COMP519: Web Programming

Autumn 2013
Acknowledgment:
The slides are derivatives of ones by Dr. David Reed.

Internet & World Wide Web Protocols
network layers
TCP/IP
domain name system, IP addresses, routing protocols
HTTP
GET/POST, headers, caching, cookies
OSI 7-Layer Model
in the 70's, computer networks were ad hoc, vendor-specific

Open Systems Interconnection model
developed by the ISO (International Organization
for Standardization) in 1984
provides an abstract model of networking

divides the tasks involved in moving information
between networked computers into 7 task groups
each task group is assigned a layer

Each layer is reasonably self-contained, so
can be implemented independently
changes/updates to a layer need not effect other
layers
Protocol Layers
Application layer
describes how applications will communicate
e.g., HTTP, FTP, Telnet, SMTP
Presentation layer
describes the form of data being transferred & ensures that it will be
readable by receiver
e.g., floating point formats, data compression, encryption
Session layer
describes the organization of large data sequences & manages
communication session
e.g., coordinates requests/responses (traffic flow)

Transport layer
describes the quality and nature of data delivery
e.g., how retransmissions are used to ensure delivery

Network layer
describes how a series of exchanges over various data links can deliver
data across a network
e.g., addressing and routing

Data Link layer
describes the logical organization of data bits transmitted on a particular
medium
e.g., frame sequencing, error notification
Physical layer
describes the physical & electrical properties of the communications media
e.g., voltage levels, data rates, max distances
Protocols define how messages are sent and received, i.e. the
format and order of them, what happens when messages are
transmitted and received.

Efficient protocols are what allows the Internet to operate in the
decentralized manner that they do.

Packet routing using these protocols also allows for reasonably
good error checking/correcting methods, and reliability of
transmissions.

What is a protocol after all?
Layer Protocols
across the network, processes at the same level can (seemingly) communicate
e.g., Web server & browser run at the application level, communicate via HTTP


in reality, actual communication takes place at the physical layer
upper layers can only communicate with those above and below

at the source, as data is passed down the layers:
the protocol for each layer adds control information to the data
at the destination, as data is passed up the layers:
the protocol for each layer strips and analyzes the control information for that layer
Internet Protocol Suite
The Internet Protocol Suite is a set of communications protocols used for information
transfer on the Internet (and other networks).

Network layer: Internet Protocol (IP)
provides generalized packet network interface
handles routing through the Internet
connectionless and unreliable (in contrast, the telephone system is a connection service)

Transport layer: Transmission Control Protocol (TCP)
provides a virtual circuit over which two processes can communicate
supplies logic to give reliable, connection-oriented session
FTP (file transfer) and HTTP are built on top of TCP
Internet Protocol (IP)
The Internet Protocol divides information into packets for delivery.

Internet Protocol adds packet routing info (20 bytes).
Time-to-live (TTL):
indicates amount of time packet is allowed
to remain before being discarded

Source address:
IP address of host sending the packet

Destination address:
IP address of host to receive the packet

Options:
options such as sender-specified routing
or security

Transmission Control Protocol (TCP)
TCP adds information that provides a virtual circuit, including message formatting,
circuit management, flow control, error correction
Source & destination ports
a port is analogous to a mailbox

Sequence number:
identifies its place in sequence
(byte # in overall message)

Acknowledgement number:
specifies the next byte # in sequence,
if destination does not receive it in X
amount of time, will notify sender

Control flags:
used to set up connection (3-way
handshake: request, ack, ack),
mark as urgent, terminate connection,
Data encapsulation and transmission
source
application
transport
network
link
physical
H
t
H
n
M
segment H
t
datagram
destination
application
transport
network
link
physical
H
t
H
n
H
l
M
H
t
H
n
M
H
t
M
M
network
link
physical
link
physical
H
t
H
n
H
l
M
H
t
H
n
M
H
t
H
n
M
H
t
H
n
H
l
M
router
switch
message M
H
t
M
H
n
frame
IP addresses
IP addresses are numerical labels assigned to
computers in a network, used for identification
and addressing (passing information)
IP addresses (under IP version 4) are 32 bits long
10010011 10000110 00000010 00010100
written as a dotted sequence
147.134.2.20

divided into 5 classes
class A: start with 0, then 7-bit code
2
24
= 16,777,216 hosts in subnetwork
class B: start with 10, then 14-bit code
2
16
= 65,536 hosts in subnetwork
class C: start with 110, then 21-bit code
2
8
= 256 hosts in subnetwork
class D: start with 1110
used for multicasting
class E: start with 11110
reserved for future use

