Network+ Important Stuff.

Broadcast Topologies.

• Ethernet is an example of a bus topology – Token Ring and FDDI are

examples of ring topologies.
• Token Ring nodes are never connect together physically in a ring. The stations
are connected physically to a hub in a star topology. The hub electronically
implements a logical ‘ring-in-a-box’.

Data Link Sublayers.

• The LLC (Logical Link Control) Sublayer, which accepts and delivers
packets from the network layer above. It allows more than one network layer protocol
to be used.
• The MAC (Media Access Control) Sublayer, which is responsible for
generating frames appropriate to the particular network interface in use. It allows
more than one network adapter to be supported.

IEEE Standards.

• 802.2 – the IEEE standard for LLC layer. It not only defines an interlink
between the media access methods and the Network layer of the OSI model, but also
provides framing, addressing and error control.
• 802.3 (CSMA/CD) – A MAC layer implementation – usually know as
Ethernet – which defines how various types of cabling work with Ethernet, Fast
Ethernet and Gigabit Ethernet.
• 802.5 – also a MAC layer implementation, it is a specification for Token Ring.
It uses a star configuration at speeds of 4Mbps and 16Mbps.
• These standards now also come under the aegis of ISO e.g. ISO 8802-2.

Bridges

• They examine the header of each frame to determine the source and
destination MAC addresses.

Spanning Tree Algorithm.

• STA avoids bridging loops by ensuring that some bridge interfaces do not
participate in bridging – they are ‘turned off’. These interfaces can be ‘turned on’
again if another bridge fails and the interface is required to maintain connectivity.

1
 Routing vs Bridging.

• Routers do not normally forward broadcasts.

• Internetworks should be built around routers where possible.
• Bridges should be used to limit traffic within LANs.

Brouters.

• Non-routable protocols such as NetBEUI.

• Can forward both Routable and non-routable traffic.

Typical IRQs used by devices.

• 3 – Available (unless used for 2nd serial port [COM2, COM4] or bus mouse).
• 4 – COM1, COM3
• 6 – Floppy-disk controller.
• 7 – Parallel port [LPT1].

Typical Base I/O Port setting used by devices.

• Port: 300 to 30F Device: Network adapter card.

• Port: 310 to 31F Device: Network adapter card.

Thinnet Connectors and Terminators.

• The lightweight specification of thinnet results in signal attenuation

(weakening of the signal) over distances in excess of 185 metres – which is the
recommended maximum length of a run of thinnet cable.

Safety considerations – Plenum Grade Cable.

• Standard PVC based cable gives off poisonous fumes when it burns. For this
reason, cabling which is to be installed above false ceilings or under floor should be
of Plenum grade.

Cable Descriptions.

Cable Description

RG-58/U Solid copper core.

RG-58 A/U Stranded wire core.
RG-58 C/U Military specification of RG58 A/U
RG-59 Broadband transmission such as cable television.

2
 RG-6 Larger in diameter and rated for higher frequencies than RG-59, but
used for broadband transmissions as well.
RG-62 ArcNet networks.

Straight-Through Cable Connections.

• Straight-through cables are used to connect end stations, such as PCs and
routers, to hubs.

Cable Categories.

• Category 1 – This refers to traditional UTP telephone cable, which can carry
voice but not data. Most telephone cable prior to 1983 was Category 1 cable.
• Category 2 – This category certifies UTP cable for data transmissions up to 4
Mbps. It consists of four twisted pairs.
• Category 3 – This category certifies UTP cable for data transmissions up to
10 Mbps. It consists of four twisted pairs.
• Category 4 – This category certifies UTP cable for data transmissions up to
16 Mbps. It consists of four twisted pairs.
• Category 5 – This category certifies UTP cable for data transmissions up to
100 Mbps. It consists of four twisted pairs of copper wire.

Structured Cabling.

• The 100m limit on the length of cable between NIC and hub for Ethernet
includes the length of any patch cables used. These should always be included in any
calculations.

Synchronous Modems.

• Because of the higher data rates supported by leased lines, they are generally
used in conjunction with synchronous modems. These modems ensure that sender
and receiver stay ‘in sync’ and so enable longer streams of data to be sent before
being checked for possible corruption.

Asynchronous Modems.

