation, let many users work at once, and run complex queries quickly.
Name one feature of project management software that helps a team meet a deadline.
A Gantt chart or Kanban board showing task dependencies and progress, so delays are spotted early.
By the end you can: explain the benefits and drawbacks of networking devices, and the features of PAN, LAN, MAN, WAN and VPN.
7.3.1 Why connect devices?
Benefits of connecting devices
Drawbacks
share files, printers and internet connections; communicate (email, chat, VoIP); central backups and security; collaborate on the same data; manage devices centrally
cost of equipment and set-up; needs specialist staff; security risks (malware can spread); a failure of a key device can stop everyone; performance drops when the network is busy
Network types grow in size from a personal area network to a wide area network. Each larger network links smaller ones together.
7.3.2 Network types
Type
Number of users
Connection and coverage
Example
PAN
one person
a few metres; Bluetooth, USB
phone to earbuds and a smartwatch
LAN
a few to thousands
one building or site; Ethernet cables and Wi-Fi
school or office network
MAN
many organisations
a town or city; fibre links
council network linking libraries
WAN
millions
countries and continents; leased lines, fibre, satellite
the internet; a bank linking its branches
VPN
authorised remote users
runs over another network, usually the internet, encrypted
staff working from home reaching office files
A virtual private network (VPN) creates an encrypted connection over a public network, so remote users can reach private resources securely.
Key terms
LAN
local area network: one site.
WAN
wide area network: large geographical area.
VPN
virtual private network: an encrypted tunnel across a public network.
Coverage
the geographical area a network reaches.
Think it through
Give one drawback of networking computers in a small office.
Malware can spread from one infected computer to all the others, or the whole office stops if the router fails.
Why does a VPN matter for someone on public café Wi-Fi?
Everything is encrypted between the laptop and the VPN gateway, so others on the café network cannot read it.
By the end you can: compare the features, benefits and drawbacks of copper Ethernet, fibre-optic and wireless connections.
7.3.3 Connectivity methods
Method
How it works
Benefits
Drawbacks
Copper / Ethernet (e.g. Cat6 UTP)
electrical signals over twisted pairs of copper wire
cheap, easy to fit, reliable, fast (1–10 Gbps), can carry power (PoE)
max about 100 m per run; affected by electrical interference; cables must reach each device
Fibre optic
pulses of light through glass strands
very high speed and bandwidth; long distances (km); immune to interference; hard to tap
expensive; fragile; needs specialist installation and equipment
Wireless (access points)
radio waves between devices and a wireless access point (Wi-Fi 802.11)
mobility; no cables; quick to add users; supports phones and tablets
slower and less reliable than cable; range limited by walls; interference; must be secured (WPA2/WPA3)
No medium wins on everything: networks mix copper to desks, fibre for backbones and long runs, and wireless for mobile devices.
Key terms
UTP
unshielded twisted pair: the usual copper network cable.
Wireless access point
connects wireless devices to a wired network.
Attenuation
a signal getting weaker over distance.
Interference
unwanted signals that corrupt data.
Try it in Packet Tracer: wireless access point
Build a switch with two wired PCs and an AccessPoint-PT.
Set the SSID and WPA2-PSK passphrase on the AP (Config → Port 1).
Give a laptop a wireless card (WPC300N) and the same settings.
Ping a wired PC from the laptop.
Exam focus
Scenario questions often ask you to recommend a connection: weigh distance, speed, cost, interference and mobility against what the organisation actually needs.
Think it through
Why would a hospital use fibre between buildings?
The distance is too far for copper, fibre carries huge amounts of data such as scans, and it is not affected by electrical interference from medical equipment.
Give one reason an office might still wire desks with copper when it has Wi-Fi.
Wired connections are faster and more reliable for desktop PCs, and do not share bandwidth with every wireless user.
Watch
A short revision video from the MrBrownCS series.
Watch and think:
Why is Ethernet usually faster and more reliable than Wi-Fi?
By the end you can: compare star, mesh and tree topologies, physical and logical topologies, and client-server, thin client and peer-to-peer models.
Physical topologies describe how devices are cabled together. Star and extended star (with switches in the middle) are used in almost every modern LAN.
