Exam 1 essays Flashcards
1
Q
Edge vs. Core - understand the concept to describe it, sketch it, and explain it
A
- Edge: end systems, access networks, links
- hosts: clients and servers (servers are often in data centers)
- core: packet switching, circuit switching, network structure
- interconnected routers, network of networks
- key functions: routing and forwarding
2
Q
what are the ways that residents access the internet? Describe. Draw a schematic.
A
Point to Point access:
- DSL (digital subscriber line): use existing telephone line to central office DSLAM
- Dialup via Modem: can’t surf the web and be on the phone at the same time
- Cable Modems: network of cable and fiber that attaches homes to ISP router, the home shares access to the router, available via cable TV companies
- Home Network: router connects to wireless access point and creates own LAN point
3
Q
What are the ways that companies access the internet? Describe.
A
- Enterprise access network (ethernet): end systems connect to ethernet switch that are connected to institutional mail, web servers, and institutional routers and ISP’s
- Companies also can access the internet wirelessly through access points at a base station
4
Q
Describe three types of guided, physical media. Which is best for which kind of situation? Why?
A
Transmits signal out of computer onto something you can touch
- Coaxial Cable: two concentric copper conductors that are bidirectional and broadband capable to hold multiple channels on one single cable (LAN use)
- Fiber Optical Cable: glass fiber that carries light pulses through high speed operation point to point transmission, there is also low error rate that spaces repeaters far apart and are immune to electromagnetic noise. greater security, higher bandwidths
- Twisted Pair: two insulated copper wires that is the most popular network cable used in data networks due to lower cost
5
Q
Describe two types of unguided, wireless media. Which of these is best for which kind of situation and why?
A
Signals propagate freely
- Radio: signal that is carried through electromagnetic spectrum with no physical wire, bidirectional (terrestrial microwave, Lan, wide area, satellite)
- Wireless (wifi) or Wide Area Cellular: requires no physical wire, all in the “air” come in directional (point to point) or omnidirectional (waves in all directions)
6
Q
Explain the difference between circuit switching and packet switching. Draw a diagram of each. Which one is more efficient? why? what are silent periods?
A
- Packet switching is more efficient and shares more than one pipe. This type allows the use of many different routers More durable and efficient. not reliable for interaction of real time (skype) most common with email
- circuit switching is typically on the phone. only use pipe the is completely dedicated to your conversation. less efficient.
- silent periods: in circuit switching, reduces the mean bandwidth per voice flow and reduces the interferences with other transmitters
7
Q
Explain the difference between TDM and FDM. What do these stand for? Draw a diagram of each. which one is more commonly used in networks? what is the purpose of multiplexing in the first place?
A
- Time Division Multiplexing (TDM): take signals and squish them onto one pipe using time. used in organizations moving big amounts of data
- Frequency Division Multiplexing (FDM): frequency waves on AM/FM that take signals and add/subtract frequencies to allow more signal on the same wire/space
- multiplexing is used because it is cost effective to a business
8
Q
Draw the TCP/IP layer model and explain each layer in detail using the handout in class that has the virtual circuit on it. What layer does this virtual circuit appear on? why is it called a virtual circuit?
A
- Application: the users access to the network. uses application software. ex. internet explorer, skype, outlook. Defines the message on the internet
- Transport: links application layer to network layer. if we want to be reliable it will set up a virtual circuit. establishes an end to end connection. Breaks long messages up into small ones and numbers them, detects lost messages and resends them
- Network: in charge of routing and address finding
- Data Link: moves packet from one router to the next. controls the physical layer. formats the start and stop of messages. corrects errors.
- Physical: the physical connection between the sender and receiver
- Virtual Circuit: not real but appears physical to the user. on the Transport layer
9
Q
How does TCP/IP model compare to the OSI 7 layer model? what are models used for?
A
- OSI 7: transport layer guarantees the delivery of packets, follows horizontal approach, has a separate presentation layer, provides both connectionless and connection oriented service, has 7 layers
- TCP/IP: the transport layer does not guarantee delivery of packets, vertical approach, does not have a separate presentation layer, network layer provides connectionless service, 5 layers
- models are used as a reference tool and also to five a visual of how the system works
10
Q
What are the 4 causes of delay? Explain in detail the 4 causes of delay and use a diagram to do this. What is end to end delay made up of? Are there other places that we can find delay in networks too?
A
- Processing: associated with how sophisticated your processor is in your switch
- Queueing: waiting to get onto a pipe. delay happens when to many people are trying to access a pipe at once
- Transmission: how quickly or slowly the actual pipe runs. copper = slow
- propagation - physical distance to someone
- end to end is made up of processing, transmission, and propagation
- bad wifi connection can cause delay
11
Q
What is the role of packet switching in statistical multiplexing?