• The cheapest modems are asynchronous. This means that they do not
exchange timing signals, which would involve more expensive circuitry. The lack of a
common clock at either end means that the sender and receiver could get ‘out of sync’
when transferring long streams of data. Data is transferred in short bursts. This also
has the effect of slowing down communications.

3
 However, dial-up telephone connections still account for most of the Internet
access even today. The problem with using voice-quality telephone lines is
that they were designed to carry analogue voice signals not digital data
signals.

UART.

• The Universal Asynchronous Receiver/Transmitter (UART) takes the parallel

data stream from the Bus and turns it into a serial bit stream to send to the modem.

RS-232 Connector.

• RS-232 supports the exchange of timing signals for synchronous

communications.

DB-9 Connectors.

• For connection to an asynchronous modem.

Null Modem Connection.

Note on Terminology:

• DTE – Data Terminal Equipment – a standard designation for a computer or

terminal attached to a modem.
• DCE – Data Circuit-terminating Equipment - a standard designation for a
modem.

Modem Speeds.

• Bit Rate – This is the number of bits transmitted per second.

• Baud Rate – Modems use a variety of signal transitions – frequency
transitions, amplitude transitions, phase transitions, as well as combinations of these –
to transmit data. The number of transitions per second is the baud rate.

4
 Standard Bps Introduced Notes
V.22BIS 2400 1984 An old standard. Sometimes included with
the purchase of a computer.
V.32 9600 1984 Sometimes included with the purchase of a
computer.
Kflex56 57,600 1996 Lucent/Motorola, incompatible with x2.
V90 57,600 1998 Emerging 57,600 standard.

Modem Troubleshooting.

• COM 3 uses IRQ4 and I/O address 03E8h.

• COM 4 uses IRQ3 and I/O address 02E8h.

Command (AT +) Example Meaning

D ATD01818298229 Dial 0181 829 8229
H ATH Hang Up
10 – 19 AT10, AT11 etc Displays modem info and other functions.

Collision Detection.

Before sending data, CSMA/CD stations ‘listen’ to the network to see if it is already
in use. If it is, the station wishing to transmit, waits. If the network is not in use, the
station transmits. A collision occurs when two stations listen for network traffic,
‘hear’ none, and both transmit. In this case, both transmissions are damaged and
this fact is detected by each station, which must retransmit at some later time.
Random back-off algorithms determine when the colliding stations retransmit. The
randomness helps to avoid the situation where both stations try to retransmit at the
same time.

Maximum Cable Length and Minimum Ethernet Frame Length.

Stations detect collisions by ‘listening’ to the channel as they transmit – if what

they ‘hear’ is not what they are sending, then a collision has occurred. In order for
both stations to detect a collision, each must still be transmitting when the others
frame arrives. This condition is responsible for the limitation on the maximum
length of the cable and the minimum length of an Ethernet frame (64 bytes).
Although the signal travels along the channel at near the speed of light, it still
requires a finite time to travel from one end of the cable to the other – the longer
the cable, the bigger the frame needs to be in order to ensure that any two stations
will both detect the collision. If a station detects a collision before it has transmitted
64 bytes it broadcasts a jamming signal to ensure that the other station also detects
the collision.

5
Ethernet Limitations.

The 5-4-3 Rule for Thicknet and Thinnet.

No more than five segments of Ethernet cable may be connected by four repeaters
(providing that only three segments are active – i.e. have stations attached). This to
aid collision detection described earlier.

FDDI Fault Tolerance.

FDDI provides a number of fault-tolerant features. The primary fault-tolerant

feature is the dual ring. If a station on the dual ring fails or is powered down or if
the cable is damaged, the dual ring is automatically ‘wrapped’ (doubled back onto
itself) into a single ring as shown above. In this diagram, when station 3 fails, the
dual ring is automatically wrapped in stations 2 and 4, forming a single ring.
Although station 3 is no longer on the ring, network operation continues for the
remaining stations.
In a similar way, FDDI can compensate for a wiring failure: stations wrap the ring
within themselves when wiring between them faults.

TCP.

TCP provides a connection-orientated, reliable, sequenced byte-stream service to

applications. TCP only provides one-to-one communications (unicasts). Sessions
must be established between communicating applications. TCP accepts a byte
stream from applications, which it breaks into segments.
Reliability is ensured by a system of sequence numbers applied to each segment. If
the destination host does not acknowledge a segment within a certain timeout
period, the segment is retransmitted. If a segment is persistently unacknowledged,
TCP will close the session and inform the application that the connection has
failed.
Error control is also provided by checksums, calculated from both the headers and
data of each segment.

UDP.

UDP provides a connectionless, unreliable transport service. Delivery of UDP

datagrams is not guaranteed: instead a ‘best effort’ to deliver is made. Applications
using UDP must supply their own mechanisms to ensure reliable delivery if this is
required. They must also present the transport layer with data in the form of
datagrams, to which UDP adds a simple header.

6
 Classes of IP Address.

5 Classes – A, B, C, D and E.

Class 1st Octet 2nd Octet 3rd Octet 4th Octet

A NetID HostID HostID HostID
B NetID NetID HostID HostID
C NetID NetID NetID HostID

• Class A addresses were designed for networks with a very large number of
hosts.
• Class B addresses were designed for medium-sized networks.
• Class C addresses were intended for small networks, such as LANs.
• Class D addresses are used for multicasting to a set of hosts – e.g. all
routers. Only those hosts registered for the multicast address will receive the packet.
• Class E addresses are reserved for experimental use.

Class 1st Octet Bit Pattern 1st Octet Network and Hosts
per Network
A 0XXXXXXX 1 - 126 126 Networks
16,777,214 Hosts
B 10XXXXXX 128 - 191 16,382 Networks
65,534 Hosts
C 110XXXXX 192 – 223 2,097,150 Networks
254 Hosts
D 1110XXXX 224 – 239 Multicast

E 1111XXXX 240 – 255 Reserved

IP Address Restrictions.

Network ID

• All bits set to 0 – reserved for Local Network.

• All bits set to 1 – reserved for Broadcast.

Host ID

• All bits set to 0 – reserved for NetID.

• All bits set to 1 – reserved for Local Broadcast.

• If the Network ID has all bits set to 1, then this is considered to be a

‘broadcast to all networks’ address. However, by default, routers do not forward
broadcasts to other networks, as this could cause excessive network traffic.

7
 Static vs Dynamic Routing.

• These can be populated manually – static routing – or through a routing

protocol such as Routing Information Protocol (RIP) or Open Shortest Path First
(OSPF) – dynamic routing.

Dynamic Routing Protocols.

• There are a number of IP routing protocols. The most widespread is

probably RIP. RIP is an open standard. Another open routing protocol is OSPF.
These two protocols represent two different approaches to dynamic routing –
Distance-Vector routing and Link-State routing. Another widely used routing
protocol is Cisco’s proprietary Internet Gateway Routing Protocol (IGRP), but is
beyond the scope of this course.

Routing Information Protocol.

• Each router broadcasts its routing table to all other routers on connected
subnets. Each router compares its own routing table with the broadcast from its
neighbour. Should a shorter path to a network be discovered, it then updates its
own table. Routes utilising the least number of hops are considered optimal. A hop
count of 16 is considered unreachable.

Problems with RIP.

• Slow Convergence – failed routes can take a long time to propagate through
the network. Spurious routes (routing loops) can be learnt while the network is
unstable. Hence the needs TTL to kill endlessly looping packets that would otherwise
congest the network.
• Network Traffic – broadcasts are frequent and contain the entire routing
table.
• Scalability – as the network becomes large the size of routing tables increases
proportionately.
• Fixed Subnet Masks – in order to minimise network traffic, subnet masks are
not broadcast along with routes. The protocol relies on the subnet mask being the
same throughout the given class A, B, or C network.

Open Shortest Path First (OSPF).

8
 OSPF uses a Link State algorithm – each router tells the world (other routers) about
its neighbours.
OSPF does not use undirected broadcasts. Instead it unicasts to other routers using
reliable TCP connections, thus reducing network traffic.
Link state advertisements, (LSAs), consisting of a router’s attached network Ids and
interface costs are exchanged between routers. LSAs are advertised upon startup and
when changes in the internetwork topology are sensed. LSAs are sent using directed
or multicast traffic not broadcasts.
Routers employing link state algorithms build a database of LSAs and use the
database to calculate optimal routes. These are then added to the routing table. All
routing information exchanged between link state routers is synchronised and
acknowledged.

Advantages of OSPF

• Scalability – OSPF routers only send information about neighbours, so that

the routing updates do not increase proportionately to the size of the network –
making OSPF a better option for large networks.

• Variable Subnet Masks – because information is being sent less frequently,

OSPF can include subnet mask information in its routing updates – permitting the use
of variable subnet masks.

Classless Inter-Domain Routing CIDR.

Also know as supernetting, CIDR may be used to consolidate several class C network
addresses into one logical network. CIDR is described in RFC1518/1519. to use
supernetting, the IP network addresses that are to be combined must share the same
high-order bits, and the subnet mask is shortened to take bits away from the network
portion of the address and add them to the host portion.

To prevent class B network Ids from being depleted ISOC allows ranges of class C
networks to be offered out.

The above scheme would lead to large numbers of network entries on the routers of
Internet Service Providers (ISPs). To avoid this, ISOC supports supernetting, which
offers out ranges of class C addresses as though they were single subnets of class B
networks thus reducing the number of entries in routing tables.

Ipv6.

• 128 bit addresses

• Displayed in Dotted Hexadecimal (up to 32 digits).
• Very large address space.
• Extensible Headers – for future standards and technologies.

Remote Execution Utilities.

9
 Remote Login (RLOGIN) – as the name implies, this utility allows a user to log in to
a remote system.

Printing Utilities.

A UNIX (or other) server running lpd (line printer daemon) acts as a print server.

• Line Printer Remote (LPR) – submits a print job to a print queue on a

remote host running a line printer daemon.
• Line Printer Queue – queries the status of a print queue on a remote host
running a line printer daemon.

Diagnostic Utilities.

Microsoft Windows environments.

• PING (Packet InterNet Groper) – PING is a tool that helps to verify IP-level
connectivity. When trouble shooting, the PING command is used to send an ICMP
echo request to a target name or IP address.
• TRACERT – TRACERT is a route tracing utility. TRACERT uses the IP
TTL field and ICMP error messages to determine the route from one host to another
through a network.
• PATHPING – PATHPING combines the features of PING and TRACERT,
while also showing the packet loss rate at each router between source and destination.
• IPCONFIG and WINIPCFG – IPCONFIG is a command-line utility that
prints out the TCP/IP-related configuration of a host. When used with the /ALL
switch, it produces a detailed configuration report for all interfaces, including any
configured serial ports (RAS). IPCONFIG is used on Windows NT and Windows
2000 systems. WINIPCFG is the Windows 9x implementation.
• FINGER – this connectivity command displays information about a user on a
specified system running the Finger service. Output varies based on the remote
system.
• NSLOOKUP – was added in Windows NT 4.0 and is a very useful tool for
troubleshooting DNS problems, such as a hostname resolution. A query is sent to a
DNS server to resolve a name to an IP address or vice-versa.
• HOSTNAME – returns the local computer host name.
• NETSTAT – displays protocol statistics and current TCP/IP connections.
• NBTSTAT – is a useful for troubleshooting NetBIOS name resolution
problems.
• ROUTE – this diagnostic command manipulates network routing tables.
• ARP – the ARP command is useful for viewing the ARP cache of IP
addresses resolved to hardware addresses.

Ports and Sockets.

10
 A port number is a 16-bit number used to identify an application on a host. Port
numbers in the range 1 – 1024 are assigned (IANA) to well-known server
applications. They are referred to as well-known port numbers.

Well-known Port Number Application

20 FTP (File Transfer Protocol) – Data.

21 FTP (File Transfer Protocol) – Control.
23 Telnet.
25 SMTP (Simple Mail Transfer Protocol).
53 DNS (Domain Name System).
80 HTTP (Hypertext Transfer Protocol).
110 POP3 (Post Office Protocol – Version 3).

A port number is enough to specify an application within a host system. However,

more information is needed to specify uniquely a target application across a network.

This is achieved by a socket, which a triple of numbers comprising:

• IP Address.
• Protocol Number.
• Port Number.

Multiple Subnets.

• DHCP broadcasts will not pass through routers unless they act as BOOTP
Relay Agents (i.e. unless they support RFC 1542). If a DHCP server is to serve
multiple subnets, the RFC 1542 routers will forward the broadcast requests to the
DHCP server.

Hostname Resolution using the HOSTS File.

• One way of resolving host names or FQDNs is to use a local HOSTS file,
which can be created using any standard text editor. This method is suitable for small
installations, which do not access many internet hosts, as it requires files to be
maintained manually on each computer.

Format of the HOSTS File.

A text-based file of the form:

IP_address hostname [alias alias …] # Comment

Each entry is limited to 255 characters.

The file usually contains the following entry:

127.0.0.1 localhost loopback

11
 which defines the names for local loopback testing.

The file is read sequentially till a match is found for the requested name. Whether the
entries are case-sensitive depends on the operating system.

DNS Name Space.

• DNS Name Space – Hierarchical Name Space providing a unique Fully

Qualified Domain Name (FQDN) for every host on the internet.
• Each domain has the authority for the names immediately under it.
• Root-level Domain – Top of the hierarchy responsible for the Top-Level
domains. Has a NULL label but can be represented by a dot (.).

NetBIOS Name Resolution using the LMHOSTS File.

Computer names can be resolved if the NetBIOS name and IP address mappings are
specified in the LMHOSTS file – a simple text file.

The LMHOSTS text file has the format:

IP Address NetBIOS Name #Keyword # Comment

Name Resolution by Broadcast.

Limitations of Broadcast.

Routers do not usually forward broadcasts. Some routers can be configured to forward
broadcasts for UDP ports 137 and 138, however this might cause network traffic
problems. Therefore name resolution by broadcast is typically confined to the
local subnet only.

• HTML – the Hypertext Markup Language – a plain text language for creating
and formatting documents, based on the principles of non-hierarchical Hypertext
navigation, devised by Ted Nelson in 1965.
• URL – the Uniform Resource Locator – a convention for finding and
accessing such documents on the Web.
• HTTP – the Hypertext Transfer Protocol – a TCP/IP protocol allowing
computers to transmit and request documents.

Firewalls.

• A firewall is a secure interface between a local network and the public

internet, designed to prevent outsiders from accessing the LAN, and insiders from
carrying out any potentially hazardous or inappropriate activity with the outside.

12
 Proxy Servers.

• The Proxy Server establishes internet sessions on behalf of its clients (using its
valid IP address 200.2.2.31). it communicates with its clients using addresses in a
reserved network, such as 10.0.0.0 (invisible to the Internet).

Using E-mail.

• POP3 is an internet protocol defining standards for e-mail communication

(mail retrieval).

Simple Mail Transfer Protocol – SMTP.

• SMTP specifies a standard for the exchange of electronic mail messages

between systems.
• To send mail, the client system opens a TCP connection to the server using
well-known port 25.

File Transfer Protocol – FTP.

• FTP uses TCP as its transport protocol. In addition to basic file transfer it
provides:
• Interactive Access – allowing users to list files in the remote directories they
need to access.
• Format Representation – allowing users to specify the type and the format of
stored data e.g. binary or text (ASCII or EBCDIC).
• Authentication control – via login name and password (though not
encrypted).

Trivial File Transfer Protocol – TFTP.

• The Trivial Transfer Protocol is a lightweight file transfer protocol, often used
in devices, which do not have any disk storage. On these devices, software is often
stored in ROM or flash memory so small programs are desirable. Typically, these
devices use TFTP to download operating system and configuration files from servers.
The servers need to be located by broadcast, so UDP rather than TCP must be used
as the transport protocol.

Terminal Emulation – TELNET.

13
• TELNET is a remote terminal protocol that uses TCP as its transport protocol
and allows a user at one site to establish a connection, using well-known port
number 23, to a login server at another site.

Simple Network Management Protocol – SNMP.

• SNMP is a protocol that allows computers using SNMP management software

to interrogate and control computers running SNMP agents. (Remember sets Traps).

Authentication Traps

• Agents can be configured to send a trap to their managers, if they receive a

request for information from a manager outside their community.

Making Network Resources Available.

• A directory or printer on the machine, which is acting as a server, is selected

and shared to the network with a special name – the sharename, which can be up to
15 characters in length in a Microsoft environment.
• Any remote system wishing to access the resource will use a combination of
the servername – the computername of the system offering the shared resource – and
sharename in the format \\servername\sharename. This is known as the Universal
Naming Convention (UNC) format.

Choosing a File System.

• For multiple OS boot configuration, boot partition must be FAT (NTFS

cannot read files from DOS.

Differential Backup Strategy.

• If the amount of data is too large to make frequent full backups impractical, a
differential backup strategy can be used. Full backups are performed during slack
times – perhaps the weekend. They may occupy multiple tapes. More frequent
differential backups can be performed daily to back up only the data that has changed
since the last full backup.
• There should be multiple sets of tapes that can be rotated to provide better
resilience against tape corruption.

Incremental Backup Strategy.

14
• If data is changing rapidly and there is a lot of it, differential backup may
become impractical, as the size of the backups will increase with time from the last
full backup. Incremental backups only backup data since the last backup. The speed
and size of the backups is reduced. Entire sets of tapes should be considered for
recycling.
• A downside is that in the event of having to restore data, the full backup
tape and all subsequent incremental tapes must be processed. This lengthens the
recovery time. In the case of the differential strategy, only the last full backup
and the latest differential backup tape need to be processed.

Network Troubleshooting.

• A great deal can be learnt about network infrastructure problems through the
use of a standard software. The use of the PING and TRACERT commands can
pinpoint network components – usually hubs, cables or devices such as routers and
bridges – which are causing problems. Once the search has been narrowed down, it is
often necessary to use other tools to determine the exact nature of the problem. Some
of the more common techniques are described below.

Device LEDs.

• Most network devices – including LAN adapters – have one, or more, LED
(Light Emitting Diode) Indicators, which report the status of the device. Check the
documentation that accompanied the device to determine whether the LEDs indicate
that it is functioning correctly.

Hardware Loopback.

• This is a device that connects the transmit pins to the received pins of an
Ethernet 10BaseT NIC, and is supplied by the manufacturers of the NIC if the
diagnostic software supports the hardware loopback option. This device checks the
send and received capabilities of the NIC without sending signals onto the entire
network. It is not possible to fully test a NIC without a hardware loopback connector.

Terminators.

• Ethernet cables need to be terminated – by a 50 ohm resistance in the case of

10Base2. hub-based systems have termination built into the hubs themselves. When
checking a 10Base2 system, ensure that all the terminators are in place then, if
necessary, use a multimeter- see below – to ensure that the resistance between the
inner and outer conductors of the terminator is correct.

Multimeters.

15
• Separate individual Voltmeters, Ammeters, Ohmmeters etc have been replaced
by devices known as Multimeters capable of operating in different modes measuring,
as a minimum, the following:
• AC Voltage.
• DC Voltage.
• Resistance.
• Continuity of current.

Analogue and digital Multimeters are available:

• Analogue multimeters allow manual adjustments to the test resistance of the

device to compensate for internal test cell resistance and employ a meter or moving
arm to indicate any values obtained.
• Digital multimeters are self-calibrating, requiring no manual adjustment and
employ LED or LCD readouts or displays.

A dial on the multimeter is usually set to select the required sensing function.

Digital multimeters usually provide more test functions than an analogue device –
often including transistor testing, conductivity readings, capacitor efficiency etc.

Digital multimeters are able to measure standard current ranges more accurately but
typically cost more than the analogue devices.

Cable Testers.

• There are a variety of cable testers available. These are used to check for
breaks in the cable, poor signal quality and whether the wires have been correctly
connected to the terminators at each end.

Time Domain Reflectometer (TDR).

• A TDR is a specialised form of cable tester. It is used to detect breaks in a

cable as well as to determine the distance of the break along the cable. It does this by
sending a signal down the cable and measuring the time for the echo to return from
the break.

Protocol Analysers and Network Monitors.

• We have already seen an example of a network monitor in this course.

Microsoft’s Network Monitor. Protocol analysers and network monitors have similar
functionality – capturing frames and analysing their structure and contents. By
examining the data exchanged between tow network applications, they can b used to
determine problems with the applications. Protocol analysers are usually implemented
as dedicated hardware solutions, while network monitors are usually software
packages designed to run on a network computer.

