(Ms Word Version attached for better viewing and printing)

Networks / Data Communications

NTC 360

Lecture Week Three

INTRODUCTION

Now that we have covered most of the media types available for our networks Protocols that run over the top of the media and the standards they follow, the final topic here is choosing a design of a network by choosing media types and infrastructure that will best suit the needs of the business.

INFRASTRUCTURE

When to use hubs verses routers and switches is not always a clear-cut decision. In many cases the decision is made due to cost not necessarily best choice. For most office environments a hierarchical design with 10/100/1000 Mbps Ethernet hub/repeaters are adequate. However, there are times when high-speed transfer is worth the cost of installing switches. A good example may be in the motion picture industry where large full motion video clips are developed and edited on computer screens. Due to the size of these kinds of files and the critical nature of time in the industry full duplex switching in the LAN environment may be called for along with Gigabit Ethernet or higher = ).

There is an old saying out there that was “route when you can, bridge if you must” This saying still holds true through time. However, switching has for the most part replaced most bridges today. Routing is the best means of isolating and reducing network traffic, which in the end equates to better network performance, however the speeds of the new switches can be just as effective these days. There are some cases where specific non-routable protocols must be used due to application specific requirement and bridging must be deployed. The problem with bridging is that any broadcast protocols used will be proliferated throughout the network causing access traffic where it does not need to be. Bridges have come along way over the past twenty years which have improved filtering and protocol blocking, however, is still not as effective as routing.

Now let’s talk about routing. What does this mean, routable, not routable? Why do I care? Imagine all of the data in the world going to all of the computers in the world. None of the data would get through because of all of the traffic and anyone would be able to see anyone else’s data. To prevent this we have devices called a Routers Bridges and switches.

The Router acts as a traffic cop. It looks at the address of where the data needs to go. It looks in its little guidebook (the routing table) and sends the data on its way. It does not let data pass unless the destination is on the other side of the Router. This keeps network traffic down to a minimum. A Router can be a hardware device or software. Most operating systems today have routing software built in. If your computer has two or more NICs, the computer can act as a Router.

At the network layer as we said, each host gets an IP address. This address is obtained by the network layer of the sending machine and made part of the header information of the network layer packet. The data is then sent along the wire. Each of the hosts on that wire sees the data; if the packet does not have its address, the host ignores the data. If your network has multiple segments there is usually a Router connecting the two segments together. Finally the packet of data gets to a Router. The Router reads the address and looks in its routing table to see if it knows which path to send the data on.

This is like the clerk in the mailroom. He/she reads the information on the envelope and places the package in the proper mailbox. If the clerk does not have the recipient on the list he/she will make some inquires; do you know this person? What department does (s)he work in? Once the determination is made, the clerk will enter that information into his/her address book for the future. This address book is the routing table. That’s great, anytime mail is sent to this new person the mail clerk knows what to do with it.

Oh great, there is another problem! There are two mail clerks and each have their own address books. The other mail clerk just received mail for this new person but the clerk does not have the new person in the address book. Can you think of a solution? How about making a deal with the other mail clerk. Anytime you find out about someone new you let me know and I will do the same.

Routers can be configured to do the same thing. This is called dynamic routing. There are also standards for how this is done. Routing tables in the routers are configured differently depending on the protocol. Many of the protocols are routable.

Many times there are multiple routes to the destination. In the routing table there is a parameter that gives the cost of each path. The term cost is not in dollars and cents, it is in value of the link. Part of the header information in an IP packet is a value called, time to live. This value is used to make sure a data transfer is not floating around the system forever. Each time the message gets to a Router this number is decreased by one. When the number gets to “0” the message is discarded. Let’s say a message is sent, it gets to a Router the Router sends it to another Router, and that Router forwards it again. One of the routers has an invalid entry and the message just keeps getting sent from Router to Router and never gets to the destination. If you get too many of these floating around, the network will be bogged down with traffic. That is the reason for the time to live parameter.

Bridge

A bridge’s function can be to convert data from one topology to another called translation such as Token Ring to Ethernet or to connect multiple segments of a network with non-routable protocols to communicate. Lastly a bridge can translate protocols also. For instance; let’s say your company had a segment running IPX protocol and another segment with IP. The bridge can connect the two together encapsulating TCP/IP in IPX and vice versa if need be so that the two systems can communicate.

Today most bridges also act similarly to routers, meaning that it would only convert and send data from one topology to the other if the destination address was on the other topology. Not so long ago these were two different functions and required two separate pieces of hardware. Most likely the way it is done today is that the company will have a commercial Router which is modular. The Router will have multiple NICs. There will be one or more NICs per topology.

Here is a brief bit of ancient history for you. About 10 years ago or so there used to be a device called a brouter. You guessed it. It was a Router bridge combination. It sold for about $6000. Today I have a bridge, Router, and firewall device that I purchased for $50.

