Chapter 3 Application Layer Functionality and Protocols CIS 81 Networking Fundamentals Rick Graziani Cabrillo College [email protected] Last Updated:

Chapter 3Application Layer Functionality and Protocols

CIS 81 Networking Fundamentals

Rick Graziani

Cabrillo College

[email protected]

Last Updated: 2/23/2008

2

This Presentation

For a copy of this presentation and access to my web site for other CCNA, CCNP, and Wireless resources please email me for a username and password. Email: [email protected] Web Site: www.cabrillo.edu/~rgraziani

3

Note

This presentation is not in the order of the book or online curriculum. This presentation also contains information beyond the curriculum.

Applications: Interface Between the Networks

5

Application Layer: OSI and TCP/IP Models

The Application layer, Layer seven, is the top layer of both the OSI and TCP/IP models.

Provides the interface between the applications we use to communicate and the underlying network.

6

Application layer protocols are used to exchange data between programs running on the source and destination hosts.

There are many Application layer protocols and new protocols are always being developed.

Email

HTTP (www)

HTTP HTTP

7


Functionality of the TCP/IP application layer protocols fit roughly into the framework of the top three layers of the: OSI model: Application, Presentation and Session layers.

Most early TCP/IP application layer protocols were developed before the emergence of: personal computers, graphical user interfaces and multimedia objects.

These protocols implement very little of the functionality that is specified in the OSI model Presentation and Session layers.

8

The Presentation Layer

The Presentation layer has three primary functions: Coding and conversion of Application layer data to ensure that data

from the source device can be interpreted by destination device. Compression of the data in a manner that can be decompressed by

the destination device. Encryption of the data for transmission and the decryption of data upon

receipt by the destination. Compression and Coding formats:

Graphics Interchange Format (GIF) Joint Photographic Experts Group (JPEG) Tagged Image File Format (TIFF).

9

The Session Layer

Create and maintain dialogs between source and destination applications. Handles the exchange of information to:

initiate dialogs keep them active restart sessions that are disrupted or idle for a long period of time

Most applications, like web browsers or e-mail clients, incorporate functionality of the OSI layers 5, 6 and 7.

10


Common TCP/IP Protocols Domain Name Service Protocol (DNS) is used to resolve Internet

names to IP addresses. Hypertext Transfer Protocol (HTTP) is used to transfer files that make

up the Web pages of the World Wide Web. Simple Mail Transfer Protocol (SMTP) is used for the transfer of mail

messages and attachments. Telnet, a terminal emulation protocol, is used to provide remote access

to servers and networking devices. File Transfer Protocol (FTP) is used for interactive file transfer

between systems.

Note: Usually a single server will function as a server for multiple applications

11

RFCs: Request For Comments

The protocols in the TCP/IP suite are generally defined by Requests for Comments (RFCs). Maintained by IETF (Internet Engineering Task Force) There are a few in there for fun -

ftp://ftp.rfc-editor.org/in-notes/rfc1882.txt.

12

Application Layer Software

Network-Aware Applications Applications are the software programs used by people to

communicate over the network. They implement the application layer protocols and are able to

communicate directly with the lower layers of the protocol stack. Email Clients Web Browsers

User applications

Services

System Operations

Within the Application layer, there are two forms of software programs or processes that provide access to the network: applications services

13


Application layer Services Other programs may need the assistance of Application layer services

to use network resources such as: File transfer Network print spooling

These services are the programs that interface with the network and prepare the data for transfer.

User applications

Services

System Operations

14


Application layer uses protocols that are implemented within applications and services. Applications provide people a way to create messages. Application layer services establish an interface to the network. Protocols provide the rules and formats that govern how data is

treated.

Bottom line: When discussing an application like "Telnet" we could be referring to the

application, the service, or the protocol.

15

Application Layer Protocol Functions

Application layer protocols are used by both the source and destination devices during a communication session.

The application layer protocols implemented on the source and destination host must match.

Protocols: (This will become clearer later! Herding cats.) Establish consistent rules for exchanging data. Specify the structure and type of messages that are exchanged.

