Networking Protocols and HTTP

• Chapter 2 discusses the application layer of computer networking using a top-down approach.
• The chapter covers network application principles, the Web and HTTP, electronic mail, and DNS.

Network Applications Principles

• Focuses on how programs running on end systems communicate.
• The end systems run on different operating systems.
• The end systems communicate over the network.
• With network application programming, writing software for network-core devices is not necessary.
• Rapid application development is made possible by creating programs to run on end systems.

Application Architectures

• Two possible structures for network applications include client-server and peer-to-peer(P2P).

Client-Server Architecture

• This architecture involves a server that is an always-on host with a permanent IP address
• The server uses data centers for scaling
• Clients communicate with the server can be intermittently connected
• The client's IP addresses may be dynamic
• The client does not communicate directly with each other

P2P Architecture

• An always-on server is not needed
• Arbitrary end systems can directly communicate, requesting service from and providing it to other peers.
• There is self-scalability due to new peers bringing new service capacity, as well as new service demands
• Peers are intermittently connected and change IP addresses are assigned dynamically
• Complex management may be needed

Processes Communicating

• A process is a program that runs on the host device
• Processes communicate with each other using inter-process communication if they are running on the same host device
• Messages are exchanged if the processes run on different host devices
• A client process initiates communication
• A server process waits to be contacted
• Applications with P2P architectures have client and server processes

Sockets

• A socket is used by a process to send and receive messages to and from
• A socket is analogous to a door
• The sending process shoves messages out the door
• The sending process relies on transport infrastructure on the other side of the door to deliver messages to the socket at the receiving process

Addressing Processes

• An identifier is needed for a process to receive messages
• The identifier includes the IP address and port numbers
• The host device has a unique 32-bit IP address
• Common example port numbers include 80 for the HTTP server and 25 for the mail server
• For sending an HTTP message to gaia.cs.umass.edu web server, the IP address is 128.119.245.12, and the port number is 80.

Application Layer Protocol Definitions

• Types of messages exchanged, such as request and response
• Message syntax, including what fields in messages are and how fields are delineated
• Semantics, which focuses on the meaning of information in the fields
• Rules for when and how processes send and respond to messages
• Open protocols are defined in RFCs
• Examples of open protocols incude HTTP and SMTP
• Proprietary protocols exist, for example, Skype

Transport Service Needs

• Data integrity; some apps, like file transfers, require 100% data integrity, while others, like audio can tolerate some loss
• Throughput; some, like multimedia, require a minimum amount to be effective, while others use whatever they get
• Timing; some, like online gaming, require low delay to be effective
• Security; encryption and data integrity are needed

Internet Transport Protocols

• TCP service offers reliable transport, flow control, and congestion control between sending and receiving processes
• TCP does not guarantee timing, throughput, or security, and is connection-oriented, requiring setup between processes
• UDP service provides unreliable data transfer between sending and receiving processes.
• UDP does not provide reliability, flow or congestion control, timing, throughput, or security, and is connectionless
• Common applications include:
• E-mail uses application layer protocol SMTP and underlying protocol TCP
• Web, remote terminal access, and file transfer use HTTP, Telnet and FTP application layer protocols and TCP protocol respectively
• Streaming multimedia and Internet telephony uses HTTP, RTP and SIP application layer protocols, with TCP or UDP protocols

Web and HTTP

• A web page consists of objects, which consist of a base HTML file with several referenced objects
• Objects can be HTML, JPEG images, Java applets, or audio files
• A base HTML file includes several referenced objects
• Each object is addressable by a URL, like www.someschool.edu/someDept/pic.gif

HTTP Overview

• HTTP is a web's application layer protocol and follows a client-server model.
• The client is the browser that uses HTTP to request, receives, and "displays" web objects
• The server sends (using HTTP protocol) objects in response to requests
• HTTP uses TCP
• The client initiates a TCP connection (creates socket) to server, port 80
• The server accepts the TCP connection from the client
• HTTP messages are exchanged between the client and the web server
• The TCP connection is usually closed
• HTTP is stateless and does not store any information about past client requests

HTTP Connections

• Non-Persistent HTTP connections are at most one object sent over TCP connection
• Non-Persistent HTTP connections are closed by the server upon completion of a request
• Downloading multiple objects requires multiple connections
• Persistent HTTP connections allow multiple objects to be sen over TCP connection
• In Persistent HTTP connections, a single TCP connection can be used by the client and server

Non-persistent HTTP Example

• Suppose a user enters the URL: www.someSchool.edu/someDepartment/home.index in their browser
• The URL contains text and references to 10 jpeg images
• The client initiates a TCP connection to the HTTP server at www.someSchool.edu on port 80
• The server waits for the connection at port 80 and accepts, notifying client of new connection
• The client sends an HTTP request message with containing URL into TCP connection socket
• The HTTP server receives the request, forms and sends a response with the requested object into the socket
• steps repeat for each of the 10 images
• The HTTP server closes the TCP connection

Non-persistent HTTP Response Time

• RTT (Round Trip Time) is time for a small packet to travel from client to server and back
• The equation for HTTP response time is: one RTT to initiate TCP connection + one RTT for HTTP request and first few bytes of HTTP response to return + file transmission time
• The equation for non-persistent HTTP response time is: 2RTT+ file transmission time

Persistent HTTP

• Requires 2 RTTs per object when using non persistent HTTP
• There is OS overhead for *each TCP connection with non persistent HTTP
• Browsers will often open many parallel TCP connections to fetch all the referenced objects
• Server leaves the connection open after response
• Subsequent HTTP messages between client and server are sent over the open connection
• Clients send new requests as soon as referenced objects are encountered

