Content Distribution

Question 1

HTTP Request Line

Answer 1

Method (verb, POST HEAD etc.), URL, Version

Question 2

Common HTTP headers

Answer 2

Referrer, User Agent

Question 3

HTTP response status line

Answer 3

httpversion, response code, location , server, Allow, content-encoding, content-length, last modified

Question 4

HTTP 100s

Answer 4

Information

Question 5

HTTP 200s

Answer 5

Success

Question 6

HTTP 300s

Answer 6

Move, redirect

Question 7

HTTP 400s

Answer 7

Error 404 file not found

Question 8

HTTP 500s

Answer 8

Server Error

Question 9

How many TCP connection for request and response in early HTTP

Answer 9

1

Question 10

Problem with Multiple short lived TCP connections

Answer 10

3-way handshake, slow start, servers in TIME_WAIT

Question 11

Solution for the One TCP connection problem

Answer 11

Persistent Connections

Question 12

What are presistent connections

Answer 12

Multiple request/response on a single TCP connection (multiplexed), Delimiters indicate the end of request, Content length

Question 13

TIME_WAIT

Answer 13

he timeout in TIME_WAIT is just an amount of time after which we can safely assume that if the other end didn’t send anything, then it’s because he received the final ACK and closed the connection.

Question 14

What is pipelining in persistent Connections

Answer 14

Client send request as soon as it encounters a reference object

Question 15

Where can Caching occure

Answer 15

Client can cache, networks can cache and servers can cache

Question 16

CDN

Answer 16

Content Distribution network

Question 17

How does 304 relate to cache?

Answer 17

stands for NOT MODIFIED. The client should use the locally cached version.

Question 18

Can DNS return near by servers

Answer 18

Yes, dig show this info

Question 19

What is CDN

Answer 19

Overlay network of caches (Deliver content to client from optimal location) Geographically disparate servers

Question 20

Challenges in running a CDN

Answer 20

How?, Where?, HOw to find? How to choose server? how to direct clients?

Question 21

Goal of CDN

Answer 21

replicate content on many servers

Question 22

Server Selection

Answer 22

Which server? Lowest Load, Lowest Latency, Any “alive” servers

Question 23

Out of server selection which option is normally choosen

Answer 23

Lowest Latency

Question 24

Content routing

Answer 24

How to direct clients to a server

Question 25

Types of Content Routing

Answer 25

Routing (e.g. anycast (course)), application based (http redirect (delays)), Naming-based (eg DNS (fase))

Question 26

Benefits CDNs peer w/ISPs

Answer 26

better throughput (lower latency), redundancy, burstiness (lower transit cost)

Question 27

ISP peer w/CDNs

Answer 27

good performance for customers, lower transit costs

Question 28

Bittorrent

Answer 28

file sharing, large file distribution, fetch content from peers, file split among peers

Question 29

Seeders

Answer 29

Have a full copy of the file "Create the initial copies"

Question 30

leechers

Answer 30

Only have partial copy of the file

Question 31

freeloading

Answer 31

Disconnecting after the file has finished downloading

Question 32

Bittorrent solution for freeloading

Answer 32

Chocking

Question 33

Chocking

Answer 33

temporary refusal to upload chunks. If can't download from a peer don't upload to it. Eliminates freeloading problem

Question 34

How does bittorrent ensure chunks get swapped out

Answer 34

Rarest Piece first

Question 35

What s Rarest Piece first

Answer 35

Determine which pieces are most rare among clients download those first

Question 36

Since rarest piece may not be available initially what is used

Answer 36

Random piece first

Question 37

What is bittorrents end game?

Answer 37

Actively request missing pieces from all peers

Question 38

Distributed hash table is a form of?

Answer 38

Structured content overlay

Question 39

Types of distributed hash tables

Answer 39

CHORD consistent hashing

Question 40

CHORD

Answer 40

scalable distributed "lookup service" Key->value

Question 41

Benefits of chord

Answer 41

Scalable, Provable correctness, performance

Question 42

Chord Motivation

Answer 42

Scalable location of data in a large distributed system

Question 43

What is the key problem of chord?

Answer 43

key lookup

Question 44

Main idea of Consistent Hashing

Answer 44

Keys and nodes map to same ID space

Question 45

Two options when implementing consistent hashing

Answer 45

Option1: Every node know location of every other node. Lookups: O(1). Tables: O(N). Option2: Node knows only successor Table O(1), lookups O(n)

Question 46

chords solution to the key lookup problem

Answer 46

Finger table. (Read the chord paper)