Types: Request, response, acknowledgement, error message, etc. Defines the dialogues, ensuring with transmissions met by expected

responses, and with the correct service invoked.

16

Application Layer Protocol Functions

Applications and services can use multiple protocols. Encapsulate the protocol or encapsulated by this protocol Invoke other protocols

Using a web browser (HTTP): May invoke:

DNS, ARP, ICMP May use:

TCP, UDP, Ethernet, PPP Uses

IP

17

Client Server Model

Client: the device requesting the information Server: the device responding to the request is called a server. The client begins the exchange by requesting data from the server. Server responds by sending one or more streams of data to the client. In addition to the actual data transfer, this exchange may also require

control information, such as: user authentication the identification of a data file to be transferred

18

Servers

A server is usually a computer that contains information to be shared with many client systems. Web server Email server File or database server Applications server

Some servers may require authentication of user account information and vary permissions.

Example, if you request to upload data to the FTP server, you may have permission to write to your individual folder but not to read other files on the site.

19

Servers

The server runs a service, or process, sometimes called a server daemon. Daemons (like other services) typically run in the background and are not

under an end user's direct control. Daemons are described as "listening" for a request from a client.

Programmed to respond whenever the server receives a request for the service provided by the daemon.

When a daemon "hears" a request from a client: It exchanges appropriate messages with the client, as required by its

protocol, Proceeds to send the requested data to the client in the proper format.

20

Application Layer Services and Protocols

Servers typically have multiple clients requesting information at the same time.

For example, a Telnet server may have many clients requesting connections to it. These individual client requests must be handled simultaneously and

separately for the network to succeed. The Application layer processes and services rely on support from lower

layer functions to successfully manage the multiple conversations.

Application Layer Protocols

22

HTTP (WWW)

FTP

SMTP (email)

Telnet

(file transfer)

(remote login)

DHCP(IP address resolution)

DNS

(file sharing)P2P

(domain name resolution)

(file sharing)SMB

We will examine HTTP in detail.

23

DataHTTP Header

TCP Header

IP Header

Data Link Header

Data Link Trailer

IP PacketData Link Header

Data Link Trailer


Data Link Trailer


Data Link Trailer


Data Link Trailer


Data Link Trailer


Data Link Trailer

DataHTTP Header

TCP Header

IP Header

Data Link Header

Data Link Trailer

Reminder of encapsulation/decapsulation

24

Focus on Application Header and/or Data

We will examine how the application (header) and/or data communication with each other between the client and the server.

“Later” we will look at what roles the other layers, protocols (TCP, IP, etc.) play.

HTTP

HTTP

25

HTTP (HyperText Transfer Protocol)

HTTP – The Web’s application layer protocol. RFC 1945 and RFC 2616 Implemented in:

Client program Server program

Current version: HTTP/1.1 Encapsulated in TCP (more later)

HTTP Server

HTTP Client

HTTPHTTP

26

HTTP (HyperText Transfer Protocol)

Web page (also called a html document) Web page consists of objects

Objects (examples): HTML file JPEG image GIF image JAVA applet Audio file

<head>

<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />

<title>Rick Graziani, Cabrillo College</title><style type="text/css">

<!--

body {

margin-left: 0px;

margin-top: 0px;

margin-right: 0px;

margin-bottom: 0px;

The base HTML file references other objects in the page.

27

Web Browser - Client

Browser – The user agent for the Web. Displays requested Web page and provides navigational and

configuration features. Browser and client may be used interchangeably in this discussion. HTTP has nothing to do with how a Web page is interpreted (displayed) by

the client (browser).

