Transcript OSI vs TCP/IP models

Networking Fundamentals
John Bellavance CCNI
Data Networks
Developed because companies wanted to exchange info over long
distances. At first they used sneakernet, but sharing data using floppy
disks was not efficient.
•The solution was to network the resources (printers,servers) to
increase productivity and save money.
•Companies in the 80s created a variety of network software and
hardware, with their own standards. As a result they were
incompatible with each other.
Internetworking
History
One solution was to create LAN
(Local Area Network) standards
which provided guidelines for
creating hardware and software..
ENIAC – the first large scale electronic digital
computer, weighed 30 Tons.
1947 – Transistors (tiny ON/OFF switches)
were invented at Bell labs.
1976 - APPLE II – Starts the PC
Revolution
Network Protocols
Protocol suites are collections of protocols that enable network
communication between hosts. A protocol is a formal description of a
set of rules and conventions that govern how devices on a network
communicate.Protocols determine the format, timing, sequencing,
and error control of data communication. Without protocols the
computer cannot create or rebuild the stream of incoming bits from
another computer into the original data.These network rules are
created and maintained by many different organizations, such as
IEEE, ANSI, TIA.
LANs – Operate in a limited geographical area.
Allow many users to access high-bandwidth media
Provide full-time connectivity to local services.
Connect physically adjacent devices
Ex: Ethernet, Token Ring, FDDI
WANs (Wide Area Networks)
WANs interconnect LANs. They make it possible for businesses to
communicate across great distances. WANs create a new class of
workers called Telecommuters – never leave home to go to work.
•Operate over large geographically separate areas.
Allow users to engage in real-time communication with other users.
Provide full-time remote resources connected to local services.
Provide e-mail, WWW, file transfer and e-commerce services.
•Ex: Modems, ISDN, DSL, Frame Relay, T or E carrier.
MANs (Metropolitan-Area Networks)
Spans a city and connects LANs, for example a bank with several
branches might use a MAN. Rowville SC is a MAN using a wireless
link.
The MAN interconnects users is a geographic area lager than the
LAN.
MANs interconnect several LANs by bridging them with backbone
lines.
Specialized Networks Located within the LAN
•Storage-Area Networks (SANs)- is a high performance link
between server-to-storage, storage-to-storage, or server-to-server.
•SANs offer
Performance- allow access to disk or tape arrays by two or more
servers at high speed.
Availability- disaster tolerant using mirrored disks.
Scalability- uses a variety of technologies and allows easy
relocation.
DATA CENTER – is a globally coordinated network of devices
designed to accelerate the delivery of information over the
Internet infrastructure. Service providers can use these services
and avoid congestion by distributing the load. The Data Center
can deliver the download of a movie over the Internet much
quicker.
The Internet Business Exchange™ Center – Critical Hubs in
the Core Infrastructure of the Internet.
Our Internet Business Exchange™ (IBX®) centers serve as core
hubs for critical IP networks and Internet operations worldwide.
With direct access to more than 200 networks, network and Internet
operations through direct interconnection to the largest aggregation
of networks for unmatched service diversity, flexibility and
reliability. At Equinix, customers can directly access the providers
that serve over 90% of the world's Internet networks and users.
Intranets- Intranet server are different than public web servers, in
that the public does not have access to the organizations intranet.
Extranets- is and intranet that is partially accessible to authorized
outsiders with a password and username.
Virtual Private Networks (VPN) – is a private network constructed
within a public network infrastructure like the Internet.The the
telecommuter can access the company headquarters’ network through
the Internet by building a secure tunnel between and PC and a VPN
router in the headquarter.
•Access VPNs- provide access to remote user using dial-up, ISDN,
DSL.
•Intranet VPNs- allow access to employees only.
Extranet VPNs- allow access to users outside the organization.
Advantaged: Allow privacy and security, encryption for
clients(Bank) and allow employees access to the corporate network
securely.