A
statistical multiplexing: giving priority of the time slots on the pipe to whoever uses it the most. the people on the most get more time slots
12
Q
What is a regional ISP vs. a Tier 1 ISP? What is the PoP? Describe how ISP’s hand off packets amongst each other. Draw a sketch to help us understand this issue. What is peering?
A
- regional ISP: where their ISP’s in the region connect. Each regional ISP then connects to tier 1 ISPs
- tier 1 ISP: similar to global transit ISP, but Tier 1 do not have a presence in every city in the world
- Pop: a group of one or more routers at the same location in the providers network where customers ISP can connect to provider ISP
- Peering links: directly connect their networks together so that all of the traffic between them passes over direct connection rather than through upstream intermediaries
- Internet exchange point: a meeting point where multiple ISPs can peer together
13
Q
Why and how do packets get dropped in networks? What do we do if we drop a packet?
A
- a queue preceding link in buffer has a finite capacity and a packet arriving to a full queue gets dropped
- UPD: you will not know if you lost a packet
- TCP: packets get numbered and if one is missing it will get resent
14
Q
Give 4 places where bottlenecks might occur in a sample network and draw a sketch of this to help us understand this
A
- bottleneck: a point where the flow of data is impaired or stopped entirely
15
Q
Draw a picture of a packet sniffer and explain how it work
A
- a promiscuous network interface reads and records all packets passing by
16
Q
What is an application layer message?
A transport layer message?
A network layer datagram?
A link layer frame?
A
application: sent from the sending host to the transport layer
transport: encapsulates the application layer message
network: adds source and destination end system addresses
link layer: the continuing datagram picks up link layer header information and creates a link layer frame
17
Q
What is a botnet?
What is a DoS attack?
What is a DDoS attack?
A
- botnet: compromised host may also be enrolled in a network of thousands of similarly compromised devices, collectively known as a botnet
- DoS Attack: a broad class of security threats, it renders a network, host, or other piece of infrastructure unusable by legitimate users
- DDoS Attack: the attacker controls multiple sources and has each source blast traffic at the target
18
Q
What is the difference between a virus, a worm, and a trojan horse?
A
- Virus: a malware that requires some form of user interaction to infect the user’s device ex. clicking on an email attachment
- Worm: a malware that can enter a device without any explicit user interaction
- Trojan Horse: a program that claims to rid your computer of viruses but instead puts viruses on it
19
Q
What is analog?
What is digital?
How can we convert between analog and digital?
What is the sampling rate and how does that relate to the quality of a replicated digital bit stream with something like a CD? Skype?
A
- Analog: your voice vibrates a wire which causes the signal and electrons to move down the wire in waves (vinyl record)
- Digital: converts analog to binary using sampling
- Convert: Sampling a voice in slots of time (CD = 8,000/sec) and then quanticizes it and then digitizes it and then sends the 0s and 1s down the wire and mimics the original voice
- Sampling rate: the average number of samples obtained in one second
- CDs have a higher sampling rate of they are better quality than skype
20
Q
What are some common status codes of HTTP?
A
- 200 ok
- 301 moved permanently
- 302 not found
- 304 not modified
- 400 bad request
- 401 unauthorized
- 403 forbidden
- 404 not found
- 505 http version not supported
21
Q
Draw a detailed chart of how to keep user state with cookies and explain that and how it works. How does it work with a shopping cart?
A
How to keep state
- Protocol endpoints: maintain state at sender/receiver over multiple transactions
- cookies: http message carry state. they permit sites to learn a lot about you (ex. name, email, card number)
- shopping carts: your cookies remember what items you put in your shopping cart
22
Q
Explain the difference between a port and a socket. Draw and explain what process, socket and then how the underlying transport protocol choice is made.
A
- Port: a logical connection place, specifically, using the internet’s protocol, TCP/IP, the way a client program specifies a particular server program on a computer in a network
- Sockets: process sends/receives messages to/from its socket
- Socket to door analogy:
1. sending process shoves message out door
2. sending process relies on transport infrastructure on other side of door to deliver message to socket at receiving process
23
Q
What is the difference between a client server architecture and a peer to peer architecture? Which one is better for what types of applications and why?
A
- client server: everything is located and controlled by a server with a permanent IP address. It is very secure but also expensive. used with big networks or sensitive data
- peer to peer: no central server with everything on it. each of the peers have a piece of information so no one has all of it. fast but very poor security. used for content delivery or file sharing networks
24
Q
Describe P2P architecture, discuss some of the pluses and minuses of it, then give 3 commercial examples of a P2P architecture.
A
- P2P: no central server with everything on it. IP addresses can change and different peers can come and go (churn). each peer has a piece of information so no one has all of it
- pluses: move files between computers efficiently. faster than client server
- minuses: security is awful. no way to password protect.