7.3.4 Topologies
Topology
Features
Benefits
Drawbacks
Star
every device has its own cable to a central switch
a cable fault affects one device; easy to add devices and find faults
the switch is a single point of failure; lots of cable
Mesh
devices connect to many (full mesh: all) others
many routes, so it keeps working if links fail; good for WANs and wireless mesh
expensive and complex; many connections to manage
Tree (hierarchical)
stars joined in layers from a central root
scales well; easy to extend by adding branches; faults contained in a branch
if the root or a trunk link fails, whole branches are cut off
The physical topology is how the cables and devices are actually laid out. The logical topology is how data flows between devices. They can differ: a network can be cabled as a star but pass data around logically like a ring or bus.
Three network models. The difference is where the files, logins and processing live.
7.3.5 Network models
Model
Benefits
Drawbacks
Client-server
central storage, logins, security and backups; scales to many users
servers and skilled staff cost money; if a server fails, services stop
Thin client
cheap, simple desk devices; data stays on the server; easy central management
depends completely on the network and server; poor for heavy graphics
Peer-to-peer
cheap and easy to set up; no server needed
hard to manage and secure beyond a few devices; backups scattered; slower when a peer is busy
Key terms
Physical topology
the actual layout of devices and cables.
Logical topology
the way data flows through the network.
Single point of failure
one component whose failure stops the whole system.
Thin client
a simple device relying on a server for processing and storage.
Think it through
Why is a tree topology suited to a large school on several floors?
Each floor or block can be a star connected to a central switch, so it is easy to extend, and a fault stays within one branch.
Give one reason a two-person start-up might choose peer-to-peer.
It needs no server or network administrator, so it is cheap and quick to set up for a few devices.
a device that requests and uses services: PCs, laptops, phones
Switch
connects devices within a LAN; forwards frames only to the correct port using MAC addresses
Router
connects different networks; forwards packets between them using IP addresses; usually links a LAN to the internet
Internet connection
the link from the organisation to its ISP (fibre, cable, DSL, 4G/5G, leased line)
Internet backbone
the very high-capacity links and routers (tier 1 networks, exchange points, undersea cables) that join networks worldwide
A switch reads the source MAC address of each frame to learn which device is on which port, then forwards frames only to the right port.Your internet connection links your home or office to an ISP, whose network joins the high-capacity backbone that carries traffic between networks and continents.
Common server types
Server type
Service
File server
shared folders with permissions
Web server
hosts websites (HTTP/HTTPS)
Mail server
sends and stores email (SMTP, IMAP/POP)
Print server
manages print queues
Database server
stores and queries data for applications
DHCP / DNS server
hands out IP addresses / turns names into IP addresses
Proxy server
requests web content on behalf of clients; filters and caches
Key terms
Switch
connects devices in a LAN using MAC addresses.
Router
connects networks and routes packets using IP addresses.
ISP
internet service provider.
Internet backbone
the high-capacity core links of the internet.
Try it in Packet Tracer: home network to the internet
Build a home LAN (PCs, switch or home router).
Add a Cloud-PT and a cable or DSL modem between the home router and the cloud.
In the cloud, connect the modem’s interface to the Ethernet interface leading to a server.
Ping the server, then swap DSL for cable and compare.
Based on the teacher’s ‘Simulating a home network to the internet’ practical.
Think it through
What is the difference between the role of a switch and a router?
A switch connects devices within one network using MAC addresses; a router connects different networks and forwards packets between them using IP addresses.
Why are undersea cables part of the internet backbone?
They carry most data between continents at very high capacity, which satellites cannot match.
By the end you can: describe the function and related protocols of each layer of the seven-layer OSI model and the four-layer TCP/IP model.
The OSI model splits communication into seven layers; the TCP/IP model used on the internet groups them into four.
7.3.7 OSI model
OSI layer
Function
Example protocols / devices
7 Application
network services for applications
HTTP, HTTPS, FTP, SMTP, DNS
6 Presentation
formats, encrypts and compresses data
TLS/SSL, JPEG, ASCII
5 Session
starts, manages and ends sessions
session set-up and checkpoints
4 Transport
end-to-end delivery, ports, splitting into segments, reliability
TCP, UDP
3 Network
logical addressing and routing between networks
IP, RIP, OSPF · routers
2 Data link
frames, MAC addresses, error detection on a link
Ethernet, Wi-Fi · switches
1 Physical
transmits bits as signals over the medium
cables, radio, fibre · hubs
The four-layer TCP/IP model used on the internet, with protocols from the specification at each layer.Encapsulation wraps the data in headers, like putting a letter in an envelope, then in a parcel, then on a lorry.
