To install horizontal cabling we have two run two cables, for one voice and the other for data, one must be four pair UTP for voice (100BaseT4) they will be connected to two telecommunications outlets or connectors at each work area, I chose 100BaseT4 (100ohm) because it can be segmented up till 100 metres. The second network media could be a 2 pair Cat 5 STP 150 ohm (100BaseTX) with its appropriate connector. This uses a single Cat 5 pair in each direction, which also minimises EMI from nearby signal pairs (voice) and environmental noise. Also connected to the horizontal cabling will be the patch cord, which allows equipment to be connected to the work area, the total length of this cannot exceed 10 metres or 33 feet. When installing horizontal cabling EIA/TIA requires that all grounding and bonding must conform to EIA/TIA – 607.
Backbone cabling is the cabling that acts as the primary path for traffic that is sourced from and destined for, other networks. As a lot of data will be travelling down this channel I am recommending the Cat 5e UTP, which is a variant of Cat 5, it has the same bandwidth of 100Mhz and bandwidth of 100Mbps. This will connect all the wiring closets together in the building and this cable is fairly new so we don’t need expensive rewiring in the future. All the wiring for the backbone cabling will radiate from a central location which will located in the middle of the building so the cable has to run through the walls to the wiring closet which will be situated in a room or cabinet that holds the cross-connects, and other hardware needed to connect horizontal wiring to backbone wiring. When installing horizontal wiring as the name suggests the cable runs horizontally above ceilings or along the floor. Horizontal wiring is normally installed during the buildings construction, so once completed it is less accessible than backbone wiring.
Patch cords which connect the workstation to the wall jack are limited in length, no longer than 6 metres is recommended.
Other networking equipment to be used are hubs, switches, routers and repeaters. Extending the analogy I mentioned earlier, I have 5 floors with the LAN centrally located on the middle floor. I have extended the network over the 5 floors. Other networking equipment I need is the hub, this will provide full bandwidth to each client. The workstations are connected using UTP, and traffic on the bus is controlled using CSMA protocol and all the stations share the available bandwidth. The hub also provides a number of ports, which a re logically combined. They can also be buffered so if the hub or port is busy packets are held. This should be a great advantage to the network as each port has exclusive access to its bandwidth and utilises existing cabling and other network components.
Routers are also used in extended star topology. Routers are more useful than bridges and switches in this network as it separates networks logically. A TCP/IP router can segment the network based on IP subnets. Routers can also act as a firewall, which is a good security feature. By configuring the router with access lists (which protocols and hosts have access) which also enforces security by restricting access.
Switches are also good devices to use on the network as it decreases the amount of extraneous traffic on each individual network, but as we are already using a router, which does the same thing, this would be a waste of money, so this isn’t necessary. The hub, router and patch panels will be located in the MDF, where the central point of an extended star topology is. Each segment has its own unique address, and a certain number of devices able to connect to the cable and a limit on its length. Bridges can be used to increase the number of attached stations but they may overload during high periods of traffic, and this may be strenuous on the backbone.
Question 3
Extend Q2., above, to produce a training manual for staff engaged in the design and layout of UTP-based Ethernet LAN’s. [25 marks]
Lets have 3 floors with 10 computers on each floor, five meters apart, with the main computer on the second floor.
Equipment
30 computers, 10 on each floor
1 Server on the second floor
100BaseT4 –Voice cable (Horizontal Cabling)
100BaseTX – Data cable (Horizontal Cabling)
Cat 5e UTP (Backbone Cabling)
Patch Cords
Router
Repeater
2 Telecommunications closets
Hub
- Set-up all the computers on the three floors, 10 on each floor and no more than 5 metres apart. Place the Server in a separate room on the second floor.
- Next step is to make a wiring closet on each floor.
-
To make this closet we have to set up a hub on the 2nd floor with the server as well as a repeater and router on the other two floors.
