The internet

Structure of the internet, DNS, TCP/IP stack and Application layer protocols

(Structure of) The Internet

A global network of interconnected networks that communicate using a common set of standards and protocols

Individuals and organisations connect to the internet through an …

… internet service provider (ISP)

Edge router

A router that links one network to another

Subscriber router

• Small scale, low-cost devices with two interfaces (one for each network).

• Often provided to homes and small businesses.

• Type of edge router

Enterprise router

• Used to connect large businesses and ISP networks to the internet.

• Very powerful and expensive

• They are a type of edge router

Router

A networking device that forwards data packets between computer networks

Core router

• Routers that are part of the internet backbone

• These routers have multiple interfaces, all of which can work at the highest speed simultaneously

Components of an IP packet

header, payload and trailer

Explain and describe packet switching

• A method of communicating data across a network

• Data is split into chunks called packets which have labels called headers.

• The header includes the IP address being sent to and order of the packets.

• Each packet is sent on the most convenient / avoidable route meaning they may arrive in a different order to which they were sent

• Once packets arrive at the receiver they are reordered

A hop

A router-to-router link

How do routers determine the best 'next hop'

They use a routing table (set of rules) to decide what to do with incoming packets. Each packet is treated individually and a separate routing decision is made for each one.

Time to live (TTL) counter in a packet header

• Used to prevent packets being infinitely passed in a loop by routers

• The value is initially set when the packet is created, and reduced by one every time it goes through a router.

• If the counter reaches zero, the packet is discarded.

Type of Service (ToS) field in the packet header

• Makes it possible to mark packets with a priority level

• Allows packets to request to be placed at the front of the queue of packets to be routed

• Routers may choose to implement or ignore these requests.

Circuit switching

• A method of communication over a network.

• A direct link is created between two devices and a fixed bandwith is reserved for the duration of the link. Data can then be transferred for the of duration of the link.

• Sometimes used for phone calls

What is the key requirement of circuit switching?

The transfer and download rates must be identical on both devices.

What is the domain name system (DNS)?

A hierarchical system used to name internet resources and find the IP address associated with a given domain name

DNS is the abreviation of …

Abbreviation of domain name system

What is a DNS lookup?

The process of getting the IP address associated with a domain

Sequence of events in a DNS lookup for a URL (e.g. https://www.raspberrypi.org/learn/index.html)

• Browser sends the hostname part of the URL (e.g. www.raspberrypi.org) to a domain name system (DNS) server.

• The DNS server checks for the hostname in its lookup table

• The DNS server sends the corresponding IP address back to the browser

Sequence of events to load a URL (e.g https://www.raspberrypi.org/learn/index.html)

• Browser sends the hostname part of the URL (e.g. www.raspberrypi.org) to a domain name server. This is usually provided by your ISP.

• The DNS server checks for the hostname in its lookup table

• The DNS server sends the corresponding IP address back to the browser

• The browser sends a request for the resource identified in the path (/learn/index.html) to the web server located at the IP address

• The server then returns the resource to the browser

Explain how the domain name system plays a role in the websites being loaded.

• DNS is used to resolve IP corresponding to URL

• Request sent to DNS resolver.

• Resolver checks its cache and if it doesn’t hold the URL, it passes it in to the TLD Name server which checks its cache and returns the answer or passes on to the authoritative Name Server.

• The IP address is returned back up to the requesting client.

• Or an error if no resolution can be found.

TLD is the abreviation of …

Abreviation of top level domain

2LD is the abreviation of …

Abreviation of second level domain

What is the benefit for humans of using domain names/ URLs instead of IPs?

Domain names/URLs are much easier to remember

Recursive name server

• A type of DNS server that handles a high volume of DNS queries and will store a lot of IP addresses in cache (but not all).

• It returns the IP address corresponding with a given domain name if it has it stored. If not, it performs a full DNS lookup to find the associated address and then returns this

DNS root servers

• DNS servers at the root of the internet controlled by ICANN.

• Stores the IP of all TLD servers

• It returns the IP address of the TLD server needed to the recursive name server

Top level domain (TLD) server

• Stores the IP address of the authoritative name servers for each of its registered domains

• It returns the IP address of the authoritative name server associated with the domain to the recursive name server

What does the authoratative name sever do?

It returns the IP address of the webserver to the recursive name server.

URL is the abreviation of …

… abreviation of uniform resource locator

Uniform resource locator(URL)

A unique address for a webpage or other internet resource

Techniques networks use to exchange data

circuit switching and packet switching

Advantages of packet switching

• Packet switching can transfer data over very large networks to allow communication globally

• There are multiple routes that can be used between devices, so if one path breaks, another can be used

• Packets have methods to check that data arrives intact, such as checksums

Disadvantages of packet switching

• Time is spent deconstructing and reconstructing the data packets at every router

• Must wait for all packets to arrive before data can be received.