IPv6 extends address size to 128 bits
extensions support authentication, data
integrity, confidentiality
Domain Name System
rarely do applications deal directly with IP addresses
a hierarchical system of domain names can be used instead
top level domains: edu, com, gov, org, net,

commonly: hostname.subdomain.domain (possibly many subdomains)
e.g., csc.liv.ac.uk



domain names serve as listings in a phone book for the internet, i.e. when you
enter a domain name in a web browser, a look up is done to find the
corresponding IP address

domain names also simplify things in that if the physical network changes (e.g. IP
addresses are modified), the domain names can remain the same while the
phone book entries are updated with the new IP address




Domain Name Servers (cont.)
a domain name server (DNS) is a machine that keeps a table of names
and corresponding IP addresses (i.e. this performs the look up to
match domain names to their corresponding IP address)

there are 13 root servers in the world
while there are only 13 root servers, there are multiple physical servers (that mirror
them) that operate in different geographic locations to provide reliable service in
case of hardware failures

when an application specifies a host name,
go to local domain name server and try lookup in its stored cache
if not stored there, then local DNS requests address from a root server
root server determines appropriate name server & forwards request
Routing Protocols
routers (or gateways) are special purpose machines on the Internet that
determine the path for packets from source to destination
when a router receives a packet, inspects the destination address
looks up that address in a routing table
based on the contents of the table, forwards the packet to another router (or to its final
destination if possible)
Routing Information Protocol (RIP)
describes how routers exchange routing table information
uses hop-count as the metric of a path's cost

Open Shortest Path First Protocol (OSPF)
more robust, scalable protocol than RIP
doesn't exchange entire tables, only updates changed links

Internet Control Message Protocol (ICMP)
adjunct to IP, notifies sender (or other router) of abnormal events
e.g., unreachable host, net congestion
TCP/IP
IP provides the method of distributing data in packets, can distribute packets to a
destination via different routes, and can handle congestion in this manner.

TCP provides for virtual circuits, error detection and correction, automatic repeat
requests, and other reliable end-to-end communications, using the IP
protocol for transmission (TCP is designed for accurate, not necessarily
timely, communications).

TCP handles reassembling packets at the destination, as they can arrive out of
order due to the IP routing scheme to avoid congestion.

TCP/IP together provide for the most commonly used functions on the internet,
namely e-mail, file transfer, and the basis of the HTTP protocol for web page
documents.
World Wide Web
the Web is the worlds largest client/server system

communication occurs via message passing
within browser, select URL of desired page
browser requests page from server
server responds with message containing
type of page (HTML, gif, pdf, zip, )
page contents
browser uses type info to correctly display page
if page contains other items (images, applets, ), browser must request each separately
Hypertext Transfer Protocol (HTTP):
application-level protocol for distributed, collaborative, hypermedia information systems

generic, stateless, object-oriented
can be used for many tasks, such as name servers & distributed object management systems
underlying language of the Web
HTTP
HTTP/1.0 allows only connectionless message passing
each request/response requires a new connection
to download a page with images requires multiple connections
can overload the server, require lots of overhead

HTTP/1.1 provides persistent connection by default
once client & server connect, remains open until told to close it (or timeout)
reduces number of connections, saves overhead
client can send multiple requests without waiting for responses
e.g., can request all images in a page at once
GET request
most URLs (Uniform Resource Locator) have the form:

protocol://serverName URI

e.g., http://www.csc.liv.ac.uk/~martin/index.html

to retrieve a document via HTTP from the server, we can issue a GET request
using the HTTP protocol

GET URI HTTP/1.1
Host: serverName
Web server only knows the contents of the GET request message

this request is automatically generated by browser when you select a URL
could also come from a link checker, a search engine robot,

can come directly from a telnet connection using port 80
GET example
bash-3.1$ telnet www.csc.liv.ac.uk 80
Trying 10.128.0.3...
Connected to www.csc.liv.ac.uk
(10.128.0.3).
Escape character is '^]'.
GET /~martin/index.html HTTP/1.1
Host: www.csc.liv.ac.uk