HTTP Client

28

Web Server

Web Server – Stores web objects, each addressable by a URL. Implement the server side of HTTP. Examples:

Apache Microsoft Internet Information Server

HTTP Server

http://www.apache.org/

29

HTTP Request Message

Request Message Request line Header lines

ASCII Text Request line: Method field

GET, POST and HEAD The great majority of Requests are GETs

GET /~rgraziani/ HTTP/1.1

Accept-Language: en-us

User-Agent: Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 6.0; SLCC1; .NET CLR 2.0.50727; Media Center PC 5.0; .NET CLR 3.0.04506; InfoPath.1)

Host: www.cabrillo.edu

Connection: Keep-Alive

HTTP Client

HTTP Server

Some data omitted for brevity

30


Request Line

GET - Browser/client is requesting an object

/~rgraziani/ - Browser is requesting this object in this

directory (default is index.html)

HTTP/1.1 - Browser implements the HTTP/1.1 (1.1 is

backwards compatible with 1.0)

Note: HTTP GET is also used by some P2P applications like Gnutella and Bittorrent.






31


Request Line

GET: - Used by browser/client to request an object.

POST: - Used when user has filled out a form and sending

information to the server. (Forms do not have to

use POST.)

- Example: words in a search engine

HEAD: - Similar to a GET, but the server will responds with a

HTTP message but leaves out the requested object.

PUT: - Used with Web publishing tools, upload objects.

DELETE: - Used with Web publishing tools, delete objects.






32


Header Lines

Accept-Language:- User prefers this language of the object

User-Agent: - The browser type making the request

Host: - Host on which the object resides

Connection: - Client/browser is telling the server to keep

this TCP connection Open, known as a

persistent connection.

- We will talk about this later in TCP

(transport layer)






33

HTTP Response MessageHTTP/1.1 200 OKDate: Fri, 22 Feb 2008 16:34:18 GMTServer: Apache/2.0.52 (Red Hat)Last-Modified: Thu, 15 Nov 2007 19:33:12 GMTContent-Length: 15137Connection: closeContent-Type: text/html

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

<html xmlns="http://www.w3.org/1999/xhtml">

Some data omitted for brevity

HTTP Client

HTTP Server

34




Response message: Status line Header lines Entity body

35




Status Line

HTTP/1.1 – Server is using HTTP/1.1

200 OK - Status code, request succeeded and information is

returned in response

36

HTTP Response MessageHTTP/1.1 404

Status Codes

200 OK

- Status code, request succeeded and information is returned in response.

301 Moved Permanently

- Requested object has been permanently moved.

400 Bad Request

- Generic error message, request not understood by server.

404 Not Found:

-The requested document does not exist on server.

505 HTTP Version Not Supported

- The requested HTTP protocol version not supported by server.

37




Header Lines

Date: – Server is using HTTP/1.1

Server: - Status code, request succeeded and

information is returned in response

Last-Modified: – Date/time when object created or modified

Content-Length: – Number of bytes in object being sent

Connection: – Server going to close TCP connection after

sending the requested object.

Content-Type: – Object in entity body is HTML text

38




Entity Body

<!DOCTYPE html PUBLIC etc.:

– HTML text and other objects to be used by the browser/client

39

HTTP Request and Response MessagesGET /~rgraziani/ HTTP/1.1





HTTP Client

HTTP Server

HTTP/1.1 200 OKDate: Fri, 22 Feb 2008 16:34:18 GMTServer: Apache/2.0.52 (Red Hat)Last-Modified: Thu, 15 Nov 2007 19:33:12 GMTContent-Length: 15137Connection: closeContent-Type: text/html



HTTP

HTTP

40

User-Server Interaction: Cookies

Web servers are considered stateless – they do not maintain state information, keep track of the user. Higher performance – allowing the server to handle thousands of

simultaneous TCP connections (later). Web servers use cookies to track users. Cookies defined in RFC 2109

41

User-Server Interaction: Cookies

Web server installs cookies on client when: Accessed the web site for the first time (Web server does not know client

by name.)

and/or User provides information to the web server. (Web server now knows

client by name.) HTTP on Web server responds with a Set-cookie: header with an ID.

This ID is stored on the client’s computer. Each time client/browser accesses web site. The GET includes Cookie: or

User_ID or similar with the ID.

HTTP Client

HTTP Server

HTTP Requests (GET) now include ID

HTTP: Response Set-cookie: ID

HTTP Requests: GET (first time)

Web server can now track clients activities on the web site.