Advantages of circuit switching

• Data arrives in a logical order which results in a quicker reconstruction of the data.

• Enables two users to hold a phone call without delay in speech.

Disadvantages of circuit switching

• Bandwidth is wasted during periods of time in which no data is being sent.

• It uses switches which means electrical interference may be produced which can corrupt or destroy data.

• It can tie up sections of the network which cannot be used by others until transmission has been completed

• Devices must transfer and receive data at the same rate.

4 layers of the TCP/IP protocol stack (top to bottom)

• Application layer

• Transport layer

• Internet layer (aka network layer)

• Link layer (aka data link layer)

Name of data block produced by the application layer

Application data

What is the name of data block produced by the transport layer?

Segments (e.g. TCP segment) or datagrams (UDP datagram)

Name of data block produced by the internet layer.

Packets

Name of data block produced by the link layer

Frames (e.g. Ethernet frames)

IP packet trailer

Contains a checksum and a cyclic redundancy check. These pieces of metadata are used to detect whether any errors have occurred during transmission.

Application layer

• Based at the top of the TCP/IP stack

• Specifies what protocols need to be used to relate the application to what it’s being used for

• For example, if the application is a browser then it would select a protocol such as HTTP, POP3, FTP

What happens at the application layer?

• The protocol used is decided based on the application, e.g. HTTPS for browser based service, SMTP / IMAP for messaging service

• When sending, it passes data on to transport layer

• When recieving, it gets data from transport layer and unpacks the data ready for display.

What does the application layer do on the recipient device?

• The transport layer passes data to the application layer

• The application layer unpacks the data and presents the data in the form it was sent

Transport layer

• Establishes an end-to-end connection between the source and recipient computers.

• Can provide either a reliable or unreliable delivery service using TCP or UDP respectively

• When sending, it splits the data into segments and adds the the source and destination port numbers to the segment headers

• When recieving, it removes the segment headers and reassembles the segments back into the original data

TCP is the abreviation of …

abreviation of transmission control protocol

Transmission control protocol

• ‘Reliable’ protocol - ensures none of the data is lost

• Splits data into TCP segments which are often refered to as TCP packets

• Labels TCP segments with their sequence number, the total number of segments the original data was split up into and the port number being used for communication.

• The recieving end acknowledges each segment it recieves.

• If the sender does not receive an acknowledgement, the segment has been lost, and so is resent

UDP is the abreviation of …

abreviation of user datagram protocol

What does the transport layer do on the recipient’s device?

• The internet layer passes data up to the transport layer

• The transport layer removes the segment headers and reassembles the segments.

• It then passes the data up to the application layer

User datagram protocol

• ‘Unreliable’ protocol - doesn’t ensures none of the data is lost

• Used when the speed of data transfer is the most important criteria, and the application can tolerate some data loss

• UDP is often used by video calling applications and media streaming services

Port number

An addressable location on a network that links to an application (or process). They are used to return data to the corret place (correct application).

Socket (address)

• Combination of an IP address and a port number.

• Facilitate asynchronous communication - a socket is designated the source or the destination, depending on which side (client or server) is transmitting the data

• Sockets are used to specify which device packets must be sent to and the application being used on that device

Well-known port numbers

• Servers use well-known port numbers so they do not need to be looked up by the client that is making a request

• These numbers lie in the range 0-1024

Network/Internet layer

• IP is the standard protocol used

• Segments from the transport layer are encapsulated by the IP protocol to create IP packets

• Each packet is given a header to assist with the delivery of data. The most important header data is the scource and destination IP addresses.

• Routers operate at this layer

Link layer

• Bottom of the TCP/IP stack

• When sending, it receives data from internet layer. It adds the source and destination MAC addresses to the packets, then transmitts the data across the physical network (e.g. using cables or wirelessly)

• When recieving, it removes the source and destination MAC addresses and passes the data back up to internet layer

What does the network layer do when it receives the data from the link layer?

• It removes the packet header and trailer (which contain information such as the sender and recipient IP addresses)

• It passes the data up to the transport layer

What does the link layer do when it receives data from the sender device?

It removes the source and destination MAC addresses and passes the data up to the internet layer

HTTP(s) is the abreviation of …

abreviation of hypertext transfer protocol (secure)

Hypertext transfer protocol secure

The encrypted version of HTTP

What is an IP packet header?

The label at the front of a packet which contains numerous fields, including:

• The sender and recipient IP addresses

• Protocols being used

• Order of the packets

• The time to live / hop limit

• (Type of service is a bonus)

Payload

the raw data being transmitted