- 3 examples: BitTorrent, kankan, skype
25
Define and describe TCP and UDP. Which one is better for which types of applications
- Transmission Control Protocol: there is a reliable transport so if not all packets come through, it will ask to resend them. Has flow and congestion control so it wont get overloaded. more secure. better for email and bank transactions
- User Datagram Protocol: establishes loss tolerant connections between the internet and application. Used for YouTube videos when you don't care if some packets get lost and decreases quality. Or when you want to send data quickly
26
Know what applications use TCP and what applications use UDP and why. What layer of the TCP network model is where TCP or UDP resides?
- TCP and UDP are located on the transport layer of the TCP network model
- TPC: Email, Remote terminal access, web, file transfer
- Ether: streaming multimedia, internt telephony
27
Give an overview of an http request and response cycle
- hypertext transfer protocol (HTTP): is the webs application layer protocol
- client initiates TCP connection to server port 80. server accepts connection. HTTP messages are exchanged. TCP connection closes.
28
How is the stateless nature of HTTP worked around on the internet? Describe cookies and how they work, give a chart to demonstrate this.
- stateless is when the server does not maintain any information about requests from past clients
- cookies are used to create state on the server. They will help store the information
29
What is persistence and non persistence in HTTP? Draw a picture showing packets flowing with each version. Which is better and why?
- persistent: one connection left open after a server responds. HTTP messages that are sent back and forth are sent over the same connection. the client can send requests as soon as it encounters referenced object and all objects can return one request
- non persistent: initiation and acceptance for every step. the server closes the TCP connection when client gets response. whole process must be completed if client wants to request something else
- most of the time http is persistent because it is easier to keep the same connection
30
Describe how web caching works
1. user/client goes to visit a website
2. requests are sent to the web cache
3. if the object is in the cache then the cache will return the object to the client
4. if the object is not in the cache then the cache will request it from the origin server, the return it to the client
31
Describe how caching is updated. why is this diagram used? what is happening with http for caching? how does a conditional GET work?
- cached information usually lasts for 2 days and updated as servers process replies
- this design is used for cost and delay. without the cache all requests would need to be sent to the origin servers from clients. to get a faster link would cost too much and without it there would be a huge delay
- when the cache doesnt know something it asks the server, http is the protocol used to transfer this information
- conditional GET: goal is to not send an object to the cache if it has the most updated version (IF MODIFIED SINCE)
32
Describe FTP, how it works, how its different than http, and draw a diagram of it. Does it use TCP or UDP? Why? Discuss in and out of band with respect to FTP. Does it maintain state? why or why not?
- File Transfer Protocol (FTP): application layer protocol used to push data (whole movies). uses two pipes. 1st pipe does the "handshake" the 2nd pipe pushes the data through
- http does not have two pipes
- uses TCP so large files get transmitted properly
- out of band: FTP uses a seperate control connection from the file
- in band: http sends request and response in same connection
- FTP must maintain state. Server must associate connection with an account and keep track of their directory
33
Draw and end to end connection of an STMP email connection between Alice and Bob and label the protocols used along the way. When Bob pulls his mail off his server, what protocols can he use?
- STMP = simple mail transfer protocol
1. Alice uses user agent to compose a message to bob@school.edu
2. As user agent sends message to mail server, message places in queue
3. client side of STMP opens TCP connection with Bs mail server
4. STMP client sends As message over TCP connection
5. Bs mail server places the message in Bs mailbox
6. B invokes his user agent to read the message
- he can use PoP3, IMAP, or http
- email wasn't designed to do what it does. they are interlaced because of the result of the popularity of email and new features
34
What is the purpose of interlacing these several different protocols. is email push or pull technology? what does MIME stand for? how and why is it used? which is better PoP or IMAP?
- push and pull: STMP pushes emails to servers and PoP3, IMAP, and Web-Based email pull the email from servers
- Multipurpose Internet Mail Extensions: acts as an internet standard for email. content is encoded when it is sent out and decoded once it get to its destination
- IMAP is better than PoP3 because it keeps all messages on one place on a server, allows user to organize messages in folders
35
What are the benefits of P2P? Do these benefits outweigh the costs? why or why not?
- benefits: everyone doesn't get an entire file. reduces traffic and delay. works faster than client server.
- benefits do not outweigh the cost because security is so poor
36
Describe how skype works using a diagram. is this a good product. why or why not? in your view. discuss how BitTorrent works with a diagram. What is a Distributed Hash Table and why might we need this?
- skype uses VoIP and P2P but uses firewalls and protection
- skype is a good product because you generally talk to people who are relatively close to you which is what P2P excels in. Also, it is relatively safe because you know who you are connecting with
- BitTorrent: gives a user a tracker to get chunks of a file from people close by
- Distributed Hash Table: used in P2P provides a lookup system with key value pairs. this is needed so people don't disturb other peoples downloads