42

HTTP Request and Response MessagesGET /jpeg/cap81/cam0.36705623.rgb888.enc HTTP/1.1

<information omitted>

Cookie: SLSPOTNAME5=Cowells; SLSPOTNAME4=Waimea%20Bay; SLSPOTNAME3=Pipeline; SLSPOTNAME2=38th%20Ave%2E; SLSPOTNAME1=Cowells; SLSPOTID5=4189; SLSPOTID4=4755; SLSPOTID3=4750; SLSPOTID2=4191; SLSPOTID1=4189; OAX=R8bfwEbcU08ABCBu; USER_ID=5551212 <not my actual user-id>; <rest of informaton omitted for brevity>

HTTP Client

HTTP Server

HTTP/1.1 200 OKDate: Fri, 22 Feb 2008 19:00:15 GMTServer: Apache/1.3.34 (Unix)Last-Modified: Fri, 22 Feb 2008 18:51:47 GMTETag: "760a31-18ce-47bf19c3"Accept-Ranges: bytesContent-Length: 6350Keep-Alive: timeout=15, max=257Connection: Keep-AliveContent-Type: text/plain <information omitted>

HTTP: Cookie 5551212 included

HTTP data customized for Rick Graziani

43

Web Caching

Web cache or proxy server – Web cache satisfies HTTP requests on the behalf of the Origin Web server. Own disk storage Keeps copies of recently requested objects

Typically installed at ISP or larger institutions. Advantages:

Reduces the response time for client requests, especially if there are any bottlenecks in the network.

Reduces traffic on institution’s access link to the ISP (Internet).

Client

Client

Origin Server

Orgin Server

HTTP RequestHTTP

Request

HTTP RequestHTTP

Request

HTTP Response

HTTP Response

HTTP Response

HTTP Response

Web Cache or Proxy Server

44

Web Caching

1. Client/browser sends HTTP Request to Web cache (Proxy server).2. Web cache checks to see if it has a local copy of the object.

2a. Local copy: Web cache sends object to client’s browser. 2b. No Local copy: Web cache sends HTTP request to origin server.

3. Origin server sends object to Web cache.4. Web cache stores a local copy of the object.5. Web cache forwards copy of the object to the client browser.Note: TCP connections are also created between Client and Web Cache; Web

cache and Origin server (later).

Web Cache or Proxy Server

Client

Client

Origin Server

Origin Server

HTTP RequestHTTP

Request

HTTP RequestHTTP

Request

HTTP Response

HTTP Response

HTTP Response

HTTP Response

45

Web Caching (Extra)

Problem – stale cache Object may have been modified on Origin server since copy was cached

by the Web cache. Solution – Conditional GET

Request method: GET Includes header: If-Modified-Since:

Web cache sends Conditional GET to Origin server to see if there is a newer version of the object. No newer version: Send current local object Newer version: Replaces current object and forwards newer version.

46

Web Cache – Steps (Extra)

1. Client/browser sends HTTP Request to Web cache (Proxy server).

2. Web cache checks to see if it has a local copy of the object. No local copy

3. Web cache sends HTTP request to origin server.

4. Origin Web server sends HTTP response with object requested

5. Web cache stores local copy of the object with last-modified date.

6. Web cache forwards object to client/browser.

1. Web cache receives another request for this HTTP object.

2. Web cache sends a Conditional GET to the Origin Web server, with If-modified-since: header.

3. Origin Web Server returns: No change: HTTP Response message 304 Not Modified, no object.

Web cache sends local object. Change: HTTP Response 200 OK, with object.

Web cache replace object and forwards update object.

47

HTTPS

HTTPS (Hypertext Transfer Protocol over Secure Socket Layer) is a URL scheme used to indicate a secure HTTP connection.

HTTPS is not a separate protocol combination of a normal HTTP interaction over an encrypted:

Secure Sockets Layer (SSL) or Transport Layer Security (TLS) connection

48

FTP (File Transfer Protocol)

FTP was developed to allow for file transfers between a client and a server. Used to push and pull files from a server running the FTP daemon (FTPd). Uses get and put commands. RFC 959

FTP Client

FTP Server

49

Client initiates a TCP control connection with FTP server using port 21. This connection remains open until the user quits the FTP application. TCP port 21 connection includes:

Username and password is sent over TCP port 21. Remote directory changes

This state information significantly reduces total number of sessions on server.

