Web Basics and Protocols CS 502 – 20020129 Carl Lagoze

Cornell CS502

Web Basics and Protocols

CS 502 – 20020129Carl Lagoze

Acks to McCrackenSyracuse Univ.

Cornell CS502

In the beginning….

Cornell CS502

In the beginning…

Cornell CS502

ARPANET

• DoD funded through leadership of Licklider

• Inspired by move from batch to timesharing

• Allowed remote login

Cornell CS502

Packet Switching

• Invented in early 1960’s by Baran, Davies, Kleinrock

• digital, redundant, efficient, upgradeable (software)

• 1969 ARPANET first network implementation

Cornell CS502

Packet Switching

• Network messages broken up into packets• Each pocket has a destination address• Pass and forward model – router gets packet,

examine, decides where to send next• Message reassembled on other end

Cornell CS502

Layered Protocol Model

Cornell CS502

TCP/IP Protocol Suite

• IP – packet delivery• TCP – virtual circuits, packet

reassembly• ARP/RARP – address

resolution

Cornell CS502

Protocol Layers

Cornell CS502

Internet Issues (Internet 2)

• Demands of multimedia applications• Virtual circuit reservations – bandwidth and

quality of service guarantees• Real time streaming protocols• State saving

Cornell CS502

Internet Governance

• Internet Society (ISOC) – Evolution, social & political issues

• Internet Architecture Board (IAB) – Oversees standards process

• Internet Engineering Task Force (IETF) – standards development

• Internet Assigned Names Authority (IANA) – protocol # assignment

• Internet Corporation for Assigned Names and Numbers (ICANN) – IP and DNS addresses

• World Wide Web Consortium (W3C) – web standards and evolution

Cornell CS502

Internet Documents

• STD’s – Official IETF Internet standards– http://www.rfc-editor.org/rfcxx00.html

• RFC’s – “Requests for Comments” to IETF community for information, standardization– http://www.ietf.org/rfc.html

• Internet Drafts – IETF working documents– http://www.ietf.org/ID.html

• W3C Reports (recommendations, drafts, notes)– http://www.w3.org/TR/

Cornell CS502

Well-Known Protocols

• Telnet – external terminal interface, RFC 854 (1983)

• FTP – file transfer, RFC 959 (1985)• SMTP – mail transport, RFC 821 (1982)• HTTP – distributed, collaborative hypermedia

systems, RFC 1945 (1.0 1996), RFC 2616 (1.1 1999)

Cornell CS502

Daemons and Ports

23

80

21

telnetd

httpd

ftpd

Socket(Virtual Circuit)

Cornell CS502

Basic Socket ServerProgramming

Cornell CS502

HTTP

• HTTP is…– Designed for document transfer– Generic

• not tied to web browsers exclusively• can serve any data type

– Stateless• no persistent client/server connection

Cornell CS502

HTTP Session

• An HTTP session consists of a client request followed by a server response

• Requests and responses:– are sent in plain text– conform to the HTTP syntax– consist of start line, headers, blank line, and

message body

Cornell CS502

HTTP Request

• Start line– Consists of method, URL, versionGET index.html HTTP/1.1– Valid methods include:

• GET, POST, HEAD, PUT, DELETE• Headers

– HTTP/1.1 requires a Host: header• Body content

Cornell CS502

HTTP Request Methods

• Methods include– GET: retrieve information identified by the URL– HEAD: same as get but don't get message body

(content)– POST: accept the request content and send it to the

URL– PUT: store the request content as the given URL

Cornell CS502

HTTP Response

• Start line– consists of HTTP version, status code, and

descriptionHTTP/1.1 200 OKHTTP/1.1 404 Not Found

• HeadersContent-type: text/html

• Content

Cornell CS502

HTTP Response Codes

• Respose coded by first digit– 1xx: informational, request received– 2xx: success, request accepted– 3xx: redirection– 4xx: client error– 5xx: server error

Cornell CS502

HTTP Content Body

• Header fields can affect content interpretation– required header field: Content-type– others: Content-Encoding, Content-Length, Expires,

Last-Modified– added by web server - we will configure some of

these later

Cornell CS502

Serving a Page

• User of client machine types in a URL

client( N etscape)

server( Apache)

http: / / www.sm allco.com / index.h tm l

Cornell CS502

Serving a Page

• Server name is translated to an IP address via DNS

client( N etscape)

server( Apache)

http: / / www.sm allco.com / index.htm l

192.22.107.5

Cornell CS502

Serving a Page

• Client connects to server using IP address and port number

client( N etscape)

server( Apache)

http: / / www.sm allco.com / index.htm l

192.22.107.5

192.22.107.5port 80

Cornell CS502

Serving a Page

• Client determines path and file to request

client( N etscape)

server( Apache)

http: / / www.sm allco.com / index.h tm l

Cornell CS502

Serving a Page

• Client sends HTTP request to server

client( N etscape)

server( Apache)

http: / / www.sm allco.com / index.h tm l

GET index .html HTT P/ 1.1

Cornell CS502

Serving a Page

• Server determines which file to send

client( N etscape)

server( Apache)

http : / / ww w.sm allco.com / index.htm l"index.htm l" is really/ etc/ httpd / htdocs/ index.htm l

Cornell CS502

Serving a Page

• Server sends response code and the document

client( N etscape)

server( Apache)

http : / / w ww.sm allco.com / index.htm l

HTTP/ 1.1 200 OKContent- type: text/ htm l

[ contents of index.h tm l]

Cornell CS502

Serving a Page

• Connection is broken

client( N etscape)

server( Apache)