Transcript Networking II - DePaul University

Networking II
Communication Protocols
TCP/IP
Communication software
• Regardless of the hardware used to connect networked computers,
they also require communication software in order to
communicate effectively
• The software rules and language used by computers on a network
to communicate with each other is known as the network protocol
• Rules include things like:
– Transmission speed
– Pre-agreed upon codes for things like “are you ready yet—I’m about to
begin sending a data file” or “Did you receive the file I just sent”
– The communication software used by computers on a network must
agree to follow the same protocol
• Eg: If two computers don’t agree on transmission speed (one
“speaks” at 56,600 bps and the other “listens” at 28,800, the
message will not get through.
Forms of communication software
• Communication software can take a variety of forms
• Computers whose exclusive (or main) networking
needs are to speak with other computers on a LAN may
use network operating system software (NOS) such as
Novell’s Netware or Microsoft’s Windows Server or
Apple’s AppleTalk
• In the same way that an OS hides the inner workings of
many of a computer’s doings, a NOS hides the inner
workings of network communications
• A NOS must be able to respond to requests from other
computers on the LAN and must be configured to agree
on the various protocols
Quick review of some network types:
• Peer-to-Peer
• Client Server
• Host-Terminal
Host-Terminal Network
•
A very basic and somewhat dated form of communication involves a host which
has all of the information.
–
•
•
•
•
•
A series of “dumb terminals” can connect to that host
The “dumb” terminal (that is, a keyboard and a monitor without any processing
power) connects to a host
The host holds all of the important information
All the terminal can do is send questions to the host. The host sends back the
answers which are printed on the screen. There is no graphical user interface.
That is, it’s all text-based.
Was very popular in reservations systems such as when purchasing airplane or
train tickets
Less common nowadays
Peer-to-Peer network
•
•
Users can share resources and files on their computers with other users on the
network
No centralized source
– E.g. If the shared printer on your network happens to be in someone elses’s office and
that person is out for the day with their computer off, you’re out of luck.
•
•
Has the advantages of being pretty easy to set up and not requiring a dedicated
server
Disadvantages of not being centralized and not having the security features usually
found in dedicated network software
Client-server network
• The network can centralize files and resources in one or more servers
• Computers connected to the network can access the server to request
those resources (files, peripherals, etc)
NOS and LAN Type
• The choice and functionality of a NOS depends on the type of network.
• Many networks are set up according to a client-server model
– One or more computers act as dedicated servers to serve data (files, music,
spreadsheets, etc), peripherals (print server), or other network resources
– These are typically faster, higher capacity computers
– The majority of the computers act as clients who ‘request’ these resources from
the server
– The server will run NOS Server software, and the clients will run NOS Client
software
• Small networks are sometimes designed using the peer-to-peer model
(P2P) in which each computer acts as both a server and a client.
– That is, any one computer can make a request of any other computer on the
network
– Many home desktop OSs such as Windows and Mac systems have this type of
networking pre-installed
– This is how you can share files and peripherals among different computers in your
house
• Many networks are hybrids of these two models (Client/Server and P2P)
An example of P2P in Use
• This dialog box allows me to
share the ‘My Music’ folder on
my computer with any other
computer on my network.
• However, some kind of P2P
software must be present both
on this computer and on all
other on the network in order
for them to be able to share the
resource.
Example – Client for MS Networks
• Here is an example showing P2P
software installed on a computer
• Most Windows PCs come installed with
a basic NOS client called ‘Client for
Microsoft Networks’
• This client allows the PC to
communicate with other computers on
the same network (e.g. other
computers at your home) to do things
like share files, folders, printers, and
other network resources
•
The last option (not shown) is TCP/IP – we’ll discuss this shortly
Example: a PC with Novell’s Netware installed
Intranets
• These days, many LAN administrators no longer bother
purchasing and installing NOS software. Instead, they design
their network protocol based on an intranet system.
• An intranet is a system built around the protocols and
standards used by the internet.
The Mother of all Networks: The Internet
• We’ll discuss specifics of the internet in a later lecture
• For now we will focus on the network protocol used by the
internet, TCP/IP
Why TCP/IP?
• Two major advantages over other network protocols /
NOSs
• Internetworking: this protocol was developed as an
experiment in internetworking – that is, connecting
different types of networks and NOSs
– E.g. A Novell network can talk to an AppleTalk network provided
that they can both “speak TCP/IP”
• Open Standard: The TCP/IP specs are not owned by any
company. Any networked computer can freely install
TCP/IP
• As a result, when the internet was initially conceived (as
something called ARPANET), TCP/IP was chosen as the
networking protocol
Overview of TCP/IP
•
TCP/IP is one way of controlling how messages can be sent out over a network.