For each file transferred, TCP opens and closes a TCP data connection on port 20.

More later on TCP ports and connections.

TCP data connection port 20

TCP control connection port 21

Username and password

Change directory on Server

Copy file from client to server – Connection Closed

TCP data connection port 20

Copy file from server to client – Connection Closed

TCP control connection port 21

Quit FTP Application – Connection Closed

FTP (File Transfer Protocol)

50

SMTP – Simple Mail Transfer Protocol

Email – One of the killer applications of the Internet.

51

Internet mail involves: User agents

Allows users to read, reply, compose, forward, save, etc., mail messages GUI user agents: Outlook, Eudora, Messenger Text user agents: mail, pine, elm

Mail servers Stores user mail boxes, communicates with local user agents and other

mail servers. SMTP

Principle application layer protocol for Internet mail Sent over TCP

Mail access protocols: POP3, IMAP, HTTP

SMTPSMTP

POP3 IMAP

User agent User agentMail server Mail server


52

SMTP RFC 2821 Transfers messages from sender’s mail server to recipient’s mail

server Push protocol, not a pull protocol

Push (from client to server or server to server) Pull (from server to client)

Retrieving email Historically, users would log into local mail server to read mail. Since early 1990’s, clients use mail access protocols:

POP3 IMAP HTTP

SMTPSMTP

POP3 IMAP



53


POP3 (Post Office Protocol) RFC 1939 Limited functionality Uses TCP port 110 Download-and-delete mode

Retrieves messages on server and store the locally Delete messages on server

Download-and-keep mode Does not delete messages on server when retrieved.

Problem Difficult to access email from multiple computers – work and home. Some email may have already been downloaded on another

computer (work) – download-and-delete To read email from another computer, must leave on server –

download-and-keep Does not provide means for user to create remote folders on mail

server

54

IMAP (Internet Message Access Protocol) RFC 2060 Mail not downloaded, but kept on server Received email is associated with user’s INBOX Users can create and manage remote folders Users can retrieve portions of the email:

Message header: Subject line and Sender

Web-based email Introduced with Hotmail in mid-1990’s Communicates with remote mailbox using HTTP HTTP is used to push (client to server) and pull the email (server to

client)

SMTP

IMAP HTTP



SMTP

55

SMTP

Mail software, processes used: MTA and MDA MUA (Mail User Agent) – Email client software. MTA (Mail Transfer Agent) – Software that governs transfer of email

between mail servers. Includes UNIX sendmail, Microsoft Exchange Server, Postfix, and Exim

MDA (Mail Delivery Agent) – Software that governs transfer of email from mail servers to clients. On Unix systems, procmail and maildrop are the most popular MDAs.

MTA receives email from the

client's MUA passes email to the MDA

for final delivery uses SMTP to route email

between servers

56

Telnet

Telnet provides a standard method of emulating text-based terminal devices over the data network.

Server

Telnet Telnet

57

Telnet

Allows a user to remotely access another device (host, router, switch). A connection using Telnet is called a Virtual Terminal (VTY) session, or

connection. Telnet uses software to create a virtual device that provides the same

features of a terminal session with access to the server command line interface (CLI).

Telnet clients: Putty Teraterm Hyperterm

Server

TelnetTelnet

58

Telnet

Telnet supports user authentication, but does not encrypt data. All data exchanged during a Telnet sessions is transported as plain text. Secure Shell (SSH) protocol offers an alternate and secure method for

server access. Stronger authentication Encrypts data

59

DHCP – Dynamic Host Configuration Protocol

IP addresses and other information can be obtained: Statically Dynamically (DHCP)

60

DHCP

DHCP Information can include: IP address Subnet mask Default gateway Domain name DNS Server

DHCP servers can be: Server on LAN Router Server at ISP

61

DHCP

We will discuss DHCP more when we discuss IPv4.