HTTP Request Message

• There are two types of HTTP messages: request and response
• The request message contains ASCII (human-readable format) and consists of request line (GET, POST, HEAD commands) and header lines
• The messages use carriage return and line-feed characters.
• HTTP's general format starts with the method, the URL, and the version, followed by the header field name, then the entity body

Uploading Form Input

• POST is often the method used, where the web page includes form input, with the input uploaded to the server in the entity body
• Using the URL method, the GET method is called, with the input uploaded in the URL field of request line
• For example; www.somesite.com/animalsearch?monkeys&banana

HTTP Method Types

• HTTP/1.0
• GET, POST, and HEAD
• HEAD asks the server to leave the requested object out of the response
• HTTP/1.1
• GET POST and HEAD
• PUT uploads the files into the entity body specified in the URL field
• DELETE deletes the file specified in the URL field

HTTP Response Message

• The response messsage has a human-readable format that begins with the HTTP version and status code plus a short description, then header lines, blank line followed by the server response object
• Examples of HTTP response status codes:
• 200 OK means that the request was successful and the object is being sent
• 301 Moved Permanently means that the object has moved, and the new location is specified later in this message
• 400 Bad Request means that the request was not understood by the server
• 404 Not Found means that the document was not found on this server
• 505 HTTP Version Not Supported means that the server doesn't support the HTTP version from client

Cookies

• Cookies enables sites track users
• Many Web sites use cookies requiring four components:
• a cookie header line of the HTTP response message,
• a cookie header line in the next HTTP request message,
• a cookie file kept on the user's host, and
• a back-end database at the web site

What Cookies are Used for

• Cookies are used for authorization, shopping carts, recommendations, and user session state
• Keep state involves protocol endpoints by maintaining state at the sender and receiver over time
• Cookies are sent using the http message carrying state
• Cookies can impact user privacy by allowing sites to learn more about the user and track their internet usage

Web Caches (Proxy Server)

• Satisfy client request without involving origin server
• A user sets the browser to access the Web via a cache.
• A browser sends all HTTP requests to cache
• The object is returned if it is in cache
• A requests is started from the origin server if the object does not exist.

Web Caching Benefits

• Reduces the time it takes to respond to client requests
• Reduces the network traffic on an institution's access link
• Makes the Internet dense with caches enables "poor" content providers to effectively deliver material

Caching Example

• An institution's access link utilization is 99% and total delay is 2 secs the the assumptions made from the original access link
• By increasing the link speed, access utilization drops to 9.9%
• By installing a local cache and supposing the hit rate is 0.4, the utilization is 0.58, and delay to 1.2 seconds

Conditional Get

• When an object is cached, it is useful not to send the object again if the cached version to up-to-date because this reduces the network traffic and avoids transmission delay.
• Server: response contains no object if cached copy is still up-to-date, uses HTTP/1.0 304 Not Modified
• Cache: specify date of cached copy in HTTP request
• If-modified-since:

Electronic Mail Components

• Three major components of electronic mail are user agents, mail servers, and SMTP
• User agent is often a "mail reader" that is used to compose, edit, and read mail messages
• This includes Outlook, Thunderbird, and Apple Mail
• It is used for outgoing and incoming messages stored on the local server or local machine

Electronic Mail: Mail Servers

• A mailbox contains all the incoming messages for a user
• The outgoing messages are stored in the message queue to be sent
• SMTP protocol is used between other servers to connect via a client, to send the data; or "service," to receive the data

Electronic Mail: SMTP

• SMTP or Simple Mail Transfer Protocol uses TCP to transfer messages from client to server on port 25
• Transfer requires sending server to reiceving server
• The process contains three phases:
• handshake or greeting phase,
• transfer of messages phase
• closure

Mail Message Format

• SMTP is a protocol for exchanging files
• RFC 822 outlines the standard format
• Information is transported as header lines, Body: the "message". This is an ASCII set of Charachter set only

SMTP Interaction

• SMTP uses persistent connections
• SMTP requires messages (header & body) to be in 7-bit ASCII
• SMTP servers use \r\n . \r\n to determine the end of messages

Mail Interactions

• Alice uses her user agent to compose the message
• Alice uses the mail server, which sends the message to her mail server
• A client side SMTP opens the TCP connection in Bob's Mail Server
• SMTP connects using the TCP Protocol
• Bob places the message in box to be read
• The invoked is sent

Mail Accessing Protocols

• Accessing protocols include: SMTP - sends/place to receiver’s server
• mail access protocol: protocol on retrieval from server
• Internet Mail Access Protocol: more features
• HTTP: (gmail), Hotmail mail

DNS

• System names are distributed in the hosts. Name servers are called to to to the domain and vice versa
• DNS is called a Domain Name System and the process is: distributed database is used by layer application protocol. The Host name, name of server (name) and address translation are connected.

DNS Serivices

• Translation with a internet IP host code
• host addressing
• Alias Name is mail server
• Loads the distribution
• Repeats the Web server. Correspond IP Names

DNS Hierarchies

• Clients want to join, and use the IP on www.amazon.com
• Uses a root DNS Server and uses client queries the Root name server for the the first time
• Client joins the connects DNS Server connects and joins www.mazon.com to join

DNS Root Name Server

• Server connected to name that resolves code
• Contacts the autoritate name server
• Name of domain and code given
• Returns name to computer, a local name is given

TLD, Autoritate Servers

• Domain is called for domains and job
• Each company has server to give host name
• Maintain host
• Run each organization

DNS and Local Server Name

• Each server has a ISP Server, such as a company, university, or residential location, all have local servers
• When host makes DNS queries, it sends a query to its DNS Server
• Act like a proxy and forwards the names out

Questions about HTTP status codes (301, 505), Telnet commands for initiating TCP connections, URL components, email transfer protocols, and HTTP's stateless nature. Includes non-persistent HTTP.