–
–
•
TCP: Transmission Control Protocol
IP: Internet Protocol
Some of the ‘rules’ specified by this protocol:
–
–
–
Every piece of data sent over a network must be broken into a series of small packets.
A packet is typically 1000-1500 bytes
Each packet must contain:
•
•
•
•
•
•
The data (eg a small piece of an MP3 song)
Sender’s location (IP address)
Receiver’s location (IP address)
A sequence number (so that the packet can be reassembled back into its original whole)
And a few other things
This whole process takes milliseconds!
–
–
–
–
–
Breaking a message into packets
Sequencing the packets
Assigning addresses to each packet
Routing packets to the destination
Reassembling the packets into the whole
A simplified view of a packet
Routing
•
Every packet is routed through a series of routers to its ultimate destination.
–
–
–
–
•
•
Different packets from the same message may take a different route, but all the packets will ultimately make it to the
appropriate destination.
One at the destination, the packet is reassembled into the whole message (e.g. the complete MP3 song)
The TCP part is responsible for assigning a proper sequence to each packet so that it can be reassembled
IP part is responsible for coordinating the sending and receiving addresses
The reason the internet as a whole never goes “down” is that a single packet can
take thousands of different routes to arrive at its destination
So even if a given area is down or very slow, the packet can use a different route to
reach its destination
IP – Internet Protocol
• IP is the addressing system of the internet
• Every computer connected to the internet has a unique “IP address”
– No two computers have the same address
• An IP address is made up of 4 numbers separated by periods (“dots”)
– E.g. 140.192.1.6 is the IP address of the computer hosting the web server
www.depaul.edu
• Every packet routed through the internet contains the IP address of both
the sending and receiving computer
• Conversion between the IP address (e.g. 140.192.1.6) and the familiar
names that we like to use (www.depaul.edu) is the responsibility of a
domain name system server (DNS server)
• More on this shortly
Who assigns IP addresses?
• Addresses are allocated by a regional internet registry (RIR)
• Larger organizations are given entire blocks of addresses (e.g.
140.192 was given to DePaul)
• The organization can then allocate the remaining two sets of
numbers to all of the computers in its network
• For example, DePaul owns 140.192 (and can allocate the remaining
two sets of numbers as it wishes)
– E.g. it delegated the 1.6 (140.192.1.6) at the end to the university’s main
web server
– Similarly, ABC News owns 199.181 and can delegate the remainder as it
wishes
Static v.s. Dynamic IP Addresses
• An IP address can be static or dynamic
• A static IP address is permanently assigned to a computer
– Most web servers have static IP addresses
• If a computer has a dynamic IP address, it means that the IP
address for that computer changes
– This is almost always the case when you connect to the internet using
WiFi. The ISP providing the internet connection assigns you a unique IP
address for the duration of your connection. When you disconnect, the
address is freed up for someone else to use.
– The next time you connect, you get a brand new IP address, which, again,
will last until you disconnect
– This situation sometimes applies to broadband (always-on) connections as
well
What’s my IP Address ?
• www.whatsmyipaddy.com
Why we don’t typically use/see IP addresses
• Clearly, www.depaul.edu is a lot more user-friendly than 140.192.1.6
• TRY IT! Go to your browser and type: http://140.192.1.6
• The reason we don’t have to deal with those ugly numbers is thanks to the
internet’s domain name system (DNS)
• Some examples of domain names:
– depaul.edu
– newyorktimes.com
– facebook.com
• DNS servers maintain tables that map domain names (e.g. depaul.edu) to
IP addresses (e.g. 140.192)
• E.g. Whenever you type in a user-friendly domain name (e.g. depaul.edu)
into a web browser and click ‘Go’, a DNS server somewhere kicks in an
immediately converts the domain name into an IP address
– All of the subsequent routing is done using IP addresses – not domain names
DNS
Running out of space…
•
The exponential growth of the internet means that we are running out of domain
names.
– In addition to more people connecting to the internet, many people have several
connections (e.g. desktop, laptop, cell phone, PDA, etc)
– Recall that IPv4 uses 4 groups of numbers for each IP. Each number can range from 1 to
256. This allows for approximately 4.3 billion addresses. We’re already running out!
•
For this reason, the various standards organizations involved with governance and
maintenance of the internet are developing a new version of IP from version 4
(IPv4) to version 6 (IPv6)
– IPv6 has several advantages, chief among them, many, many more possible IP addresses
(3.4×1038)
– Over the next few years, the majority of internet devices will have IP addresses
conforming to the IPv6 standard
•
A valid IPv6 address: 2001:0db8:85a3:08d3:1319:8a2e:0370:7334
•
The switch has been mandated, but is nowhere near being fully implemented. The
two IP versions will both work for the time being.