Bridges work at the data link layer. The bridge will have a connection to the physical layer of one topology. It will dissect the data link layer encapsulation and recreate a new data link layer packet for the new topology and send it on its way along the new topology. The Bridge is an interpreter. Just like at the United Nations building. The representative from Spain is talking in Spanish to the representative for France who does not speak Spanish. There is a person in the middle who speaks both Spanish and French. The interpreter is the bridge. The interpreter in the middle listens in Spanish and speaks in French. When the French representative replies the interpreter listens in French and speaks in Spanish.

Hubs
Hubs connect all computer LAN connections into a single concentrator and can be thought of as multi-port repeaters. Hubs allow organizations to take advantage of central wiring and extend the perimeter of their network by stretching bulky cable. Hubs do not modify or filter the packet (work at the physical OSI layer.) Because they are configured in a star pattern, troubleshooting became much easier than older wiring configurations. A disadvantage of the hub is that all devices connected to a hub in essence share the bandwidth to the hub, so congestion can become a problem.

Switching
A switch can be used in place of a router or bridge it or can incorporate these functions it selects a path or circuit for sending data packets to its next destination. In general, a switch is a simpler and faster mechanism than a router, which requires knowledge about the network and how to determine the route. The switch has a connection to each host. When a packet gets to the switch, the switch reads the MAC Frame information to see where the data is destined. The switch then creates a connection connecting the two hosts together. This is the same concept as your telephone.
A vital difference between a hub and a switch is that all the nodes connected to a hub share the bandwidth among themselves, while a device connected to a switch port has the full bandwidth all to itself. For example, if 10 nodes are communicating using a hub on a 100 Mbps network, then each node may only get a portion of the 100 Mbps if other nodes on the hub want to communicate as well. But with a switch, each node could possibly communicate at the full 100 Mbps.
A switch is usually associated with layer 2, the Data-Link Layer. However, some newer switches also perform the routing functions of layer 3, Layer 3 switches are also sometimes called IP switches.
On larger networks, the trip from one switch point to another in the network is called a hop. The time a switch takes to figure out where to forward a data unit is called its latency. These terms are the same for bridges and routers.
The price paid for having the flexibility that switches provide in a network is this latency. On the other hand, switching allows a network to maintain full-duplex Ethernet (remember full duplex (). Before switching, Ethernet was half-duplex, which means that only one device on the network can transmit at any given time. In a fully switched network, nodes only communicate with the switch, never directly with each other. Below is a picture of a switched environment.

We are finally done with the data link layer, Honest!

Gateways

Gateways were developed to fulfill the specific need to join disparate LANs or applications. Gateways can be thought of as the translator needed to decipher between an English-speaking person and French-speaking person. They can translate between different applications, protocols or LAN technologies and they operate at all levels of the OSI model except the physical layer. Some examples include: ➢ SAP to PeopleSoft gateways, ➢ TCP/IP to Novell IPX gateways, and ➢ Token Ring to Ethernet gateways.

Putting it all together

All Information Technology (IT) people are required to make the above technology work flawlessly and in seconds . How do we do this? This is the art of IT. Designing systems, networks, databases and everything else that performs all of the functions needed to keep business running at light speed (literally). Let’s start with communication networks (or Comm networks). Comm networks can look much like a spider web but not quite as pretty . It is the means by which information or data is exchanged from one location to another. In most cases this information is transmitted via copper wire, satellite, fiber optic cable, radio frequency or cellular transmissions. I know this seems like a lot of choices, but each one has its own set of benefits and detriments and of course each has different costs.

Let’s build an example of systems and networks that clarify all of this techy mumbo jumbo. Imagine for a moment that you are starting your own business, perhaps making childrens science kits for schools. First you came up with the idea, then began to formalize a business plan to check out the feasibility of this endeavor. So you are sitting at home, in your den, on your computer, researching your new concept on the Internet. What do we have here; you got it, a single user system.

Now, that you have your venture capital to start your new business, you find a nice small, low cost, office that you can rent. You will need a computer to track all of your expenses, developing marketing material, track inventory and such. You are off to the races! Your business is taking off! 50 new orders but it is too much for one person to handle the phones, the books, assembly of the kits and track inventory. So, time to hire a staff.

Three people have been added, a person to assemble the science kits and package them and an office manager. After fighting over the one computer in the office, you decide it is time to add another computer so nobody has to sit and wait for his or her turn to use the computer. Because the three of you do access the same spreadsheets and databases it is decided that the best way to be sure the information is correct, you need to have one spot for the data and everyone could access the same data, that way everyone would be working with the same information. In order to do this; it was recommended by the local Computer vendor to create a network so that both machines have access to the same data. So a two-user network was developed.

The network was designed as a peer-to-peer network since there are only two computers so a server was not necessary. Because the two computers were basically sitting next to each other it was decided to connect them together using RJ62 or coax cabling, for an Ethernet connection. One of your computers also was configured with a modem and phone line to access the Internet so that your customers can email you, and provide faster communication to your suppliers. Life was great for about six months.