HTTP/1.1 200 OK
Date: Mon, 08 Oct 2011 10:01:15 GMT
Server: Apache/2.0.58 HP-UX_Apache-based_Web_Server
(Unix) mod_perl/1.99_16 Perl/v5.8.7 DAV/2 PHP/5.0.4
Last-Modified: Mon, 01 Oct 2011 14:55:16 GMT
ETag: "ec3f-1122-9fd83d00"
Accept-Ranges: bytes
Content-Length: 4386
Content-Type: text/html

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0
Strict//EN
"http://www.w3.org/TR/xhtml11/DTD/xhtml1-
strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"
xml:lang="en" lang="en">

<head>
. . .

. . .
</head>

<body>

. . .

. . .
</body>
</html>

Connection closed by foreign host.

server response has assorted
header information, followed
by the page
Response Header Fields

the first line of the servers response contains a status code

200 OK request was processed successfully

301 Moved permanently document has been moved
304 Not modified if cached version is up-to-date

400 Bad request syntax error in clients request
403 Forbidden client is not allowed access (e.g., protected)
404 Not found file could not be found

500 Internal server error server failed
503 Service unavailable server is overloaded


Other response header fields

in addition to the status code, the servers response may include

Date response time (in GMT)

Server identification info on the server

Last-modified time document was last changed (in GMT)

Content-length size of document, in bytes

Content-type file format (e.g., html, gif, pdf)

Expires prevents browser from caching beyond date

File not found
bash-3.1$ telnet www.csc.liv.ac.uk 80
Trying 10.128.0.3...
Connected to www.csc.liv.ac.uk
(10.128.0.3).
Escape character is '^]'.
GET /~martin/foo.html HTTP/1.1
Host: www.csc.liv.ac.uk
HTTP/1.1 404 Not Found
Date: Mon, 08 Oct 2011 10:10:35 GMT
Server: Apache/2.0.58 HP-UX_Apache-based_Web_Server
(Unix) mod_perl/1.99_16 Perl/v5.8.7 DAV/2
PHP/5.0.4
Accept-Ranges: bytes
Transfer-Encoding: chunked
Content-Type: text/html

<html>
<head>
<title>404 Not Found</title>
<link rel="stylesheet" href="./fonts.css"
type="text/css">
</head>

<body>
. . .
<td id="main_content">
<div id="topmargin"><a name="skipnav"
tabindex="1"></a></div>
<!-- Beginning of main page content -->

<h1>Page Not Found</h1>
<p class="text">The requested URL
was not found on this server.</P>

. . .

</html>

Connection closed by foreign host.
if file not found, response
includes 404 status code
and generic error page
Directories as URIs
Trying 10.128.0.3...
Connected to www.csc.liv.ac.uk
(10.128.0.3).
Escape character is '^]'.
GET /~martin/ HTTP/1.1
Host: www.csc.liv.ac.uk
if a directory is specified, will
look for a file named
index.html
HTTP/1.1 200 OK
Date: Mon, 08 Oct 2011 10:18:00 GMT
Server: Apache/2.0.58 HP-UX_Apache-based_Web_Server
(Unix) mod_perl/1.99_16 Perl/v5.8.7 DAV/2 PHP/5.0.4
Last-Modified: Mon, 01 Oct 2011 14:55:16 GMT
ETag: "ec3f-1122-9fd83d00"
Accept-Ranges: bytes
Content-Length: 4386
Content-Type: text/html

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0
Strict//EN"
"http://www.w3.org/TR/xhtml11/DTD/xhtml1-
strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"
xml:lang="en" lang="en">

<head>
<title> Russell Martin's Home Page </title>
. . .
</head>

<body>

. . .

