Exam practice

LAN: local area network. WAN: wide area network. MAN: metropolitan area network. PAN: personal area network. NIC: network interface card. MAC: media access control. ISP: internet service provider. TCP/IP: Transmission Control Protocol / Internet Protocol. UDP: User Datagram Protocol. HTTP: Hypertext Transfer Protocol. HTTPS: Hypertext Transfer Protocol Secure. FTP: File Transfer Protocol. SMTP: Simple Mail Transfer Protocol. DNS: Domain Name System. DHCP: Dynamic Host Configuration Protocol. IPv6: Internet Protocol version 6. NAT: network address translation. VPN: virtual private network. URL: uniform resource locator.

Star: five computers, each with its own cable to a switch in the middle. Bus: one cable with a terminator at each end, the five computers joined to it (as in Figure 5.1).
Star. Advantages: a broken cable cuts off only one computer; easy to add a computer or find a fault. Disadvantage: if the switch fails, the whole network stops (and it needs more cable than a bus).
Bus. Advantages: needs the least cable; simple and cheap for a few computers. Disadvantage: one break in the bus cable stops the whole network, and faults are hard to find.

A hub repeats every message out of every socket, to every device. A switch learns which device (by its MAC address) is on which socket, and sends each message only out of the socket it is for.
With a hub, all the devices share its capacity and must take turns, they get in each other’s way, and every device can see every other’s traffic. A switch lets many pairs talk at once, at full speed, and keeps each message to its own device. Switches became cheap, so hubs are no longer made.

A protocol is a set of rules that devices follow to communicate: the form of the messages, their order, their meaning, and what to do when something goes wrong.
Application layer: HTTP (or DNS, SMTP, FTP, DHCP). Transport layer: TCP (or UDP). Internet layer: IP. Link layer: Ethernet (or Wi-Fi).

TCP sets up a connection, numbers every piece, checks each arrived, sends lost pieces again, and puts them in order: reliable but slower. UDP just sends each piece, with no checking or resending: faster, never waits, but pieces can be lost.

• i.Downloading a file: TCP, because every byte must arrive, in order; a file with one piece missing is broken.
• ii.A live video call: UDP, because a piece sent again would arrive too late to show; it is better to skip it and carry on.

• i.With the mask 255.255.255.0 the first three numbers are the network: 192.168.5.
• ii.No. 192.168.5.200 has the same network part, 192.168.5, so it is on the same network: the computer sends the packet straight to it, through the switch.
• iii.Yes. 192.168.6.14 has the network part 192.168.6, a different network, so the computer sends the packet to its default gateway, the router at 192.168.5.1, which passes it on.

Leading zeros off: 2001:​db8:​0:​0:​0:​ff00:​42:​8329; then the run of three zero groups becomes a double colon: 2001:db8::ff00:42:8329.

Private addresses are three ranges (10.x.x.x, 172.16.x.x to 172.31.x.x, and 192.168.x.x) set aside for use inside networks, never on the internet. Many networks use the same ones at the same time without trouble.
NAT: each device in the home has its own private address, given by the router’s DHCP. When a device sends a packet out, the router replaces the private address with the home’s one public address, and remembers which device sent it (and to which app). When the answer comes back to the public address, the router looks up its record, puts the device’s private address back, and passes the answer to that device. To the internet, the whole home looks like one device.

1. The phone asks its DNS resolver (usually run by the internet company) for the address of moecdc.gov.np.
2. If the resolver has no copy, it asks a root server, which answers: ask the .np servers, at these addresses.
3. It asks a .np server, which answers: ask moecdc.gov.np’s own servers, at these addresses.
4. It asks moecdc.gov.np’s server, which gives the address, and how long it may be kept.
5. The resolver passes the address to the phone, and keeps a copy.

Caching is keeping copies of answers for the time they allow. The next request for the same name, from anyone using that resolver, is answered at once from the copy, without asking the root, .np or the domain’s servers. Most lookups are answered from caches in a few milliseconds, and the root servers are asked only rarely.

1. DNS lookup (application layer): the browser finds the address of moecdc.gov.np, from a cache or by asking the resolver.
2. TCP connection (transport layer): the browser and the server set up a connection with the three-way handshake.
3. TLS (between the transport and application layers): because of https, they agree on a secret key, and the server shows its certificate to prove its name. Everything after is encrypted.
4. HTTP request (application layer): the browser sends GET for the page, with the Host moecdc.gov.np.
5. HTTP response: the server sends the status 200 OK and the page’s HTML.
6. The browser asks for the other files the HTML names (styles, pictures, scripts) and draws the page as they arrive.

All of these messages travel as IP packets (internet layer) across Wi-Fi, cables and fiber (link layer), router to router.

On an http:// page, the password travels as plain text: anyone on the way, on a café’s Wi-Fi or at an internet company, could read it, and could even change the page you see. On https:// it is scrambled on the way (though the padlock still doesn’t prove the website is honest, Grade 6).

1. Owner: a school’s LAN is owned and run by the school; the internet is tens of thousands of networks with separate owners, and nobody owns the whole.
2. Rules: the school’s administrator decides everything in its LAN; the internet is held together only by open protocols agreed by all and each network’s own policies.
3. Size and membership: the LAN covers a building and only the school’s devices; the internet covers the world and anyone can join through an internet company.
4. Addresses and routes: the LAN uses private addresses chosen by the school and simple routes; the internet uses public addresses handed out by regional registries, and networks tell each other their routes by BGP.

An intranet is a private network built with the internet’s tools (web pages, email) for an organization’s own people only; an extranet lets chosen outsiders, such as suppliers or partner schools, into part of it with a login.

Count sockets: a router has only a few. Start from the fiber box and the router, and use switches for the rooms.

A diagram like Figure 5.10: internet company → fiber box (ONT) → router; router → a 16- or 24-port switch in the lab (cable), joining the 10 computers and the printer; router → a small switch in the office, joining the 2 computers and the file server; router (or a switch) → an access point by cable, with teachers’ laptops and phones joined by Wi-Fi. Key: solid lines for cables, dashed for Wi-Fi, a colored line for fiber.
The router runs a DHCP server (giving every device a private address), NAT (sharing the school’s one public address) and a firewall (blocking traffic nobody asked for).
Client–server suits it: the file server keeps the office’s and teachers’ files in one place, where they can be backed up and access controlled by one person; with this many computers, peer-to-peer sharing would be hard to keep track of and to back up.