Word spread fast about the great job you were doing for your customers and you won a bid to provide the entire district of schools K thru 12 all of their science kits! This meant you had to expand, quickly, to be able to fill this new demand! They even gave you 50% of the money upfront!

Well, your three-man office was not big enough to store and assemble of the kits, so you went back to the real estate agent and found a nice small warehouse to keep your inventory and assemble the kits. This warehouse was just three blocks down the street so you could keep your current office as an administrative office and move all of the production to the warehouse. Now came a dilemma how to provide your assembly and packaging staff with the data from the computer when they are now going to be down the road. So you call your local Computer vendor rep back to help you with the problem.

Since the growth of the business your computer started running very slowly so you mentioned this to the rep as well. After talking to you for a couple of hours to understand more about your goals for the business over the next couple of years, the rep decided to redesign your system so that it can easily be expanded in the future and take care of your current expansion. You thought this was great until you saw the price tag, are you sure this is the best way, you asked? If you want to keep your business running smoothly this is the only way said the rep. So you dug out your checkbook.

The next thing you know a Windows NT file server was installed with lots of memory, CPU power, and disk space to handle all of the new inventory items, payroll, the books, marketing information and everything else you need to keep your business growing. The old coax cable was no longer as effective because you have added five more computers to the office area for all of the new staff there and wanted to access the data with more speed than you have in the past so RJ45 cabling was installed in the administration office connected to a 12 port 10/100 Mb stackable Ethernet switch. Wow was it fast, now those big expense sheets came up just as you clicked the mouse instead of going to get a cup of coffee waiting for the information to come up, at the same time your office manager was creating pay checks and filing clerk was entering time sheets. You got it, now you are the proud owner of a bonafide local area network (LAN).

You asked the rep, “what about the warehouse. How will they access the inventory program?” The rep proceeded to explain the next step. The local exchange company (LEC), Lucent, has fiber optic cabling laid under the streets through out the city and for a minimal monthly fee will allow you to use this cable and connected the PCs at the warehouse to the administration office LAN. These PCs will look like they are just in the next room. The next thing you know Lucent is installing routers in the admin and warehouse locations to connect your LAN to their metropolitan area network (MAN). Before you new it, the staff sent you an email from the warehouse saying that the needed more pine cones for the kits! All you could think of at the moment was, “isn’t technology great!” Information was flying around at the speed of light and business was growing at the same rate.

Two years later with a nicely padded checkbook it was time to expand again! Not only did you win the bid for the state of Michigan you also won the bid for Ohio! The only requirement was that the kits be made in the state they were sold, so it was time to find a warehouse in Ohio. Site found! Now, how do we organize this, you asked yourself, and back to the rep you went. “No problem!” said the rep, we will install a wide area network (WAN) so that the computers in Ohio can access the file server in Michigan so that the same inventory system, payroll, marketing etc. information can be accessed from there! “Too cool,” you said.

Lucent knocked on the door and said we are here to hook you up to our frame relay service to your site in Michigan. “Frame Relay?” you asked, “what will this do?” This will allow you to share phone lines with other customers to sent your computer data to and from Michigan at a low cost since you are sharing the phone line with other people.” “Yikes!” you said, “What if they get my data?” While wearing a sheepish grin the Lucent person said don’t worry only you can receive your data, it is specially coded for you.” Somewhat reassured you answered, “Oh, ok.”

Now, you have copied what you designed for access from Ohio into three other states and your pocket book is bursting with all of the new business you have.

Conclusion

Although the above was based on real events the names have been changed to protect the innocent . This is how technology helps business to grow and find efficiencies. With good evaluation of business needs technology can provide the means to achieve success. Computer networks are means of connecting computers and people together providing the communication needed to solve business requirements.

Discussion Questions

1. What infrastructure devices does your company use? Routers, bridges, switches? Why? What manufacturer do they use?

2. What is a virtual Circuit in a packet switched network? What types of VC are available and how are they different? Why is it beneficial to use VCs?

3. What is the difference between a centralized and a distributed control of LANS? What are the advantages and disadvantages of each? Which does your company employ? Why?

4. What is the different between Ethernet and Token ring technologies. What does your company use?

5. What are some methods to resolve network congestion?

6. Based on the Hamblen and Lawson (2008) article, what is the major innovation that this new switch offers? How does this convergence allow data centers to be virtual?

7. Based on the Beliles and Twinam (2008) article, what role does an Ethernet layer 2 switch play in virtual local-area networks (VLAN)? How can an Ethernet layer 2 switch be used to implement physical and data security policies? How does protecting a VLAN switch in a secure physical location prevent tampering?

References

Beyda, William J. (2000). Data Communications: From Basics to Broadband. Needham Heights, MA: Pearson Custom Publishing. ISBN 0-536-60540-8.

Stallings, W. (1997). Local and Metropolitan Area Networks (5th ed.). Upper Saddle River, NJ: Prentice Hall.

www.whatis.com

www.howstuffworks.com