Key terms
Protocol
an agreed set of rules for communication.
Encapsulation
each layer adds its own header to the data.
TCP
reliable, connection-based transport.
UDP
fast, connectionless transport without re-sending.
Exam focus
Learn one job and two protocols for every layer. Questions often give a protocol and ask for its layer, or ask why layers help (standards, interoperability, easier troubleshooting).
Think it through
Which OSI layers are combined into the TCP/IP application layer?
Application, presentation and session (layers 7, 6 and 5).
At which layer does encryption with TLS take place in the OSI model?
The presentation layer (layer 6), which handles encryption, compression and data formats.
By the end you can: explain the role of web, mail, routing and application protocols: HTTP, HTTPS, SMTP, POP, IMAP, RIP, OSPF, FTP, SFTP, DHCP and DNS.
7.3.10 Protocols in the specification
Group
Protocol
Role
Port
Web
HTTP
transfers web pages between browser and server (not encrypted)
80
Web
HTTPS
HTTP encrypted with TLS: protects logins, payments and personal data
443
Mail
SMTP
sends email from client to server and between mail servers
25 / 587
Mail
POP (POP3)
downloads email to one device, usually deleting it from the server
110 / 995
Mail
IMAP
keeps email on the server and synchronises it across devices
143 / 993
Routing
RIP
routers share routes; chooses the path with the fewest hops (max 15)
–
Routing
OSPF
link-state routing: builds a map of the network and chooses the lowest-cost (fastest) path
–
Application
FTP
transfers files (not encrypted)
20 / 21
Application
SFTP
secure file transfer over SSH (encrypted)
22
Application
DHCP
automatically gives devices an IP address, mask, gateway and DNS server
67 / 68
Application
DNS
translates domain names into IP addresses
53
DHCP gives each device an IP address, subnet mask, default gateway and DNS server automatically (remember DORA).DNS turns names people can remember into the IP addresses computers use. Answers are cached to save time.
Key terms
Port number
identifies which service or application data is for.
Hop
one step from one router to the next.
Link-state
a routing method where routers share a full map of links and their costs.
Try it in Packet Tracer: DHCP and DNS servers
Add a server with static IP 192.168.1.2.
Turn on DHCP (pool from .100, gateway .1, DNS .2) and DNS (A record intranet.local → 192.168.1.2) and HTTP.
Set PCs to DHCP, then browse to intranet.local.
Exam focus
Compare POP and IMAP carefully: POP downloads to one device; IMAP keeps mail on the server and syncs every device. A common question asks which suits a worker who uses a phone and a laptop.
Think it through
Why is RIP less suitable than OSPF for a large network?
RIP only counts hops (maximum 15) and ignores link speed, so it may pick a slow route; OSPF uses link costs, scales better and adapts faster to changes.
Which protocol should a web designer use to upload files to a client’s server, and why?
SFTP, because it encrypts the login and the files, unlike FTP.
Watch
A short revision video from the MrBrownCS series.
Watch and think:
Which protocol would you use to upload a website securely?
By the end you can: explain bandwidth and latency and their effect on the performance of networks and connected systems.
Bandwidth is the amount of data a connection can carry per second; latency is the delay before data arrives. They affect different kinds of application.
How bandwidth and latency affect applications
Application
Needs high bandwidth?
Needs low latency?
Why
Streaming 4K video
yes
not very
lots of data, but buffering hides short delays
Online gaming
not much
yes
small updates must arrive instantly or play lags
Video calls / VoIP
moderate
yes
delay causes people to talk over each other
Large file downloads / backups
yes
no
total time depends on how much data per second
Remote control of machinery or surgery
moderate
critical
a delay could be dangerous
Worked example: Download time
A 2 GB file is 2,000 MB = 16,000 megabits (× 8).
On a 100 Mbps connection: 16,000 ÷ 100 = 160 seconds (about 2 min 40 s).
On a 20 Mbps connection: 16,000 ÷ 20 = 800 seconds (over 13 minutes).
Latency barely matters here; bandwidth decides the time.