Digital Bandwidth
 is the measure of how much information can flow from
one place to another in a given amount of time
 Bandwidth is finite- there are limitation set by the
laws of physics, DSL uses the same copper wires for
voice but use a frequency range that is wider and
therefore DSL ca send more bits per second than
modems.
 Bandwidth is not free- a service provider charges and
you need to make the right decisions about services and
equipment.
 Bandwidth is key to analyze network performance
and designing networks Demand for bandwidth is ever-increasing- new
applications are always created, voice over IP, so we
need to anticipate the need for more bandwidth.
Pipe Analogy for Bandwidth
Highway Analogy for
Bandwidth
Bandwidth is measured in bit per second.
Bandwidth varies depending on the type of medium as well as the
LAN and WAN technologies used. The physics of the medium, be it
twisted pair copper, coaxial cable or fibre optic cable influences the
limitations of the capacity to carry data.
Ex; UTP limit is 1 Gbps.
Bandwidth determined also by the equipment, the number of users,
the amount of broadcasts and so on.
Digital Bandwidth Versus Analog bandwidth
Until recently, radio and television were sent through the air using
analog electromagnetic waves which are measured in Kilohertz and
Megahertz. Digital bit streams can carry video, voice and data
unlimited amounts of data can be sent over the smallest (lowerbandwidth) digital channel. When digital info arrives at its
destination it can be reassembled, viewed and listened to in its
original form.
Typical Media
Max. Theoretical
Bandwidth
Max. Physical Distance
50-Ohm coaxial cable
(Thinnet)
10 Mbps
185m
75-Ohm coaxial cable
(thicknet)
10 Mbps
500m
CAT5 UTP
10 Base-T Ethernet
10 Mbps
100m
CAT 5
100 Base-TX Ethernet
100 Mbps
100m
CAT 5
1000 Base-TX Ethernet
1000 Mbps
100m
2000m
Multimode Fiber 100 BaseFX Ethernet
100 Mbps
Multimode Fiber 1000 BaseSX Ethernet
1000 Mbps
220m
Singlemode Fiber 1000 base
LX Ethernet
1000 Mbps (1 Gbps)
5000m
WAN Services and Bandwidths
Type of Service
Typical User
Bandwidth
Modem
Individuals
56 Kbps
DSL
Ind.,Telecommuter, small business 12Kbps-6 Mbps
ISDN
Telecommuters and small
businesses
Frame Relay
Small institutions and reliable
WANs
128 Kbps to
2 Mbps
56 Kbps to 44
Mbps
T1
Larger entities
1.544 Mbps
T3
Larger entities
44.736 Mbps
STS-1 (OC-1)
Phone companies/Backbones
51.840 Mbps
STS-3 (OC-3)
Phone companies/Backbones
155.251 Mbps
STS-48 (OC-48)
Phone companies/Backbones
2.488320 Gbps
Data Throughput
 Bandwidth is the measure of the amount of info that
can move through the network at any given time.
 Throughput refers to the actual, measured bandwidth
at a specific time of day using specific Internet routes
while downloading a specific file
 Throughput is often less than the maximum possible
bandwidth.
 a major factor in analyzing a network’s performance
Data Transfer Calculations
Divide the file size by the network bandwidth yields
an estimate of the fastest time. The result is only an
estimate, because the file size does not include any
overhead added by the encapsulation process.
File: 100 Mbytes * 8 = 800 Mbits / 100 MBitsps = 8
sec.
Learning a concept of layers helps us understand how data
flows (how data flows, how traffic flows and the rules of
flow) and how computers communicate with each other. An
example of flow is the Electrical or water networks. Each
layer has a specific tasks and uses specific protocols.
This layered communication process each layer performs a
specific task.
We will see how the network communication process is
broken into a layered model.
We will see how data is sent out over the network to reach
its destination and the steps involved. This is helpful when
troubleshooting.
For computers to send information through a network,
communication originates from a source to a destination.