62

DNS – Domain Name System

DNS allows users (software) to use domain names instead of IP addresses

63

Name Resolution

Resolver DNS client programs used to look up DNS name information. Name Resolution The two types of queries that a DNS resolver (either a DNS client or another

DNS server) can make to a DNS server are the following:Recursive queries Queries performed by Host to Local DNS ServerIterative queries Queries performed Local DNS server to other servers

Need the IP address

64

DNS Name Resolution

User types http://www.example.com

Step 1. The DNS resolver on the DNS client sends a recursive query to its

configured Local DNS server. Requests IP address for "www.example.com". The DNS server for that client is responsible for resolving the name

Cannot refer the DNS client to another DNS server.

1

65

Step 2. Local DNS Server forwards the query to a Root DNS server.

Step 3. Root DNS server

Makes note of .com suffix Returns a list of IP addresses for TLD (Top Level Domain Servers)

responsible for .com.

DNS Name Resolution

1

2 23

66

DNS Name Resolution

Root DNS Servers There are 13 Root DNS servers (labeled A through M)

TLD Servers Responsible for domains such as .com, edu, org, .net, .uk, jp, fr Network Solutions maintains TLD servers for .com Educause maintains TLD servers for .edu There are redundant servers throughout the world.

67

Step 4. The local DNS server sends query for www.example.com to one of the

TLD servers.

Step 5. TLD Server

Makes note of example.com Returns IP address for authoritative server example.com (such as

dns.example.com server)

DNS Name Resolution4 4

5

68

Step 6. Local DNS server sends query for www.example.com directly to DNS

server for example.com

Step 7. example.com DNS server responds with its IP address for

www.example.com

6

6

DNS Name Resolution

7

69

Step 8. Local DNS server sends the IP address of www.example.com to the DNS

client.

DNS Caching When a DNS server receives a DNS reply (mapping hostname to an IP

address) it can cache the information in its local memory. DNS servers discard cached information after a period of time (usually 2

days) A local DNS server can cache TLD server addresses, bypassing the root

DNS servers in the query chain.

8

7

DNS Name Resolution

70

DNS Name Resolution

In the worst cases, you'll get a dialog box that says the domain name doesn't exist - even though you know it does.

This happens because the authoritative server is slow replying to the first, and your computer gets tired of waiting so it times-out (drops the connection) or the domain name does not exist.

But if you try again, there's a good chance it will work, because the authoritative server has had enough time to reply, and your name server has stored the information in its cache.

71

nslookup

nslookup Displays default DNS server for your host Can be used to query a domain name and get the IP address

72

DNS Name Resolution

ipconfig /displaydns After a certain amount of time, specified in the Time to Live (TTL)

associated with the DNS resource record, the resolver discards the record from the cache.

ipconfig /flushdns – Manually deletes entries The default TTL for positive responses is 86,400 seconds (1 day). The default TTL for negative responses is 300 seconds.

73

(Missing Info) DNS: 204.127.199.8

74

75

76

77

SMB – Server Message Block Protocol

The Server Message Block (SMB) is a client/server file sharing protocol. IBM developed Server Message Block (SMB) in the late 1980s to describe the structure of shared network resources, such as directories, files, printers, and serial ports.

78

SMB

Request-response protocol . Unlike FTP, clients establish a long term connection to servers. Client can access the resources on the server as if the resource is local to

the client host. SMB is sent over TCP

Prior to Windows 2000 windows used a proprietary protocol (NETBIOS) to send SMB.

Linux/UNIX have similar protocol: SAMBA

79

SMB

SMB messages can: Start, authenticate, and terminate sessions Control file and printer access Allow an application to send or receive messages to or from another

device

80

Peer-to-Peer (P2P) Networking and Applications

In addition to the client/server model for networking, there is also a peer-to-peer model.

Two or more computers are connected via a network and can share resources (such as printers and files) without having a dedicated server.

End devices (peers) can function as either a server or client.

81

P2P File Sharing

P2P (Peer-to-Peer) file sharing accounts for more traffic on the Internet than any other application (2004).

Peers (hosts) act as both clients and servers. No centralized file server. HTTP GET and responses are commonly used.