Remember: bandwidth is usually in bits per second, file sizes in bytes, so multiply bytes by 8.
Performance also depends on contention (many users sharing the same bandwidth), distance and the number of hops, congestion and packet loss (which causes re-sends), and the slowest link in the path.
Key terms
Bandwidth
the maximum data a connection can carry per second.
Latency
the delay for data to travel from source to destination.
Throughput
the data actually delivered per second.
Jitter
variation in latency, which disrupts calls and games.
Think it through
Why can a satellite connection feel slow for gaming even with good bandwidth?
The signal travels to a satellite and back, adding high latency, so each action takes noticeably longer to register.
How long does a 500 MB file take to download at 50 Mbps?
7.4 Virtual environments T Level core · 7.4.1 – 7.4.4
Virtual machines, hypervisors and virtual environments
By the end you can: explain virtual machine clients and servers, type 1 and type 2 hypervisors, and the key features, benefits and drawbacks of virtual environments.
Virtualisation runs several virtual machines on one physical computer. Type 1 is used for servers; Type 2 is handy for learning and testing.
7.4.1 Components
Component
What it is
Virtual machine (client)
a software computer: a virtual PC, or network devices such as a virtual switch or virtual router
Virtual machine (server)
a server OS running as a VM, e.g. web, email or database server
Type 1 hypervisor
‘bare metal’: installed directly on the hardware; efficient and secure; used in data centres (ESXi, Hyper-V, Proxmox)
Type 2 hypervisor
‘hosted’: an application on a normal OS; easy to set up; good for testing and learning (VirtualBox, VMware Workstation)
Virtual machines, switches and routers are software, but they behave like real devices, so whole networks can run inside one server.
Key features of virtual environments
Key feature (7.4.2)
Meaning
Increased security
each VM is separated, so a problem in one is contained; snapshots allow quick roll-back
Managed execution
the hypervisor controls and limits the CPU, memory and storage each VM can use
Sharing
one physical machine’s resources are shared between many VMs
Aggregation
resources from several physical machines can be pooled and managed as one
Emulation
a VM can imitate different hardware or operating systems
Isolation
VMs cannot see or affect each other
Portability
a VM is a set of files that can be copied, backed up or moved to another host
Benefits and drawbacks
Benefits (7.4.3)
Drawbacks (7.4.4)
cost-effective for large environments (fewer physical servers); easy central management; resilience (move VMs if a host fails); potentially lower carbon footprint (less hardware and power); improved disaster recovery (copy and restore VMs); better testing environments (safe, disposable copies); education and training (labs without extra kit)
extra load on the host hardware; slower execution than running directly on hardware; performance can be falsely represented, because a VM’s speed depends on the host and the other VMs sharing it
Key terms
Virtual machine
a software emulation of a computer.
Hypervisor
software that creates and manages virtual machines.
Snapshot
a saved state of a VM that can be restored.
Host / guest
the physical machine / the virtual machine running on it.
Exam focus
Balance benefits and drawbacks for the scenario. A school running 30 practice servers benefits hugely; a single high-performance game server may run better on bare hardware.
Think it through
Why is a type 2 hypervisor fine for a student but not for a data centre?
It runs on top of a normal OS, which adds overhead and another thing that can fail; a type 1 hypervisor runs directly on the hardware for better performance and security.
How does virtualisation improve disaster recovery?
Whole VMs can be copied or snapshotted and restored quickly on other hardware, instead of rebuilding a physical server from scratch.
7.5 Cloud environments T Level core · 7.5.1 – 7.5.3
Cloud types, benefits and delivery models
By the end you can: compare private and public clouds, explain the benefits of cloud, and describe IaaS, PaaS and SaaS and who manages what.
7.5.1 Types of cloud
Private cloud
Public cloud
Owned / used by
one organisation (on-site or hosted for it alone)
a provider shares its infrastructure with many customers (AWS, Azure, Google Cloud)
Control and security
full control; data kept separate
less control; security shared with the provider
Cost
high up-front cost; needs in-house skills
pay as you go; no hardware to buy
Scaling
limited by hardware owned
almost unlimited, very quickly
Benefits of cloud
Benefit (7.5.2)
Meaning
Portability
data and services can be reached from anywhere, on any device; workloads can move between locations
Elasticity
resources scale up at busy times and down when quiet, automatically
Fewer storage limitations
storage can grow as needed without buying disks
Cost effectiveness
pay only for what is used; no hardware, power or maintenance costs
The specification’s split of responsibilities. With SaaS the client only manages user accounts and data.