Before data can be sent across a network it must first be broken into
smaller chunks.(Data Packets, data segments or data frames)
before data can be sent in the form of electrical impulses, it must
first be broken into manageable chunks. This includes a source and
destination address (like a letter) to make communication possible.
Computer Protocols
• For data packets to travel from source to
destination, all devices on the network must speak
the same language.
• Network protocol: A set of rules that make
communication on a network possible and
efficient.
• Eg: While driving a car, other cars (should!) signal
when they wish to make a turn; if they did not,
then the roads would be chaos
Networking Standards
IOS International Organization for Standards.
Proprietary systems were created that were privately developed,
owned and controlled. This did not work well with other systems.
Open means that free usage of the technology is available to the
public.
To address the problem of different networks systems being
incompatible with each other, the International Organization for
Standardisation (IOS) researched network schemes to find a set of
rules. As a result the IOS created the OSI Model Open System
Interconnection Reference Model, this allowed network vendors to
create networks that would be compatible with other networks. The
IEEE sets standards for networking technologies at Data Link
and Physical Layers.
The Purpose of the OSI Reference Model
The primary objective of he OSI model is to accelerate the
development of future networking products.
The OSI reference model allows you to view the network functions
and how information travels through a network and what occur at
each layer.
Why A Layered Model?
• Reduces complexity- Breaks network communication into
smaller parts, making it easier to develop.
• Standardizes interfaces- facilitates standardization of
network components
• Facilitates modular engineering- (development can be
made in a modular fashion- Changes in one does not effect
another Lay).
• Ensures interoperable technology- allows different
network hardware and software to communicate with each
other.
• Accelerates evolution
• Simplifies teaching & learning- by breaking it up into
smaller parts
• Easy troubleshooting
All People Seem To Need Data Processing
The OSI Reference Model
Each layer has a unique function.
7
Application
6
Presentation
5
Session
4
Transport
3
Network
2
Data Link
1
Physical
User interface
Data presentation and
encryption
Keeping different
applications’ data separate
End-to-end connections
Addresses and best path
Access to media
Binary transmission
Application Layer
Application
Presentation
Session
Transport
Network
Data-Link
Physical
– Closest to the user and does not provide
services to any other layer.
– Provides network communication services to
applications which allow for the transfer of
files(word), Bank transactions. Network
redirectors allow applications like Word and
Excel to “see” the network.
– Responsible for identifying and establishing
the availability of communication partners
and if there are sufficient resources for
communication. Makes sure that appropriate
resources are available to initiate a connection
with destination host.
– Synchronizes and establishes an agreement on
procedures for error recovery.
– Ex:Email, HTTP, telnet, FTP, SNMP.
Presentation Layer
– Provides data representation
and code formatting. Translates
Application
between multiple data formats.
Presentation – Basically, the presentation layer is
Session
responsible for representing data
so that the source and destination
Transport
can communicate at the
Network
application layer.
Data-Link
– Compression and encryption
Physical
– Ex: Ascii, Jpeg, Tiff, sound and
movie files.
Session Layer
Application
Presentation
Session
Transport
Network
Data-Link
Physical
– Provides inter-host communication by
establishing, maintaining, and terminating
sessions between applications.
– Synchronizes dialogue between 2 hosts
presentation layers and manages their data
exchange. Session uses dialog control and
dialog separation to manage the session
parameters and login.
– Some Session protocols:
•
•
•
•
•
•
NFS (Network File System) Sun/Unix
SQL (Structured Query Language) define
database info requests
RCP (Remote Call Procedure)
ASP (AppleTalk Session Protocol)
SCP (Session Control Protocol)
X-window- Unix
Controls sessions, by determining which flows are part of the
same sessions and which must be completed before it is considered
complete.
Transport Layer- boundary between application
protocols and data-flow protocols
Application
Presentation
Session
Transport
Network
Data-Link
Physical
– Connection Oriented -Controls communication between
end-to-end hosts using pre-established pathing.
