Transcript Chapter 1

Networking Review
CIS 278 Spring 2004
Learning in CIS
• "Education is not the filling of
a pail, but the lighting of a
fire."
• William Butler Yeats
OSI Model
OSI Model
• The Application layer is where the tools
are located to support a class of software we
are familiar with. An example would be
SMTP, an application used by most email
software like Outlook and Netscape Mail to
carry email from source to destination.
OSI Model
• The Presentation layer is the encoding
layer. It is responsible for providing the
data in the proper format. For example, text
can be presented in many formats, including
ASCII and ANSI. A picture can be
displayed using either JPEG or PICT. It’s
the same information, only the encoding
technique is different.
OSI Model
• The Session layer establishes the dialogue
between applications. In a Telnet session, it
defines the login procedure and maintains
that conversation until the parties log out or
disconnect, then it handles the disconnect
process.
OSI Model
• The Transport layer handles the delivery
of complete messages from one network to
another. It keeps all message packets in
order, manages the flow of the data, and
offers a level of error detection and
correction.
OSI Model
• The Network layer delivers packets from
one network to another. It deals with
routing and addressing of packets. This is
the Internetworking layer, and is associated
with a router.
OSI Model
• The Data Link layer deals with frames
within a network. It doesn’t understand
going from one network to another. It
provides a local form of addressing (like
MAC addresses), and offers another layer of
error detection and correction that’s used
within a network. This layer is associated
with Bridges and Switches
OSI Model
• The Physical layer looks at bits. It address
issues like the number of pins on a
connector, the signal type (eg control or
data), encoding scheme (eg Manchester
Encoding), and signal level (eg 5 volts or 12
volts). It has nothing to do with addressing.
To define a network type like Ethernet
requires both the Physical and Data Link
layers. This layer is associated with a
Repeater.
TCP/IP Model
http://www.unm.edu/~network/presentations/course/chap2/sld018.htm
Comparison of OSI and TCP/IP
http://dast.nlanr.net/Training/DCWJuly99/kai_tcpip/sld005.htm
Comparison of OSI and TCP/IP
http://www.microsoft.com/windows2000/en/server/help/sag_TCPIP_ovr_model.htm
Encapsulation
Network Hardware
• Networks have evolved over time, but today
we chiefly use Hubs to interconnect our
LAN devices using a star topology,
providing one port per device. These hubs
are usually either Ethernet Repeaters or
Switches.
Network Hardware
Hub
Tower PC
Tower PC
Tower PC
Server
Network Hardware
• Repeaters keep devices all on a single
collision domain. Everything connected to
the same repeater is on the same network,
regardless of which port it is connected to.
• Switches provide a separate collision
domain for each port, but provide a single
broadcast domain, all within the same
network.
Network Hardware
• Routers are placed at the end of a network
to provide access to another network. They
are Internetworking devices.
Ethernet
• There are lots of different LAN types,
including ARCnet, Token Ring, FDDI, and
ATM, However, in this class we will
chiefly work with Ethernet.
Ethernet
• An Ethernet definition from Webopedia:
• A local-area network (LAN) architecture
developed by Xerox Corporation in cooperation
with DEC and Intel in 1976. Ethernet uses a bus or
star topology and supports data transfer rates of 10
Mbps. The Ethernet specification served as the
basis for the IEEE 802.3 standard... Ethernet uses
… CSMA/CD …. It is one of the most widely
implemented LAN standards.
Ethernet
• Webopedia definition, continued:
• A newer version of Ethernet, called
100Base-T (or Fast Ethernet), supports data
transfer rates of 100 Mbps. And the newest
version, Gigabit Ethernet supports data rates
of 1 gigabit (1,000 megabits) per second.
• (10Gbps is also now available)
Ethernet
• The original Ethernet was based on a single
coaxial cable to which all devices were
attached. These days we mostly use a hub,
UTP cable and a star topology.
Ethernet
Tower PC
Tower PC
Tower PC
Server
Ethernet
• Frames come in a variety of formats,
including Ethernet_II, 802.3, 802.2 and
SNAP. Each differs slightly from the next
and they are all more-or-less equivalent.
The important part is that all devices on a
network must use the same frame format.
Ethernet
• Frame addressing uses the 48 bit MAC
address (Media Access Control). The
address is typically displayed in
hexadecimal, with six two digit fields of 8
bits each separated by colons. E.g.
00:80:c7:a8:f0:3d. A broadcast frame is
addressed with all 1s. In hex it looks like
this: ff:ff:ff:ff:ff:ff.