7.5.3 Cloud delivery models
Model
Example
Advantages
Disadvantages
IaaS
renting virtual servers (e.g. Azure VMs, AWS EC2)
most control and flexibility; no hardware to buy
client must manage and secure the OS and everything above it; needs skills
PaaS
a platform to deploy code (e.g. Azure App Service, Heroku)
developers focus on code; provider patches the OS and runtime
less control; may be tied to one provider’s platform
SaaS
ready-made software in a browser (e.g. Microsoft 365, Gmail)
no installation or maintenance; available anywhere
least control and customisation; depends on the provider and internet
Key terms
Elasticity
automatically adding or removing resources to match demand.
IaaS
infrastructure as a service.
PaaS
platform as a service.
SaaS
software as a service.
Exam focus
Learn the responsibility split exactly as in the specification: IaaS client manages application software, system software (middleware and OS), runtime, data and user accounts; PaaS client manages application software, data and user accounts; SaaS client manages only user accounts and data.
Think it through
An online shop gets 20 times more visitors on Black Friday. Which cloud benefit helps most?
Elasticity: extra servers are added automatically for the peak and removed afterwards, so the shop only pays for them while needed.
Which delivery model gives a developer the least server administration while still running their own code?
PaaS: the provider manages the OS, middleware and runtime; the developer manages the application and data.
7.6 Resilient digital environments T Level core · 7.6.1 · 7.6.2
Resilience: benefits and methods
By the end you can: explain the benefits of resilient environments and the methods used to improve resilience.
A resilient digital environment keeps working, or recovers quickly, when something goes wrong: a failed disk, a cyber attack, a power cut, a flood or a mistake by a member of staff.
Why resilience matters
Benefit (7.6.1)
Impact on organisations and clients
Increased security
data is protected in storage and in transfer; fewer vulnerabilities for attackers to exploit
Increased reputation
protects the brand and image; customers keep their confidence and stay
Reduction in downtime
services stay available, so sales, work and customer service continue
Methods to improve resilience
Method (7.6.2)
How it improves resilience
Software updates / upgrades
planned updates add features and fixes; patches close newly found vulnerabilities quickly
Hardware replacement
rolling replacement plans swap old kit before it fails; secure disposal wipes or destroys drives so data cannot be recovered
Data and system redundancy
duplicate disks (RAID), servers, power supplies and connections remove single points of failure
Device hardening
remove unneeded ports, applications, permissions and access to shrink the attack surface
Backup systems and recovery procedures
copies kept onsite (fast restore), remote/offsite (safe from local disaster) and in the cloud; tested recovery procedures
Hot, warm and cold sites
an alternative location to run from after a disaster
Standard operating procedures
effective staff training, induction for new starters, and training for new digital systems and new or updated policies
A standby site lets an organisation keep working after a disaster. The faster the recovery, the higher the cost.RAID levels. Orange blocks hold parity: extra data calculated from the other blocks, so a lost block can be rebuilt. RAID is not a backup.
In the real world: a ransomware attack
An online retailer’s servers are encrypted by ransomware. Because it kept offsite and cloud backups, patched its systems, and had a tested recovery procedure and a warm site, it restored service within hours instead of paying the criminals or closing for weeks.
Key terms
Resilience
the ability to keep working or recover quickly after a problem.
Patch
a software update that fixes a bug or vulnerability.
Device hardening
reducing the attack surface by removing anything not needed.
Redundancy
duplicate components that take over if one fails.
Hot site
a fully running standby location for immediate failover.
Exam focus
For ‘evaluate’ questions on resilience, weigh cost against risk: a hospital or bank justifies a hot site and full redundancy; a small café might rely on cloud backups and a cold plan.
Think it through
Why must old hard drives be disposed of securely?
They may still hold personal or business data, which could be recovered and misused, breaching data protection law and damaging reputation.
Give two examples of device hardening on a new server.
Close unused ports and services, remove unneeded software and default accounts, and give users only the permissions they need.
Watch
Data security in storage and transfer is a benefit of resilient environments (7.6.1).