– Sets up session establishment and tears down VCs.
 TCP provides the ability of multiple applications to
use a single transport (Multiplexing)
– Reliability and Error Recovery- TCP segments and
reassembles the data, adding a header with control
information for sequencing and acknowledging packets
received.
– The segment header also includes source and destination
ports for upper-layer applications
– Flow control using- windowing, buffering and source
quench messages (sent by the recipient when the buffers
(memory) are nearing capacity)
– TCP uses 3 way handshake- synchronizes sequence
number between hosts.
Transport Layer
UDP is connectionless. UDP does not acknowledge the receipt of
packets, no sequencing, no virtual circuit creation, no guarantee
delivery but less overhead. Provides error detection but not recovery.
TCP uses 3 way handshake- synchronizes sequence number
between hosts, provides reliability by establishing a communication
session before sending data (Virtual Circuit) Provides error detection
and recovery.
The transport layer attempts to provide a data transport service that
shields the upper layers from transport implementation details.
Network Layer
Application
Presentation
Session
Transport
Network
Data-Link
Physical
– Provides connectivity and path selection(path
determination and switching).
– Connectionless, and Responsible for logically
addressing the packet
– Addressing is done through routed protocols such as
IP, IPX, AppleTalk
– Path Selection is done by using routing protocols such
as (RIP, IGRP are classfull routing prot. Do not include
subnet mask in routing update), EIGRP, OSPF, and
BGP.
– Routers operate at the Network Layer,
ICMP(traceroute and ping), ARP(DHCP), RARP
(diskless workstations) operate at this layer.
– Fragments data into smaller packets to accommodate
smaller MTU size (Maximum Transmission Units)
Data-Link Layer
Application
Presentation
Session
Transport
Network
Data-Link
Physical
– Handles error notification, network topology
issues, and physically addressing the frame.
– Media Access Control -Provides the transit of
data across a physical link. Access to the media
using a physical address called a MAC address
(48 bits) two part address- 3 bytes OUI, 3 for
serial number.
– Flow control and error detection.
– through either...
•
•
Deterministic—token passing
Non-deterministic—broadcast topology (collision
domains)
– Important concept: CSMA/CD
– LLC provides SAPs (service access points) for
upper layers, flow control
DATA-LINK
Prepares data from upper layers to be transmitted over a
particular physical medium, the final encapsulation.
Convert data into bits, so it can be transmitted by physical
layer.
Error detection: CRC Cyclical Redundancy Check or FCS
Frame check sequence, if an error is detected the frame is
discarded.
Frame Relay, HDLC, PPP encapsultions.
Ethernet , Token Ring 802.5
Physical Layer
Application
Presentation
Session
Transport
Network
Data-Link
Physical
– Provides electrical, mechanical,
procedural and functional means for
activating and maintaining links between
systems. Like voltage levels, timing, physical
data rates (100 Mbps), maximum transmission
distances (100 meters).
– Standards for sending data over the physical
medium through which bits flow. Media
types:IEEE
• Connectors-Ethernet 802.3 CAT 5 cable,
RJ45 jacks.
• Coaxial cable
• Fiber Optics cable FDDI Fiber Distributed
Data Interface
• The atmosphere
The OSI Reference Model
7
Application
6
Presentation
5
Session
4
3
2
1
The top 3 layers are
known as the application layers because
they deal with the
user interface, data
formatting, and the
application access.
The OSI Reference Model
7
Application
6
Presentation
Data Transport
5
Session
Layers 1-4 are known
as the data flow layers
because they control
the physical delivery of
messages over the
network and how data
is transferred.
4
Transport
3
Network
2
Data Link
1
Physical
For data to travel from source to destination, each layer at the source
must communicate with its peer layer at the destination, this is called
Peer-To-Peer communication. During this process, the protocols at
each layer exchange information, called Protocol data Units (PDU)
between each layer.each layer depends on the services of the one below
it. If a computer wants to send data to another it must package the data
for delivery, this is encapsulation. Encapsulation wraps data with the
necessary protocol information before network transit.