</body>

</html>
Connection closed by foreign host.
Redirection
bash-2.05b$ telnet www.csc.liv.ac.uk 80
Trying 10.0.64.233...
Connected to www.csc.liv.ac.uk.
Escape character is '^]'.
GET /~avf/ HTTP/1.1
Host: www.csc.liv.ac.uk
HTTP/1.1 301 Moved Permanently
Date: Tue, 06 Sep 2005 13:49:15 GMT
Server: HP Apache-based Web Server/1.3.27 (Unix)
mod_perl/1.27 PHP/4.2.2
Location: http://www.csc.liv.ac.uk/~avf/
Transfer-Encoding: chunked
Content-Type: text/html; charset=iso-8859-1


<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<html><head>
<title>301 Moved Permanently</title>
</head>
<body>
<h1>Moved Permanently</h1>
The document has moved <a
href="http://www.csc.liv.ac.uk/~avf/">here</a>.
<hr/>
<address>HP Apache-based Web Server/1.3.27 Server at
www.csc.liv.ac.uk Port 80</address>
</body>
</html>

0

Connection closed by foreign host.
Request Header Fields
the client can specify additional information in the request

User-Agent specifies the browser version

Referer tells server where the user came from
useful for logging and customer tracking

From contains email address of user
generally not used for privacy reasons

Authorization can send username & password
used with documents that require authorization

If-Modified-Since only send document if newer than specified date
used for caching
Conditional GET
bash-2.05b$ telnet www.csc.liv.ac.uk 80
Trying 10.0.64.233...
Connected to www.csc.liv.ac.uk.
Escape character is '^]'.
GET /~avf/ HTTP/1.1
Host: www.csc.liv.ac.uk
If-Modified-Since: Tue, 30 Aug 2005 14:00:00 GMT
HTTP/1.1 304 Not Modified
Date: Tue, 06 Sep 2005 14:08:58 GMT
Server: HP Apache-based Web Server/1.3.27
(Unix) mod_perl/1.27 PHP/4.2.2
ETag: "10b3e-1000-431452ef"

Connection closed by foreign host.
since the document has not been modified
since the specified date, the page is not
sent by the server (status code 304)
Other Request Methods

HEAD similar to GET, but requests header information only
useful for checking to see if a document exists, how recent

POST similar to GET, but encodes inputs differently
useful for submitting form contents to a CGI program, and is
also often used in PHP scripts for submitting information

PUT upload a document to the server
new in HTTP/1.1

DELETE delete a document from the server
new in HTTP/1.1
HEAD example
bash-2.05b$ telnet www.csc.liv.ac.uk 80
Trying 10.0.64.233...
Connected to www.csc.liv.ac.uk.
Escape character is '^]'.
HEAD /~avf/index.html HTTP/1.1
Host: www.csc.lliv.ac.uk
HTTP/1.1 200 OK
Date: Tue, 06 Sep 2005 14:23:24 GMT
Server: HP Apache-based Web Server/1.3.27 (Unix)
mod_perl/1.27 PHP/4.2.2
Last-Modified: Tue, 30 Aug 2005 12:37:03 GMT
ETag: "10b3e-1000-431452ef"
Accept-Ranges: bytes
Content-Length: 4096
Content-Type: text/html

Connection closed by foreign host.
server does not send the page, only the
header information
when we discuss CGI programming (with Perl), we will see more about how
the GET and POST methods are used to pass information to a server

information is extracted in different ways depending upon the method used
to send the data

PHP provides special built-in associative array variables ($_GET[ ] and
$_POST[ ]) that provide easy access to information that a server receives via
one of these requests
GET and POST methods (more to come)
Caching
browsers cache pages to save downloading

maintain temporary storage (cache) for recent pages

when a page is requested, check to see if already in cache

if not in the cache, issue GET request
when response message arrives,
display page and store in cache (along with header info)

if already stored in the cache, send GET request with If-Modified-Since header set to the data
of the cached page
when response message arrives,
if status code 200, then display and store in cache
if status code 304, then display cached version instead
Cookies
Netscapes solution: cookies
a cookie is a collection of information about the user, and a browser can store some
information on your machine that it can later request

server can download a cookie to the clients machine using the Set-cookie header in
a response

Set-cookie: CUSTOMER=Alex_Thompson; PATH=/; EXPIRES=Thursday, 29-Jan-2010
12:00:00

when user returns to URL on the specified path, the browser returns the cookie data
as part of its request

Cookie: CUSTOMER=Alex_Thompson
HTTP message passing is transaction-based, stateless
many e-commerce apps require persistent memory of customer interactions

e.g., amazon.com
remembers your name, credit card, past purchases, interests
Other sites use shopping carts for purchases, logins for access, etc.