NETWORKING

Basic of Networking
A computer network consists of a collection of computers, printers and other equipment that is
connected together so that they can communicate with each other. Fig 1 gives an example of a
network in a school comprising of a local area network or LAN connecting computers with each
other, the internet, and various servers.
roadly speaking, there are two types of network configuration, peer!to!peer networks and
client"server networks.
#eer!to!peer networks are more commonly implemented where less then ten computers are
involved and where strict security is not necessary. All computers have the same status, hence the
term $peer$, and they communicate with each other on an equal footing. Files, such as word
processing or spreadsheet documents, can %e shared across the network and all the computers on
the network can share devices, such as printers or scanners, which are connected to any one
computer.
&lient"server networks are more suita%le for larger networks. A central computer, or $server$, acts
as the storage location for files and applications shared on the network. 'sually the server is a
higher than average performance computer. (he server also controls the network access of the
other computers which are referred to as the $client$ computers. (ypically,teachers and students in
a school will use the client computers for their work and only the network administrator )usually
a designated staff mem%er* will have access rights to the server.
#ort Num%ers
#ort num%ers are the mechanism for identifying particular client and server applications. +ervers
select a port to wait for a connection. ,ost services have well!known port num%ers. For
example, -((# uses port ./. 0hen a we% %rowser )the client* requests a we% page it specifies
port ./ when contacting the server.&lients usually have ephemeral port num%ers since they exist
only as long as the session lasts.+ome of the common well!known (&# port num%ers are listed in
the ta%le %elow.
ICMP
1nternet &ontrol ,essage #rotocol is a set of messages that communicate errors and other
conditions that require attention. 1&,# messages, delivered in 1# datagrams, are usually acted on
%y either 1#, (&# or '2#. +ome 1&,# messages are returned to application protocols.A common
use of 1&,# is 3pinging4 a host. (he #ing command )#acket 1Nternet 5roper* is a utility that
determines whether a specific 1# address is accessi%le. 1t sends an 1&,# echo request and waits
for a reply.#ing can %e used to transmit a series of packets to measure average round!trip times
and packet loss percentages.
(he Application Layer
(here are many applications availa%le in the (&#"1# suite of protocols. +ome of the most useful
ones are for sending mail )+,(#*, transferring files )F(#*, and displaying we% pages )-((#*.
(hese applications are discussed in detail in the (&#"1# 'ser6s ,anual, vols. 1 and 7.Another
important application layer protocol is the 2omain Name +ystem )2N+*. 2omain names are
significant %ecause they guide users to where they want to go on the 1nternet.
DNS
(he 2omain Name +ystem is a distri%uted data%ase of domain name and 1# address %indings. A
domain name is simply an alphanumeric character string separated into segments %y periods. 1t
represents a specific and unique place in the 3domain name space.4 2N+ makes it possi%le for us
to use identifiers such as ra%%it.com to refer to an 1# address on the 1nternet. Name servers
contain information on some segment of the 2N+ and make that information availa%le to clients
who are called resolvers.
Components of a Network
A computer network comprises the following components8
A minimum of at least 7 computers
&a%les that connect the computers to each other, although wireless communication is
%ecoming more common )see Advice +heet 7/ for
more information*
A network interface device on each computer )this is called a network interface card or N1&*
A 9+witch6 used to switch the data from one point to another. -u%s are outdated and are little
used for new installations.
Network operating system software
+tructured &a%ling
(he two most popular types of structured network ca%ling are twisted!pair )also known as
1/ase(* and thin coax )also known as 1/ase7*. 1/ase( ca%ling looks like ordinary telephone
wire, except that it has . wires inside instead of :. (hin coax looks like the copper coaxial
ca%ling that$s often used to connect a ;ideo <ecorder to a (;.
1/ase( &a%ling
0hen 1/ase( ca%ling is used, a strand of ca%ling is inserted %etween each computer and a hu%.
1f you have = computers, you$ll need = ca%les. >ach ca%le cannot exceed ?7= feet in length.
ecause the ca%les from all of the #&s converge at a common point, a 1/ase( network forms a
star configuration.
Fig :a shows a &at=e ca%le, with a standard connector, known as an <@!:= connector.Fig :%
shows a standard &at=e 0all Autlet socket which the ca%les are connected to.Fig :c shows a
standard &at=e #atch #anel 0all Autlet socket which is used to terminate the ca%les from various
points in the school %ank to a central point.Fig :d shows a wall mounted ca%inet used to house
and protect patch panel ca%les and connectors.

1/ase( ca%ling is availa%le in different grades or categories. +ome grades,or BcatsB, are
required for Fast >thernet networks, while others are perfectly accepta%le for standard 1/,%ps
networks!!and less expensive, too. All new networks use a minimum of standard unshielded
twisted!pair )'(#*&ategory =e 1/ase( ca%ling %ecause it offers a performance advantage
over lower grades.
Network Interface Car !NIC"
A N1& )pronounced $nick$* is also known as a network card. 1t connects the computer to the
ca%ling, which in turn links all of the computers on the network together. >ach computer on a
network must have a network card.,ost modern network cards are 1/"1// N1&s and can operate
at either 1/,%ps or 1//,%ps.Anly N1&s supporting a minimum of 1//,%ps should %e used in
new installations students.&omputers with a wireless connection to a network also use a network
card
#$% an Switc&
A hu% is a device used to connect a #& to the network. (he function of a hu% is to direct
information around the network, facilitating communication %etween all connected devices.
-owever in new installations switches should %e used instead of hu%s as they are more effective
and provide %etter performance. A switch, which is often termed a $smart hu%$.+witches and hu%s
are technologies or 9%oxes6 to which computers, printers,and other networking devices are
connected. +witches are the more recent technology and the accepted way of %uilding today$s
networks. 0ith switching, each connection gets Bdedicated %andwidthB and can operate at
full speed. 1n contrast, a hu% shares %andwidth across multiple connections
such that activity from one #& or server can slow down the effective speed of other connections
on the hu%.Now more afforda%le than ever, 2ual!speed 1/"1// autosensing switches are
recommended for all school networks. +tudents may want to consider upgrading any hu% %ased
networks with switches to improve network performance C ie speed of data on the network.
Wire'ess Networks
(he term $wireless network$ refers to two or more computers communicating using standard
network rules or protocols, %ut without the use of ca%ling to connect the computers together.
1nstead, the computers use wireless radio signals to send information from one to the other. A
wireless local area network )0LAN* consists of two key components8 an access point )also
called a %ase station* and a wireless card. 1nformation can %e transmitted %etween these two
components as long as they are fairly close together )up to 1// metres indoors or ?=/ metres
outdoors*.
Fig Da8 0ireless Access point or 0ireless asestation
+uppliers would need to visit the students and conduct a site survey. (his will determine the
num%er of %ase stations you need and the %est place)s* to locate them. A site survey will also
ena%le each supplier to provide you with a detailed quote. 1t is important to contact a num%er of
different suppliers as prices, equipment and opinions may vary. 0hen the term $wireless network$
is used today, it usually refers to a wireless local area network or 0LAN. A 0LAN can %e
installed as the sole network in a school or %uilding.-owever, it can also %e used to extend an
existing wired network to areas where wiring would %e too difficult or too expensive to
implement, or to areas located away from the main network or main %uilding. 0ireless networks
can %e configured to provide the same network functionality as wired networks, ranging from
simple peer!to!peer configurations to largescale networks accommodating hundreds of users.
0hat are the advantages and disadvantages of a 0ireless LANE
0ireless LANs have advantages and disadvantages when compared with wired LANs. A
wireless LAN will make it simple to add or move workstations, and to install access points to
provide connectivity in areas where it is difficult to lay ca%le. (emporary or semi!permanent
%uildings that are in range of an access point can %e wirelessly connected to a LAN to give these
%uildings connectivity. 0here computer la%s are used in students,the computers )laptops* could
%e put on a mo%ile cart and wheeled from classroom to classroom, providing they are in range of
access points. 0ired network points would %e needed for each of the access points. A 0LAN has
some specific advantages8
1t is easier to add or move workstations
1t is easier to provide connectivity in areas where it is difficult to lay ca%le
1nstallation can %e fast and easy and can eliminate the need to pull ca%le through walls and
ceilings
Access to the network can %e from anywhere in the school within range of an access point
#orta%le or semi!permanent %uildings can %e connected using a wireless LAN
0here laptops are used, the 9computer suite6 can %e moved from classroom to classroom on
mo%ile carts
0hile the initial investment required for wireless LAN hardware can %e similar to the cost of
wired LAN hardware, installation expenses can %e significantly lower
0here a school is located on more than one site )such as on two sides of a road*, it is possi%le
with directional antennae, to avoid digging trenches under roads to connect the sites
1n historic %uildings where traditional ca%ling would compromise the faFade, a wireless LAN
can avoid drilling holes in walls
Long!term cost %enefits can %e found in dynamic environments requiring frequent moves and
changes
(hey allows the possi%ility of individual pupil allocation of wireless devices that move
around the school with the pupil.
0LANs also have some disadvantages8
As the num%er of computers using the network increases, the data transfer rate to each
computer will decrease accordingly
As standards change, it may %e necessary to replace wireless cards and"or access points
Lower wireless %andwidth means some applications such as video streaming will %e more
effective on a wired LAN
+ecurity is more difficult to guarantee, and requires configuration 2evices will only
operate at a limited distance from an access point,with the distance determined %y the standard
used and %uildings and other o%stacles %etween the access point and the user
A wired LAN is most likely to %e required to provide a %ack%one to the wireless LANG a
wireless LAN should %e a supplement to a wired LAN and not a complete solution
Long!term cost %enefits are harder to achieve in static environments that require few moves
and changes
1t is easier to make a wired network 9future proof6 for high data
transfer.
0ireless Network &omponents
(here are certain parallels %etween the equipment used to %uild a 0LAN and that used in a
traditional wired LAN. oth networks require network interface cards or network adapter cards.
A wireless LAN #& card, which contains an in!%uilt antenna, is used to connect note%ook
computers to a wireless network. 'sually, this is inserted into the relevant slot in the side of
the note%ook, %ut some may %e internal to the note%ook. 2esktop computers can also connect to
a wireless network if a wireless network card is inserted into one of its internal #&1 slots.1n a
wireless network, an $access point$ has a similar function to the hu% in wired networks. 1t
%roadcasts and receives signals to and from the surrounding computers via their adapter card. 1t
is also the point where a wireless network can %e connected into an existing wired network.
(he most o%vious difference %etween wireless and wired networks,however, is that the latter
uses some form of ca%le to connect computers together. A wireless network does not need ca%le
to form a physical connection %etween computers.
Wire'ess Network Config$rations
0ireless networks can %e configured in an ad hoc"peer!to!peer arrangement or as a local area
network.
( #oc)Peer*to*Peer Config$ration
(his is the most %asic wireless network configuration. 1t relies on the wireless network adapters
installed in the computers that are communicating with each other. A computer within range of
the transmitting computer can connect to it. -owever, if a num%er of computers are networked in
this way,they must remain within range of each other. >ven though this configuration has no real
administration overhead, it should only %e a consideration for very small installations.
Benefits an E$cationa' +ses
(he installation of ca%les is time consuming and expensive. (he advantages of not doing so are
apparent8the amount of work required and the time taken to complete it are significantly reduced
the network is accessi%le in places where wiring would have %een difficult or impossi%le with no
ca%les linking computers together, ca%le!related faults and network downtime are minimiHed.
0here a wireless network is in place, teachers or students can have continuous access to the
network, even as they move with their equipment from class to class.(he space over which a
wireless network operates is not planar %ut spherical. (herefore, in a multi!level site, network
access is availa%le in rooms a%ove or %elow the access point, without the need for additional
infrastructure.1n a location within a school where network access is required occasionally,
desktop computers fitted with wireless network cards can %e placed on trolleys and moved from
location to location. (hey can also %e located in areas where group work is taking place. As they
are connected to the network, documents and files can %e shared, and access to the 1nternet is
availa%le, enhancing group proIect work.As the range of the wireless network extends outside the
%uilding, students and teachers can use wireless devices to gather and record data outside, e.g.,
as part of a science experiment or individual performance data as part of a #> class.
(,antages of Networking st$ents
Spee-
Networks provide a very rapid method for sharing and transferring files.0ithout a network, files
are shared %y copying them to floppy disks, then carrying or sending the disks from one
computer to another. (his method of transferring files in this manner is very time!consuming.
Cost-
(he network version of most software programs are availa%le at considera%le savings when
compared to %uying individually licensed copies. esides monetary savings, sharing a program
on a network allows for easier upgrading of the program. (he changes have to %e done only
once, on the file server,instead of on all the individual workstations.
Centra'i.e Software Management-
Ane of the greatest %enefits of installing a network at a school is the fact that all of the software
can %e loaded on one computer )the file server*. (his eliminates that need to spend time and
energy installing updates and tracking files on independent computers throughout the %uilding.
Reso$rce S&aring-
+haring resources is another area in which a network exceeds stand!alone computers. ,ost
students cannot afford enough laser printers, fax machines,modems, scanners, and &2!<A,
players for each computer. -owever, if these or similar peripherals are added to a network, they
can %e shared %y many users.
/'e0i%'e (ccess.
+chool networks allow students to access their files from computers throughout the school.
+tudents can %egin an assignment in their classroom, save part of it on a pu%lic access area of the
network, then go to the media center after school to finish their work. +tudents can also work
cooperatively through the network.
Sec$rit1-
Files and programs on a network can %e designated as Bcopy inhi%it,B so that you do not have to
worry a%out illegal copying of programs. Also, passwords can %e esta%lished for specific
directories to restrict access to authoriHed users.
,ain challenges of installing a +chool Network
Costs
Although a network will generally save money over time, the initial costs can %e su%stantial, and
the installation may require the services of a technician.
<equires Administrative (ime.
#roper maintenance of a network requires considera%le time and expertise.,any students have
installed a network, only to find that they did not %udget for the necessary administrative
support.
File +erver ,ay Fail.
Although a file server is no more suscepti%le to failure than any other computer,when the files
server Bgoes down,B the entire network may come to a halt.0hen this happens, the entire school
may lose access to necessary programs and files.
T&e OSI Moe'2 +nerstaning t&e
Se,en 3a1ers of Comp$ter Networks
Intro$ction
(he Apen +ystems 1nterconnection )A+1* model is a reference tool for understanding data
communications %etween any two networked systems. 1t divides the communications processes
into seven layers. >ach layer %oth performs specific functions to support the layers a%ove it and
offers services to the layers %elow it. (he three lowest layers focus on passing traffic through the
network to an end system. (he top four layers come into play in the end system to complete the
process.(his white paper will provide you with an understanding of each of the seven layers,
including their functions and their relationships to each other. (his will provide you with an
overview of the network process, which can then act as a framework for understanding the
details of computer networking.+ince the discussion of networking often includes talk of 3extra
layers4, this paper will address these unofficiall layers as wellFinally, this paper will draw
comparisons %etween the theoretical A+1 model and the functional (&#"1# model.Although
(&#"1# has %een used for network communications %efore the adoption of the A+1 model, it
supports the same functions and features in a differently layered arrangement.
(n O,er,iew of t&e OSI Moe'
A networking model offers a generic means to separate computer networking functions into
multiple layers.>ach of these layers relies on the layers %elow it to provide supporting
capa%ilities and performs support to the layers a%ove it. +uch a model of layered functionality is
also called a 3protocol stack4 or 3protocol suite4.#rotocols, or rules, can do their work in either
hardware or software or, as with most protocol stacks, in a com%ination of the two. (he nature of
these stacks is that the lower layers do their work in hardware or firmware )software that runs on
specific hardware chips* while the higher layers work in software.(he Apen +ystem
1nterconnection model is a seven!layer structure that specifies the requirements for
communications %etween two computers. (he 1+A )1nternational ArganiHation for
+tandardiHation* standard D:J.!1 defined this model. (his model allows all network elements to
operate together, no matter who created the protocols and what computer vendor supports them.
(he main %enefits of the A+1 model include the following8
K -elps users understand the %ig picture of networking
K -elps users understand how hardware and software elements function together
K ,akes trou%leshooting easier %y separating networks into managea%le pieces
K 2efines terms that networking professionals can use to compare %asic functional relationships
on different
networks
K -elps users understand new technologies as they are developed
K Aids in interpreting vendor explanations of product functionality
(he physical layer of the A+1 model defines connector and interface specifications, as well as the
medium
)ca%le* requirements. >lectrical, mechanical, functional, and procedural specifications are
provided for sending
a %it stream on a computer network.
Components of t&e p&1sica' 'a1er inc'$e2
K &a%ling system components
K Adapters that connect media to physical interfaces
K &onnector design and pin assignments
K -u%, repeater, and patch panel specifications
K 0ireless system components
K #arallel +&+1 )+mall &omputer +ystem 1nterface*
K Network 1nterface &ard )N1&*
1n a LAN environment, &ategory =e '(# )'nshielded (wisted #air* ca%le is generally used for
the physical layer for individual device connections. Fi%er optic ca%ling is often used for the
physical layer in a vertical or riser %ack%one link. (he 1>>>, >1A"(1A, AN+1, and other similar
standards %odies developed standards for this layer.
Note8 (he #hysical Layer of the A+1 model is only part of a LAN )Local Area Network*.
Layer 7 of the A+1 model provides the following functions8
K Allows a device to access the network to send and receive messages
K Affers a physical address so a device6s data can %e sent on the network
K 0orks with a device6s networking software when sending and receiving messages
K #rovides error!detection capa%ility
&ommon networking components that function at layer 7 include8
K Network interface cards
K >thernet and (oken <ing switches
K ridges
N1&s have a layer 7 or ,A& address. A switch uses this address to filter and forward traffic,
helping relieve
congestion and collisions on a network segment.
ridges and switches function in a similar fashionG however, %ridging is normally a software
program on a &#',while switches use Application!+pecific 1ntegrated &ircuits )A+1&s* to
perform the task in dedicated hardware,which is much faster.
Layer ?, the network layer of the A+1 model, provides an end!to!end logical addressing system
so that a packet of data can %e routed across several layer 7 networks )>thernet, (oken <ing,
Frame <elay, etc.*. Note that network layer addresses can also %e referred to as logical addresses.
1nitially, software manufacturers, such as Novell, developed proprietary layer ? addressing.
-owever, the networking industry has evolved to the point that it requires a common layer ?
addressing system. (he 1nternet #rotocol )1#* addresses make networks easier to %oth set up and
connect with one another. (he 1nternet uses 1# addressing to provide connectivity to millions of
networks around the world.(o make it easier to manage the network and control the flow of
packets, many organiHations separate their network layer addressing into smaller parts known as
su%nets. <outers use the network or su%net portion of the 1# addressing to route traffic %etween
different networks. >ach router must %e configured specifically for the networks or su%nets that
will %e connected to its interfaces.<outers communicate with one another using routing
protocols, such as <outing 1nformation #rotocol )<1#*and Apen version of +hortest #ath First
)A+#F*, to learn of other networks that are present and to calculate the %est way to reach each
network %ased on a variety of criteria )such as the path with the fewest routers*.<outers and other
networked systems make these routing decisions at the network layer.0hen passing packets
%etween different networks, it may %ecome necessary to adIust their out%ound siHe to one that is
compati%le with the layer 7 protocol that is %eing used. (he network layer accomplishes this via
a process known as fragmentation. A router6s network layer is usually responsi%le for doing the
fragmentation.All reassem%ly of fragmented packets happens at the network layer of the final
destination system.(wo of the additional functions of the network layer are diagnostics and the
reporting of logical variations in normal network operation. 0hile the network layer diagnostics
may %e initiated %y any networked system, the system discovering the variation reports it to the
original sender of the packet that is found to %e outside normal network operation.(he variation
reporting exception is content validation calculations. 1f the calculation done %y the receiving
system does not match the value sent %y the originating system, the receiver discards the related
packet with no report to the sender. <etransmission is left to a higher layer6s protocol.+ome %asic
security functionality can also %e set up %y filtering traffic using layer ? addressing on routers or
other similar devices.
Layer :, the transport layer of the A+1 model, offers end!to!end communication %etween end
devices through a
network. 2epending on the application, the transport layer either offers relia%le, connection!
oriented or connectionless,
%est!effort communications.
+ome of the functions offered %y the transport layer include8
K Application identification
K &lient!side entity identification
K &onfirmation that the entire message arrived intact
K +egmentation of data for network transport
K &ontrol of data flow to prevent memory overruns
K >sta%lishment and maintenance of %oth ends of virtual circuits
K (ransmission!error detection
K <ealignment of segmented data in the correct order on the receiving side
K ,ultiplexing or sharing of multiple sessions over a single physical link
(he most common transport layer protocols are the connection!oriented (&# (ransmission
&ontrol #rotocol
)(&#* and the connectionless '2# 'ser 2atagram #rotocol )'2#*.
Layer =, the session layer, provides various services, including tracking the num%er of %ytes that
each end of the session has acknowledged receiving from the other end of the session. (his
session layer allows applications functioning on devices to esta%lish, manage, and terminate a
dialog through a network. +ession layer functionality includes8
K ;irtual connection %etween application entities
K +ynchroniHation of data flow
K &reation of dialog units
K &onnection parameter negotiations
K #artitioning of services into functional groups
K Acknowledgements of data received during a session
K <etransmission of data if it is not received %y a device
Layer L, the presentation layer, is responsi%le for how an application formats the data to %e sent
out onto the network. (he presentation layer %asically allows an application to read )or
understand* the message.
>xamples of presentation layer functionality include8
K >ncryption and decryption of a message for security
K &ompression and expansion of a message so that it travels efficiently
K 5raphics formatting
K &ontent translation
K +ystem!specific translation
Layer D, the application layer, provides an interface for the end user operating a device connected
to a network.(his layer is what the user sees, in terms of loading an application )such as 0e%
%rowser or e!mail*G that is, this application layer is the data the user views while using these
applications.>xamples of application layer functionality include8
K +upport for file transfers
K A%ility to print on a network
K >lectronic mail
K >lectronic messaging
K rowsing the 0orld 0ide 0e%
Layers ., J, and 1/
0hether a designed to %e a humorous extension or a secret technician code, layers ., J, and 1/
are not officially part of the A+1 model. (hey refer to the non!technical aspects of computer
networking that often interfere with the smooth design and operation of the network.
Layer . is usually considered the 3office politics4 layer. 1n most organiHations, there is at least
one group who is favored, at least temporarily, %y management and receives 3special4 treatment.
0hen it comes to networking,this may mean that this group always has the latest and"or fastest
equipment and highest speed network links.
Layer J is generally referred to as the 3%linders4 layer. (his layer applies to organiHational
managers who have already decided, usually with little or no current information, to dictate a
previously successful network plan.
(hey may say things such as8
31t worked in my last company, so we will use it here.4
3>very%ody says this is the right solution.431 read in an airline magaHine that this was the %est
way to do it so that is what we will do.40hat these managers seem to forget is that they are
paying a highly qualified staff to provide them with useful information. (hese managers %ypass
planning in order to make a quick decision.Layer 1/, the 3user4 layer, is in every organiHation.
ut users are much more than a layer. 0hile they are one of the reasons the network exists, users
can also %e a %ig part of the need for trou%leshooting. (his is especially true when the users have
computers at home and have decided to 3help4 the network administrator or manager %y making
changes to the network without consulting the network staff. >qually challenging is the user who
3didn6t do anything4 when the network segment in his"her immediate vicinity suddenly stopped
working. 1n these cases, the layer 1/ identification coincides with layer 1/ trou%les )and the
3121/(4 la%el some technicians have used*.
T&e Internet 3a1er
K #acket!switching, connectionless service
K >ach packet is independently routed from source
to destination
K #acket ordering issue
K 2efines packet format
K 2efines protocol ! 1nternet #rotocol )1#*
T&e +pper 3a1ers
K (wo end!to!end protocols in (ransport layer8
C (ransmission &ontrol #rotocol )(&#* ! relia%le
connection oriented protocol
C 'ser 2atagram #rotocol )'2#* ! unrelia%le,
connectionless protocol
K (he Application layer initially
C <emote login protocol ! (elnet
C File transfer protocol ! F(#
C >lectronic mail ! +,(#
C 2omain name service C
2N+
#ost*to*network 'a1er protoco's
(he -ost!to!network layer is the lowest layer of the (&#"1# reference model. 1t com%ines the
link layer and the physical layer of the 1+A"A+1 model. At this layer, data is transferred %etween
adIacent network nodes in a 0AN or %etween nodes on the same LAN. (he host!to!network
layer provides the methods to transfer data %etween network entities. 1t also provides error
detection and correction procedures, since the errors might come from the physical transfer. (he
host!to!network layer is responsi%le for physically transmitting the %it stream and reconstructing
the 3framed4 data from a received %it stream for the higher layers.
(he data transfer at this layer is normally not end!to!end transfer. 1t is in fact a data transfer from
one node to another, where 3the other node4 might %e the destination node or a node on the path
to the destination.