Encapsulation wraps data with necessary
protocol information before transit.
How Data is Encapsulated
1. Build the data.
2. Package the data for end-to-end transport.
3. Append (add) the network address to the
header
4. Append (add) the Physical address to the
data link header.
5. Convert to bits for transmission.
De-Encapsulation
Headers are looked at at each layer and removed.
1-Read the physical address MAC, strip it off the header and the
trailer, creating a packet and passed on to upper-layers.
2- If there are errors, discard the data or ask for retransmission.
Each layer depends on the services of the other below it. To provide
this service, the lower layers uses encapsulation to put the PDU
from the upper layer into its data field; then adds whatever header
and trailer is needed.
The Importance of TCP/IP
The U.S. Department of
Defense (DoD) created
the TCP/IP reference
model. The open standard
for the Internet is TCP/IP.
Makes communication
possible between any
computers on earth. The
DOD wanted packets to
get through every time,
under any conditions.
The TCP/IP model
has only four layers.
Common TCP/IP Protocols
• Application Layer
• Transport Layer
• Internet Layer
• Network Access
Common TCP/IP Protocols
• FTP - File Transfer
Protocol
• HTTP - Hypertext
Transfer Protocol
• SMTP - Simple Mail
Transfer protocol
• DNS - Domain Name
System
• TFTP - Trivial File
Transfer Protocol
Common TCP/IP Protocols
The transport layer
involves two protocols transmission control
protocol (TCP) and user
datagram protocol
(UDP).
Similarities:
• both have layers
• both have application layers, though they include very
different services
• both have comparable transport and network layers
• packet-switched (not circuit-switched) technology is
assumed
• networking professionals need to know both
Networking Devices
•End User Devices(Hosts) – Include computers, scanners, printers.
•Network Devices – Include all devices that allow end user devices
to communicate.
Hosts devices are physically connected to the network media using a
Network Interface Card (NIC)
•Install a NIC card in the motherboard expansion slot. On
a laptop, the NIC is usually the size of a PCMCIA card
NIC Specifics
•
NICs provide hosts with access to media by using a unique code
called a MAC address. This address is used to control data
communication on the network.
•
The NIC controls the hosts access to the media.
•
MAC stands for Media Access Control
•
NICs operate at Layer 2 !!
•
It translates the parallel signal of the PC into a serial format sent
over the network cable.
•
I converts 0s and 1s into an electrical signal, pulses of light or
radio waves.
LAYER 1—Repeater
• Cleans up (regenerates) and retimes the signal
at the bit level.
• Used when a network’s cabling extends beyond
its capability. UTP is 100 meters.
NICs, Repeaters, & Hubs
Repeaters can be used
to increase the distance
NIC
NIC
So what can we use if this distance
is greater than 100 meters?
What’s the maximum
distance for Cat 5 cable?
100 meters
When the signals leave the host they are strong. The
longer the cable length, the weaker and more
deteriorated the signal becomes. Repeaters amplify and
retime signals at the bit level to allow them to travel
longer distances on the media.
LAYER 1—Hub (multi-port
repeater)
The Cloud
• Simply a multi-port repeater. Active and passive hubs.
• Used to connect multiple devices, the hub acts as the centre
of the physical star, yet same contention as a bus.
• Token Ring Hub-is called a MAU (Media Attachment Unit)
FDDI is a concentrator.
Same Broadcast
Domain and
same Collision
Domain
In this picture, all hubs forward
all traffic to all devices.
To lessen the amount of LAN traffic,
businesses began to uses bridges to filter
frames based on
MAC addresses = Look at the Local
Addresses like the post office.
Bridge
LAYER 2—Bridge
• Connects two LAN segments and pass frames.
• Build tables of all Mac addresses on the network.
• Keeps traffic local by filtering traffic based on MAC
Addresses contained in the layer 2 Frame. Creates
more usable bandwidth.the bridge keeps track of
MACs on each side of the bridge and filters traffic
based on MAC addresses only.
• SEGMENTATION = Creates separate or more
collision domains.
Switch= Micro
segmentation
A switch (also know as a multi-port
bridge), can effectively replace these
four bridges.
Switch
Another benefit of a switch is that each LAN
segment gets dedicated bandwidth.
Combines the connectivity of a hub and the
traffic regulation of a bridge They switch frames
out only the port connected to the host.
10 Mbps
10 Mbps
The Cloud
10 Mbps
10 Mbps
10 Mbps
Router
Layer 3
Routers filter traffic based on IP addresses.
The IP address tells the router which LAN
segment the ping belongs to.
1
16
The Cloud
LAYER 3—Router
• Can be used to connect different Layer 2
technologies such as Ethernet, Token Ring and
FDDI.
• Makes decisions based on network addresses (IP
Addresses).
• What are the routers two main functions?
Router’s Two Main Functions
• Path Determination
• Packet Switching
• Operate at Layers 1, 2 AND 3
Other Devices
Gateways – Is an Access Server combining routing, remote access,
voice gateway, firewall and a digital modem.
DSLAM ( Digital Subscriber Line Access Multiplexer) – Is a DSL
box giving you access to the carrier netwok.
CMTS ( Cable Modem Termination System) used by cable operators
as a concentration point or hub in the cable network to provide highspeed Internet access. This would be used in a hotel or apartment
building.
Optical Platforms – Used as backbone and WANs for fibber
backbone.
Firewalls – Is either a firewall running on a router or server or a
standalone hardware device on a network.
AAA server – is a program that handles user requests for access to
network resources. They provide authentication, authorization and
accounting (who is doing what)
VPN concentrators – offer remote access and site-to-site VPN
capabilities.
Wireless NICs - Have built in antennas.
Wireless Access Points – is a wireless transceiver that acts as a
hub.
Wireless Bridge – provides high-speed ( 11 Mbps) and long range
line of sight wireless connectivity (up to 25 Miles)
Network Topologies – are defined how computers, printers and
other network devices are connected. It defines the physical layout
of wires as well as the paths used for data transmission.
Network have both Physical and Logical topologies.
The physical refers the the layout of devices and the media (cables).
Logical defines how the media is accessed by the hosts fro sending
data.
Bus
Ring
Star
Mesh
• Single backbone
Bus Topology •
All devices are directly
connected to a central
cable.
• Each end of the bus must
be properly terminated to
absorb electrical signals, so
they don’t bounce around.
Star Topology
• All nodes connected to a
central device
• Center of star is usually a hub
or a switch
• Used for Ethernet technologies.
• Each device is connected to a
central device with its own
cable, so if one device has a
problem with a bad cable, only
that device will be affected. But
if the centre of the star fails the
whole network will fail.
Extended Star Topology
• Connects individual star
topologies together.
• At the center of the star is a
hub or a switch.
• Extends the length and size
of the network.
Ring Topology
• No backbone
• A host is directly connected to
each of its neighbors
• Used for token passing logical
topologies.
Two types of rings exist: Single
and Dual ring. In a dual ring the
2 rings allow data to be sent in
both directions, this type of
setup creates redundancy (fault
tolerance).
Hierarchical Topology
Server
• Like the extended star
except a computer controls
traffic—not a hub or a
switch.
Mesh Topology like the Internet
• Each host has its own
connection to every other
host for redundancy.
• Used in situations where
communication must not
be interrupted. Usually to
connect routers in a WAN.
Logical Topologies
• Broadcast Topology
– Each host on the LAN sends its data (or
broadcasts its data) to every other host.
– Access to media is based of “first-come, firstserve.” (Ethernet works this way)
• Token Passing Topology
– Access to media is controlled by an electronic
token.
– Possession of the token gives the host the right
to pass data to its destination.