Module 4: Application Layer Study Notes

1. Client-Server Paradigm

This is the most common paradigm, where a powerful, always-on server provides services to many
client processes that initiate requests.

 Server: A process running on a host that waits for communication requests from clients.

 Client: A process that initiates communication with a server to request a service.

 Communication using TCP:

o This is the standard for connection-oriented services.

o It provides a reliable, in-order byte stream.

o Used by protocols like HTTP, FTP, SMTP, and TELNET.

o The server must be running before the client starts.

o Server Process:

1. Creates a TCP socket.

2. Binds the socket to a well-known port.

3. Calls listen() to wait for a connection request.

4. Calls accept() to block and wait, creating a new socket for the connection
when a client connects.

o Client Process:

1. Creates a TCP socket.

2. Calls connect() to establish a connection with the server.

 Communication using UDP:

o This is used for connectionless services.

o It is an unreliable service that sends independent packets called datagrams.

o Used by protocols like DNS, TFTP, and DHCP.

o Server Process:

1. Creates a UDP socket.

2. Binds the socket to a well-known port.

3. Blocks on recvfrom() to wait for a client datagram.

o Client Process:

1. Creates a UDP socket.

2. Uses sendto() to send a datagram to the server.

2. Peer-to-Peer (P2P) Paradigm

 In a P2P paradigm, there is no always-on server.

 Pairs of hosts, called peers, communicate directly with each other.

 Peers are not always on and may have dynamic IP addresses.

 This model is scalable and cost-effective (doesn't require a large server infrastructure).

 It is commonly used for file distribution and sharing (e.g., BitTorrent).

3. Application Layer Protocols

3.1 DNS (Domain Name System)

 DNS is a distributed database that translates hostnames (like [Link]) into IP
addresses (like [Link]).

 It uses a hierarchical system of name servers:

o Root Servers: At the top of the hierarchy.

o Top-Level Domain (TLD) Servers: Manage domains like .com, .org, .edu, and country-
level domains like .in.

o Authoritative Servers: Hold the actual DNS records for specific domains (e.g., the
DNS servers for [Link]).

o Local Name Servers: Your ISP or organization's server, which receives your initial
request and performs the lookup.

 DNS uses UDP port 53 for queries.

3.2 FTP (File Transfer Protocol)

 FTP is a protocol for transferring files between a client and a server.

 It uses TCP for reliable service.

 Key Feature: FTP uses two separate TCP connections:

1. Control Connection (Port 21): Used to send commands (like LIST, RETR, STOR) and
receive responses. This connection stays open for the entire session.

2. Data Connection (Port 20): Used to actually transfer the file. A new data connection
is created for each file transfer.

 It requires user authentication (username and password).

3.3 TFTP (Trivial File Transfer Protocol)

 A simplified version of FTP.

 It uses UDP, making it unreliable and connectionless.

 It does not have authentication.

  Because it's simple and lightweight, it's often used for booting diskless workstations or
transferring configuration files to network devices.

3.4 HTTP (Hypertext Transfer Protocol)

 The protocol used to access the World Wide Web (WWW).

 It follows a client-server model where the client (browser) requests resources from a server
(web server).

 It uses TCP port 80.

 Key Features:

o Stateless: The server does not keep any state information about past requests from a
client.

o Persistent vs. Non-Persistent:

 Non-Persistent: A separate TCP connection is created for every object (e.g.,

one for the HTML file, one for each image). This is slow.

 Persistent: The same TCP connection is reused to download multiple objects,

which is much more efficient.

 HTTP messages are in plain text and include:

o Request Messages: Use methods like GET (to fetch a page) and POST (to submit form
data).

o Response Messages: Contain a status code (e.g., 200 OK, 404 Not Found) and the
requested resource.

3.5 SMTP (Simple Mail Transfer Protocol)

 The primary protocol for sending (pushing) email from a mail client to a mail server, and
between mail servers.

 It uses TCP port 25 for reliable transfer.

 SMTP is a push protocol; it only sends, it cannot be used to retrieve email from a server.

3.6 POP (Post Office Protocol)

 A protocol used by a mail client to retrieve (pull) email from a mail server.

 Uses TCP port 110.

 Key Feature: POP is very simple and typically works in an "offline" model. It connects,
downloads all emails to the local computer, and then deletes them from the server.

 It is "stateless" across sessions, meaning the server doesn't keep track of what has been
downloaded.

3.7 IMAP (Internet Message Access Protocol)

 A more modern protocol for retrieving email.

  Key Feature: IMAP is more complex and allows users to manage email directly on the
server.

 Messages can be read, organized into folders, and searched, all while remaining on the
server.

 This is ideal for accessing mail from multiple devices (e.g., phone and laptop) as the state
(read, unread, deleted) is synchronized.

3.8 MIME (Multipurpose Internet Mail Extensions)

 MIME is not a protocol but a standard for extending the format of email messages.

 SMTP can only send simple 7-bit ASCII text.

 MIME adds headers to the email to specify the type of content, allowing email to include:

o Text in character sets other than ASCII.

o Attachments (e.g., images, audio, video, documents).

o Messages with multiple parts (e.g., a plain text and HTML version).

3.9 DHCP (Dynamic Host Configuration Protocol)

 Note: This protocol was also covered in Module 3. It operates at the Application Layer.

 DHCP automatically assigns IP addresses and other network configuration parameters (like
subnet mask, default gateway, and DNS servers) to hosts as they join the network.

 It uses UDP and follows a four-step process (D-O-R-A):

1. Discover: A client broadcasts a DHCPDISCOVER message to find a DHCP server.

2. Offer: DHCP server(s) respond with a DHCPOFFER message, offering an IP address

and lease time.

3. Request: The client broadcasts a DHCPREQUEST message to accept one of the offers.

4. Acknowledge: The chosen server sends a DHCPACK message to confirm the lease.

3.10 TELNET (TErminaL NETwork)

 A protocol used for remote login to a host over the network.

 It allows a user to act as if they are at a terminal directly connected to the remote machine.

 It uses TCP for a reliable connection.

 Major Drawback: TELNET is insecure as it transmits all data, including usernames and
passwords, in clear text. It has been largely replaced by SSH (Secure Shell).