82

By Peter SvenssonThe Associated PressOct. 19, 2007“Peer-to-peer applications account for between 50 percent and 90

percent of overall Internet traffic, according to a survey this year by ipoque GmbH, a German vendor of traffic-management equipment.”

83

Challenge with P2P – locating content across thousands or millions of peers.

One solution – centralized directory Approach done by Napster

Problems (non-legal problems) Single point of failure Performance bottlenecks

1 – Inform and Update

Centralized Directory Server

P2P – Centralized Directory




Peer

Peer

Peer

Peer

2 – Query for content

3 – File Transfer

Napster

84

1. Peer A starts P2P application

2. Informs centralized directory server of its: IP address Names of objects making available for sharing (MP3, videos, etc.)

3. Directory server collects information from each peer that becomes active. Dynamic database Maps IP addresses with object names

4. Peer A queries directory server for IP addresses of other peers for specific content Directory Server returns IP addresses for those peers (Peer B)

5. Peer A establishes TCP connection and downloads file (i.e. HTTP GET) from other peer, Peer B.

6. Directory server removes Peer from database when Peer closes application or disconnects from Internet (periodic messages – pings – from server).


Centralized Directory Server

P2P – Centralized Directory




Peer

Peer B

Peer

Peer A

2 – Query for content

3 – File Transfer

85

Gnutella – public domain file sharing application Fully distributed approach

No centralized server Gnutella peer maintains peering relationship (TCP connection – later) which

a number of other peers (usually fewer than 10).

P2P – Query Flooding

Peer A

Peer CPeer B

Peer D Peer E

Peer F

Query

Query

Query

QueryQuery

Query hit

Query hit

Query hitFile transfer

Gnutella

86

Peer A searches for a file

1. Peer A sends query to all neighboring peers.

2. If neighboring peer does not have file, forwards query to all its neighboring peers

3. If any peer has the file it returns a query hit message.

4. Peer A selects a peer, Peer C, to retrieve file (HTTP GET)

5. A direct TCP connection is made with selected peer, Peer C.

6. HTTP response is used to send file.

Query Flooding Non-scalable and causes a significant amount of traffic on Internet. Gnutella modified it to limited-scope flooding which limits how many peers

away the query is sent to, usually 7 to 10. (similar to TTL – later).

Peer A

Peer CPeer B

Peer D


Peer E

Peer F

Query

Query

Query

QueryQuery

Query hit

Query hit


87

How a peer joins and departs Gnutella network

1. Finding peers: Bootstrap program: Client maintains a list of peer IP addresses who are

usually up Contact Gnutella site that maintains a list

2. Client attempts to make contact with peers (TCP connection – later)

3. Client sends Gnutella ping message to peer. Forwards Gnutella ping to other peers, who continue to forward ping

until limited-scope is reached.

4. Each peer returns a Gnutella pong message including: Its IP address Number of files it is sharing Total size of the files

Peer A

Peer CPeer B

Peer D


Peer E

Peer F

Query

Query

Query

QueryQuery

Query hit

Query hit


88

Kazaa combines ideas from Napster and Gnutella 2004 – Contributed to more traffic on Internet than any other application 2007 – Bittorrent became the leading application Proprietary technology

P2P - Combination

Kazaa

89

Kazaa does not use a centralized server Group leader peers (parent)

Higher bandwidth and Internet connectivity Greater Gnutella responsibilites

Peers (child) – non-group leaders Child peer establishes TCP connection with a group leader Group leader:

maintains database directory of child peers including their IP addresses maintain TCP connections with other group leaders

Child peers query group leaders who forward the query to other group leaders Child peer selects peer for TCP connection and file transfer

P2P - Combination

Group Leader

Group Leader

Group Leader

Query

Query

Query

Query Reply

File Transfer

Chapter 3Application Layer Functionality and Protocols

CIS 81 Networking Fundamentals

Rick Graziani

Cabrillo College

[email protected]

Last Updated: 2/23/2008

Documents

Chapter 3 Application Layer Functionality and Protocols CIS 81 Networking Fundamentals Rick Graziani Cabrillo College [email protected] Last Updated: