The first option is to use analogue lines, and in this scenario, we usually have an analogue modem, pretty much like a modem that you would dial up to your ISP with.

The difference between an analogue modem and a dial-up modem is that an analogue modem doesn't dial. On the other side we would have an analogue modem as well, so we have a local client and a remote client and between the two, we have a telephone company (Telkom SA) supplied piece of copper cabling.

How the internal service of this supplied copper cabling works is again out of the scope of this course but really what this means is that we can now connect a local office to a remote office.

There are disadvantages to analogue modems,

Some advantages could be that they are cheap - they're much cheaper than any other communication mechanisms with the exception of possibly using wireless, so they still in fairly high demand locally in South Africa and there are still quite a number of installations of analogue circuits.

The next means of connecting a remote and a local server together, that we are going to discuss is by a digital wire, and again this would be supplied by your local telephone company.

A digital wire can run much faster because it 's a digital signal that 's being transmitted which means there is no conversion between an analogue signal and a digital.

Think of a modem, when you dial up to the Internet you hear the buzzing, crackling and wheezing of the modem while it 's converting your digital bits coming out of the PC into analogue sound and sending them across a piece of wire - analogue frequencies.

In digital mode, with a digital line, there is no conversion happening, which means it 's much faster. Currently the flow of digital line you can get in South Africa is 32 Kbps.

What happens on both the local and remote side is that there is a Network Terminating Unit, what they call an NTU.

An NTU is equivalent to a modem. An NTU 's job is to provide an interface that we can connect our devises.

In this scenario, we are transmitting digital data down this line rather than analogue data. The disadvantage with digital is that it's expensive. In South Africa, it 's significantly more expensive, in the order of ten times as expensive to install a digital line than it is to install an analogue line. In South Africa we refer to digital lines as DIGINET.

Overseas they run T1 and E1 lines, where T1 is 1.5Mbps and E1 is 3.4Mbps and if you compare that to our current offering of DIGINET in South Africa, it is actually the bottom of the range with a speed of 32Kbps.

So overseas you can buy T1 and E1 line which are significantly faster than anything yet available in South Africa. Yet although you can buy fast lines they are significantly more expensive.

The latest technology is ADSL, which is Asynchronous Digital Scriber Line this is a digital line, so we get the digital connection between the two but the Asynchronous Transfer means that the download speed can happen anywhere between 8 and 15Mbps.

The upload speed is restricted to between approx 256k and 2Mbps (this will depend on your Telecom provider) but it is Asynchronous Transfer, which means it doesn't send/receive these things at the same speed. ADSL is only now being rolled out in South Africa.

The next type of WAN that we get is one that uses dial-up lines.

This is a common way of connecting to the Internet and in this mechanism we have a PC connected to a modem, which can dial-up from time to time make a connection to a modem at the ISP which is in turn connected to a LAN. By dialing up, we are extending the LAN.

The other dial-up that is on offer is a digital line: ISDN (Integrated Services Digital Network). ISDN offers a dial-up digital line instead of a dial-up analogue line. It uses a technology where it offers three lines at the same time.

1.a B channel

2.another B channel

3.a D channel.

The D channel is the data channel - it 's the channel used to communicate between the ISDN equipment and it 's not available for us to communicate on but runs at 16Kbps.

Each B channel can run at 64 Kbps.

So in fact, with ISDN we've got a maximum of 128Kbps of bandwidth when we use both B channels. The advantage of ISDN for example is that it can either use both B channels and get 128Kbps or we can use a single B channel (64Kbps) reserving the remaining B channel for telephone or fax communication, while simultaneously being attached to the network.

The two B channels and a D channel offer us more flexibility and the dial-up is a digital rather than analogue.

The advantage of ISDN apart from the fact that you've got higher speed is also the connection time. The time to connect with an ISDN service is often less than 4 seconds. In other words, from the time that you dial to your ISP, until the time that you are actually connected and can start surfing the Web is less than 4 seconds.

In my set up, it takes close to 1 second to connect as opposed to an analogue modem which could take up to 30 seconds to connect.

Wi-Fi is technology for connecting clients remotely and is the fastest growing technology offered by all the major players in this market. Wi-Fi or 802.11g is wireless connectivity offering to connect between 11 and 56Mbps and even higher. The advantage of wireless technology is it 's lack of the need of physical wire/copper or Fiber to connect to the client.

In the past we've had a modem in some form, connected by a physical piece of wire to another modem, the wire is now gone and we will have a dish or an antenna talking to another antenna.

Another means of connecting is Asynchronous Transfer Mode (ATM) and this certainly offers the fastest Wide Area Connection available today. Speeds start at 155Mbps and running to approx 622Mbps, although with recent technology, we can expect speeds to be significantly higher.

If you take that and you compare that to our LAN running at 1000Mbps, 622Mbps is only running 40% slower than what our Ethernet is running.

So clearly this is where WAN 's are moving. Higher bandwidth is demanded and this can only be delivered by these types of technology at speeds high enough to satisfy the need for bandwidth. In South Africa the Telecoms company uses a combination of microwave and ATM technology to deliver service between Johannesburg, Durban and Cape Town, the three main centers. This technology can carry voice, video and data at great enough speeds to ensure some quality of service.

In the old days the LAN comprised mostly of devices called hubs or a concentrator in other words.

A hub or a concentrator was a way of concentrating network connections in a single point. We said that hub 's ran at 10Mbps and essentially if you put 10 machines into a wire that was running a 10Mbps you would see that every machine could probably only transmit at 1Mbps even if they were transmitting at their maximum.

This statement is not strictly true of course, because Ethernet is CSMA/CD, so there would be a back-off process and two machines would communicate with one another, ultimately using up their 10Mbps standard.

Hubs were shared, they were slow, they were not optimal, primarily because you had a certain number of devices that you plugged in and the performance of Ethernet would degrade to such an extent that it was preferable not to even work on the network. That was in the bad old days!

Hubs then gave way to switches, the difference between a hub and a switch is that it when workstations started communicating with one another, they would essentially form a direct connection and even though other devices were connected these two workstations would talk directly to one another.

They would create a virtual connection between the two devices that were communicating with each other.

Once the conversation was complete that connection would be broken and then if a machine wanted to talk with a different workstation it would again create a virtual connection.

So you can see that at different times, different workstations could communicate with one another without interfering with each other's traffic, because there 's a virtual connection being established.

This set up really became a point connection, it was switched. What would happen is as soon as the packets arrived, they would switch to a correct port and they would leave on the correct port without interfering with anybody else 's traffic.

As opposed to hubs, switches were much faster, there was less contention but they were also much more expensive.

Now if you relate this to our TCP/IP model you will notice that a hub really operated only at the physical layer. It had no intelligence to know which port a particular PC was on. It had no intelligence to understand how to move packets between port A and port G. Switches on the other hand are able to switch packages between one port and another based on who is connected to that port.

Switches actually offer a switching service where it builds up a table similar to our ARP table, with port number and MAC address.

So switches are much more intelligent, they can communicate at the MAC layer, they are faster, they are able to switch packets, there is less contention and as a result one gets a much better through-put.

On the downside a switch is more expensive than a hub.

In our networks today there are very few hubs left as most organizations use switches. They are available as 10Mbps, 100Mbps or gigabit switches and you pay accordingly.

A Switch creates a virtual bridge between point A and point G and the packets flowing across this bridge are only destined for point G.

In the hub scenario, the packets were delivered to all workstations on the network.

Clearly that could be a problem in terms of contention, in terms of speed, in terms of efficiency. So because hubs are shared, every time a connection is made, it had to contend with everybody else wanting to make a connection.

With switches, it 's like a bridge, where only one person is able to cross the bridge at a time.

On a network we could have a switch with a whole bunch of workstations attached to this switch. These workstations can happily communicate with one another because they are on the same logical and physical network.

There arises a problem, because a switched network is what we refer to as a flat network, in other words, in order for these machines to communicate they must all reside on the same logical network. If they don't they can't communicate.

An example of the same logical network is: where the address is 192.168.0.X, and each workstation would be a item within that address, such as workstation 1, another might be workstation 15 and yet another might be workstation 212.

They are on the same logical network, and the same physical network and they communicate with no problem.

If we attached a second switch and put all these workstations on the 192.168.0.X network, again as examples assuming that we have a workstation 2, workstation 46, and workstation 89.

Now the two would be on the same logical network and you can see that this is a fairly flat network. As long as they are on the same network they can communicate.

What happens if I changed this and said that the additional workstations were on the 172.16.4.X network? We would now have 2 logically different networks.

If we relate this to our TCP/IP model, remember at the bottom we have the physical layer and that 's Ethernet. One layer up where switching happens we have MAC addresses. Only after that is the 3rd layers where we have network addressing, which in our case is IP.

Notice that a switch doesn't operate at the network layer- it cannot operate at the network layer. Its maximum reach up the network is to the MAC layer. Clearly we've got a problem, because we now want to communicate between one network and another network and that means that switches are inadequate, they can't solve the problem.

Remember that if you look inside and you look at the switching table all it 's got is a MAC address and a port number. All it can tell you is that MAC address "X" is at port number 6 or 7 of 15 or 24 etcetera.

The difference was that gateways were responsible for transferring between one protocol and another protocol.

For example, there is a protocol called SNA, which is used, mostly by banks in their Auto Teller machines because of design elements. It 's very efficient on Wide Area Network. The bank would run TCP/IP internally but they would need to communicate with their Auto Bank Teller machines by SNA and they would need a gateway to convert between the SNA protocol and the TCP/IP protocol.

So gateways are generally referred to as a translation mechanism between one protocol and another protocol. They are still very much in use today but the distinction isn't quite as clear as it used to be.

A router really has a similar job but its job is not to communicate from one protocol to another protocol, its job is to connect from one network to another network.

Let 's look again at our example above where we had two networks, one where the workstations fell into the 192.168.0.X network, and a second switch with workstations that fell into the 172.16.4.X network.

These are completely different networks, both physically and logically and in order to connect these networks we need a router, the router is going to convert between one network and another network.

On the one arm of the router is the 192.168.00.X network and on the other arm of the router is the 172.16.4.X network.

For most companies they would have a LAN, which would be connected to a router connected to a digital line (usually), then connected to an NTU on the side of the ISP connected to yet another router and then connected to ISP Ethernet.

In looking at our TCP/IP stack, on the LAN we would be operating at a MAC layer and on the network we would be operating at layer 3, the network layer.

In our simple scenario we have a switched network, where we have a router to convert between the networks. Routers operate at yet another higher level on the TCP/IP stack, they operate at the IP (Network) layer and so they are able to distinguish between physical networks and logical networks.

The router builds up a table of IP addresses and the port number that the requests for service have been detected on, so if the workstations on the 172.16.4.X network are communicating with one another. They equally communicate with the router and on port 1 of the router we have the IP addresses for all these workstations and we have as well the MAC addresses for all those workstations.

For port 2, we have the IP addresses and the MAC addresses of the workstations that fall into the 192.16.8.00.X network and again, like these working switches, these are dynamic tables so what happens when Joe switches off his PC in the 172.16.4.X network, well, his IP and MAC are eventually aged out from the routing table, on the router.

To review the scenario: we have network A, which is the 172.16.5.x network, and network B, which is the 192.168.0.x network. In order to connect these networks we use a router.

Although we've shown network A and network B on separate physical networks, there is no reason why we couldn't combine these into one physical network. For that we are going to use a switch, we would place the workstations onto the switch and 4 of the workstations we might put on the network 172.16.4.X and 4 of the workstations we might put onto the network 192.16.8.00.X. Now they are on the same physical network.

Now how do we connect between logical network A and logical network B?

Well, we will connect via a router we would put the router into the network on to the switch and its job would be to convert between one logical network and another.

A router serves the job of translating between one network and another.

A Linux box can be used as a router. In fact on every Linux box you have a routing table.

A routing table tells us a number of things such as what IP addresses are attached and what ports are on the router. It also tells us whether the port is up or not.

In order to see a routing table you can type the "netstat -rn" command and that will show you your routing table.

	Note
	The -n option has to do with network translation and we'll talk about network translation shortly.

By using the Linux machine as a router you would theoretically have to have at least 2 networks on the same network card.

On the network interface card you would plug in B, 192.16.8.0.X network and would give it a host address of lets call it ONE. (192.16.8.0.1)

Then plug in network A, which is 172.16.4.X and you give it a host address of ONE (172.16.4.1) as well.

How can that router have 2 host addresses on ONE? Well, because this host address resides on that network, network A and this host address resides on network B, network 192.16.8.0.X.

In fact if we looked at the full IP address of the interface on network A it would be 172.16.4.1 and if we look at the full IP address on network B it would be 192.168.0.1.

Linux is quite clever because what it allows you to do is to plumb the interface and plumbing the interface is really a way of attaching multiple IP addresses to the same physical network card.

In order to do that type in the following command:

ifconfig eth0 192.168.0.1 netmask 255.255.255.0
                    

and that would give you your first interface on network B an IP address.

Then type in:

ifconfig eth0:0 
                    

and that would be the first logical interface on the same physical network 172.16.4.1 netmask 255.255.255.0.

This would give you a single network card connected to your switch, on the one side would be network 172.16.4.1 and on the other side it would be 192.16.8.0.1.

So this workstation would send its packet to the router and the router would act essentially as a go-between sending the packet to the client on the network.

Similarly when the packet returns or a reply was sent, the workstation on network A would send it back to the router and the router's responsibility would be to send it on to the correct destination.

So in this the router is acting as the go-between between the two networks. It 's essentially routing packets.

Our example consists of a very simple network but if we were to take a more complex example, you would really see the effect of routing on a network.

For this example, I'm going to draw a typical scenario of a small business connecting to the Internet. Equipment wise we have a switch and attached to that switch is all the workstations in that small business.

The small business is called ACME Widget Manufacturing Company and they manufacture ACME widgets. The client whether Microsoft, Linux or other, connects to a switch. They have a server and the client applications would be requesting services from the server. They have a router to the NTU (Network Terminating Unit) which attach's via Wide Area Link to yet another NTU which attach's to a router and to the ISP, which would in turn attach to 3 or 4 other routers which themselves might attach to NTU.

The ISP, in this case, is a nice stable one, lots of redundancies, so they have a link that goes to New York, they have second link that goes to London, and they have another link that goes to their Johannesburg office and another link that goes to Durban.

If Fred decides to get http://www.google.com (lets not worry for the moment how that translates), that translates to an IP address at 207.46.31.19 for example.

Fred lives on the network, 196.6.14.X and he happens to be host 32.

So what happens is, he says, I need to go to this address, how do I go?

Well, the address doesn't happen to be anywhere on my network so I'll go to my router and this process of going to the router is going to see a default gateway and every host on the network should have a default gateway.

If the network does not know where to send a packet, it will be forwarded to a the default gateway.

In our example with Fred, the packets leave his workstation and route to the switch (acts at layer 2), but because the switch doesn't know anything about this IP address it then switches the packets directly through the router.

The router in turn cannot find the relevant IP address but my default gateway says the ISP, well, in fact, my default gateway is the ISP address. Now here we have a network, 196.6.14 - this is also a network. And this network might be 10.0.0.2 - so this router says, well, if you don't know where to send this packet, send it to 10.0.0.2, which is that port on the router. The router gets the packet and says, OK, I don't know where to send that but what I'm going to do is, I've got my default gateway set up so that if I don't know what to do with the packet, I'm going to send it via New York. And so Fred 's packet goes from his workstation to the router connecting him to his company, across this Wide Area Network to a router within the ISP which in turn has its own default gateway saying if I don't know where to send this packet, I must send it via New York. And so it sends the packet out and off it goes.

At each point along the way, the router records this transaction, so when Google responds, the packet returns. It could essentially return via London but it is destined for the ISP 's router and when it is received by an ISP router, ISP says, OK, I know where to send that, I must send it across the Wide Area link, I must send it to the router, this router gets the packet and says I must send it back to Fred 's workstation. And so the process of sending packets around the Internet is really a process of routers actually knowing which route to take. If I asked you to travel from point A to point B, you would probably pull out a map, you would look at the directions on the map and you would choose a route and you would follow that route. Perhaps going to locate location B, you would follow one route and returning from location B you would follow another - is that feasible? Of course it is! At the end of the day, all it 's requiring is that I've told you to get to point B " you started at A, selected a route and off you've gone. If I say, which route did you select"

You might say, well, the route with the least number of traffic lights on it. That 's my default route, that 's my default means of getting from A to B and that, in a sense, would be default gateway. SO the process of talking between networks, you can see we've got at least 3 networks we're talking and possibly even more. Here's one, the 196.6.14 network, here 's the 2nd one, the 10.0.0 network. In this 10.0.0 network we only have 2 hosts - 1 and 2. In fact there 's no other hosts on that network, only 2 hosts. The 3rd network we have might be that and there might be a 4th and a 5th and a 6th. How many networks, that doesn't really matter as long as our packet knows how to get from A to B and how does it know that? Because the router knows how to route that packet through the network.