Test Frame and Packet Generators.

16
• These may be included in protocol analysers or may be provided as separate
products (usually implemented as software running on a network computer). They can
generate packets of specific types (to test particular protocols) and in large numbers
(to test the network under the stress of heavy traffic).

Tone Generator and Probe (Fox and Hound).

• A Tone Generator puts an electrical signal onto a cable. The associated probe
can pick up emissions from the signal on the cable – even at some distance. This can
be used to check for breaks in cables and to trace the location of cable when they are
in inaccessible places – e.g. under-floor.

SNMP Based Tools.

• There are a number of network management products on the market, which

use SNMP to gather information about network devices – as well as to control those
devices remotely. For these products to work effectively, the devices must run SNMP
Agent software. One or more systems on the network will run the management
software. This usually displays the information gathered in an easy-to-use graphical
format. Examples of such software include HP OpenView and CiscoWorks.

Device Diagnostic Software.

• Many network devices have diagnostic commands built into their operating
system software. These can be very useful in identifying problems with the device
itself as well as with attached cables and other devices with which it communicates.
For example, Cisco routers have an extensive range of DEBUG and SHOW
commands.

High Level Data Link Control – HDLC.

• HDLC is a layer 2 protocol, widely used in leased line networking. It is based

on an original IBM protocol and is the basis of derived protocols such as PPP.
• One of the limitations of HDLC is that the frame header does not contain
a field specifying which layer 3 protocol is being encapsulated. Consequently, it
can only be used in single network layer protocol environments.
• Cisco has a proprietary version of HDLC that does support multiple layer
3 protocols, but it is not inter-operable with standard HDLC.

Point-to-Point Protocol.

• PPP is a Data Link layer 2 protocol, which is widely used in (but not confined
to) switched-circuit network environments.

17
 Serial Line Internet Protocol – SLIP.

• SLIP is an earlier Data Link layer protocol, still found in many UNIX
environments. As the name implies, it is limited to encapsulating IP packets. This is
not its only limitation with respect to PPP, as can be seen from the comparison below.
It should only be used in circumstances where one of the systems does not support
PPP.

SLIP PPP
Only supports IP Multiprotocol support
Limited to 56kbps Not limited to 56kbps
Unencrypted Authentication Encrypted Authentication
Does not support Data Compression Supports Data Compression
Does not support Error Correction Supports Error Correction
Requires Configure IP Addresses Supports IP Address Negotiation
Asynchronous Transfer only Asynchronous and Synchronous Transfer

Dial-up Networking via the Internet.

• This sort of connection is made possible by using Point-to-Point

Tunnelling Protocol (PPTP), which allows frames to be encapsulated in IP
packets for transmission across the Internet.

Advantages:
• Lower transmission costs.
• Lower hardware costs, as only the RAS server needs to be connected to the
Internet from the corporate LAN.
• Good security – all connections are encrypted.

X.25 and the OSI Reference Model.

• The X.25 specification maps to the Physical, Data Link and Network
layers of the OSI reference model.
• Layer 3 X.25 describes packet formats and packet exchange procedures
between peer Layer 3 entities. Layer 2 X.25 is implemented by Link Access
Procedure, Balanced (LAPB).

Asynchronous Transfer Mode – ATM.

• Designed for Voice, Video and Data Transfer.

• Based on cells.
• Small data payload.

18
• Fixed length – fast, large modular switches.
• ATM is a cell-switching and multiplexing technology that combines the
benefits of circuit switching (constant transmission delay and guaranteed capacity)
with those of packet switching (flexibility and efficiency for intermittent traffic).

Overview of Network Operating Systems.

• The central purpose of a network operating sytem is to make local resources

available to a network. For example, a user at one computer on the network may
access file on another computer. Printers are another class of shared resource.
• In addition to making resources available, it is also very desirable to ensure
secure access to those resources. Only the users who need a resource should be
allowed to access it.
• In order to manage these two requirements efficiently, most modern
network operating systems also allow centralised administration of resources.

Novell NetWare.

• Netware 3.x uses Bindery (database of users, groups and resource

permissions).
• Netware 4.x uses the NetWare Directory Service (NDS), Novell’s
implementation of an X.500 directory service.

19