Ethernet
• Ethernet at the physical layer is identified
with a standardized format that often
indicates data rate, signaling type, and the
maximum length of cable in hundreds of
meters. Examples would be 10base5 and
100baseT. These specifications are often
driven by the cable selection.
Ethernet
• Ethernet cable comes in a variety of forms,
including wireless, several coaxial cables,
and several fiber optic cables. Each cable
type has a choice of connector type. The
format specified defines what kind of cable
and connector you will be using.
Ethernet
• Ethernet cabling assumes serial
communication. That means one bit at a
time. When Ethernet uses coaxial cable
there is a single data path, so
communication is half duplex. Twisted pair
or fiber optic cable both use separate
transmit and receive paths, so there is a
potential for full duplex communication.
Ethernet
• Ethernet twisted pair cabling uses balanced
signaling, so there is a transmit pair and
receive pair.
• On a UTP cable used for Category 3, 4, 5,
etc. an RJ45 connector with 8 pins is used.
Only 4 wires carry the signal.
• Ethernet uses straight through cables.
Ethernet
• Since Ethernet uses straight through cables
(pin 1 on one end connects to pin 1 on the
other end, 2 to 2, 3 to 3, etc.), the idea of
DTE and DCE is used.
• Individual signals are paired in such a way
that they will carry the signal best.
• A standard straight cable connects from a
host (DTE) to a hub (DCE).
Ethernet
• You can connect directly from a host
directly to another host for a 2 node
Ethernet network.
• You can connect directly from a hub to
another hub in order to enlarge your
network.
• In these cases, a crossover cable must be
used.
Ethernet
• A crossover cable is used when a DTE talks
directly to a DTE, or DCE to a DCE.
• To build this cable, take the transmit pair
from each end of a cable and connect them
to the receive pins on the other end of the
cable.
• Hubs often have a crossover function built
in.
Ethernet
Compliments of Charles Spurgeon
Ethernet
• A rollover, or rolled cable is used by Cisco
as a terminal cable. It is constructed by
crossing all signals; pin 1 to pin 8, pin 2 to
pin 7, etc. Cisco assumes you have an RJ45
jack on your PC as a serial port. However
most PCs use a 9 pin connector for serial
communications, so an RJ45 to 9 pin
adapter must also be used.
Ethernet
Compliments of Cables Online.com
Cisco Product Hierarchy
• Cisco uses the concept of Core Layer,
Distribution Layer, and Access Layer to
identify the target use for its products. They
are Cisco terms and although they are not
standardized the concept is valid. Suffice it
to say that some places in your network will
experience higher traffic rates than other
places, so you buy appropriately sized
hardware.
Ethernet
• Ethernet uses Differential Manchester Encoding as
its signaling method.
• This technique ensures synchronization, but it
experiences more transitions than some other
signaling techniques, thus limiting the maximum
data rate.
• The bit is identified by the presence or absence of
a transition at the beginning of the bit time.
Ethernet
Data Communications & Computer Networks, by Curt White
Internetworking
• Connecting one network to another
typically involves a link from the telephone
network.
• These digital telephone lines use twisted
pair cable and run at a multiple of 64 Kbps
• They are called T carrier services. A T-1
line provides a 1.544 Mbps point-to-point
connection.
Internetworking
• T-carrier services are charged based on
distance and data rate.
• Delivery is typically copper, using an RJ48
connector (which has the same physical
configuration as an RJ45 connector).
• These are four wire services rather than the
analog phone line which uses two wires.
Internetworking
• ISDN (a switched service) and DSL (a
similar, but non-switched service) are also
used to provide connectivity between nodes.
Internetworking
• A digital line might instead terminate at an
intermediary WAN to carry traffic to a
destination.
• Some WAN technology examples would be
Frame Relay, X.25, and the Internet.
Internetworking
• MANs may use SMDS (Switched
Multimegabit Data Service) for high speed
network connections over short distances
within range of these local networks. They
are typically not accessed via the telephone
company.
Internetworking
• To access T-services we need:
• a router to provide addressing and layer 3
services
• A data conversion device like a CSU/DSU
to convert the data from a LAN format to a
format compatible with the telephone
company.
Tools
•
•
•
•
•
•
For next week, look for these tools:
T-Stripper http://www.idealindustries.com/ht/WireStrippers.nsf
Punch tool http://www.action-electronics.com/paladin.htm
RJ45 crimper http://www.9thtee.com/networking.htm
Screwdriver
Wire cutters http://www.beadshop.com/wirecutter.html