- Next we install the wires
- Get the Cat 5e UTP for the backbone cabling and run it to the Hub and then onto all three floors to the wiring closet
- In the wiring closet there should be another hub where the backbone cabling from the hub connected to the main server on the second floor should connect.
- Install a telecommunications outlet on to the wall and use the horizontal cabling to connect the wiring closet to the backbone cabling, which can be done by using the routers and repeaters.
- Once the telecommunications outlet is fitted on the three floors, you may use the patch cord to connect the computers to the outlet and the network is now ready.
- Now you may install the NOS and run the network.
Question 4
Briefly outline network-programming features available in UNIX and Windows-based products. Your answer should consider basic concepts rather than detailed software/ programming. [10 marks]
I am very biased, and always sing the praises of Unix. I rarely use Windows, and for networking the best windows based application is NT, so I am going to compare the two. I am backing up my biased view as why I would prefer Unix to NT.
If I was to install Unix on a school-networked system, I would be able to achieve perfect system reliability, while it would be relatively immune to student attacks. Only hardware failure or an administrator error can bring the Unix system to a halt. If we used Windows NT, it’s in a more complex environment, as the windows desktop opens access to everything for the insider attack. This will cause a serious drain on management time, as they would have to become PC support staff, and look after the network, if Unix was implemented, administration would only have to be part time.
Security – Both operating systems have security leaks in their system, whilst Unix is open about its security leaks, and publish fixes for their software whenever a security problem is found, Microsoft diminish them as user errors. Microsoft don’t leak out security defaults in it’s operating systems.
Stability – If heavy server systems are in use, in a large environment, some vendors recommend rebooting NT every week, to get rid of the random junk that the system has left over.
Scalability – UNIX systems can be used on simple PCs to the most powerful allowing you to obtain Unix machines for each and every occasion.
Open – Unix has been built on open standards, allowing anyone to access the operating system, Microsoft have recently introduced this, but is in the trial process.
Standard – This is where Microsoft have got one over Unix, all the Microsoft suites are able to communicate with one another, giving you the same feel of the Microsoft environment throughout, in other words inter-operate..
Application Program Interface (API) – Microsoft uses NetBEUI, this isn’t as diverse as Unix’s pipes which is their version of API. In Unix you may invoke, redirect pipes easily writing the right code. Although Microsoft’s may be easier to use, as it is a more graphical user interface but it’s limited.
Integrated NOSs – Windows NT can run client at workstation and use Novell Netware for networking gaining best from each NOSs, this dominates the market.
Question 5
Compare network capabilities offered by Unix, Linux and Windows. Your answer is to make reference to NOS network operations and protocols. [20 marks]
Microsoft – The protocol architecture view of Microsoft Windows products are very similar such as W95, W98 and NT. Numerous utilities are available for availability and performance monitoring of network applications and services. Such as , this uses the Internet Control Message Protocol (ICMP) to determine the availability and responsiveness of network hosts. These utilities sometimes integrate related network monitoring features. Traceroute utilities track the path taken by packets as they travel from source destination. NS Lookup determines the physical IP address (or addresses) associated with an Internet domain name using this online name server lookup service. Microsoft has 3 communication protocols, which are supported at layers 3, and 4 - NetBEUI, IPX, TCP/IP. NetBEUI is the Microsoft implementation of IBM’s Netbios. This is a method of interfacing software with network services, it is connection orientated and ideal for local networking and fast. However it is non routable and unsuitable for large networks.
UNIX - In 1984, the University of California at Berkeley released version 4.2BSD, which included a complete implementation of the TCP/IP networking protocols. Systems based on this and later BSD releases provided a multi-vendor networking capability based on Ethernet networking. Many networking support tools were included such as remote login, file transfer, electronic mail, and other important features.
Unix can be easily implemented on all types of computers, forming a large network. This is an advantage over Microsoft, which requires computers to upgrade their hardware to make efficient use of its software. Unix supports many protocols, such as, NFS - allows you to treat files across a network exactly the same as files on your own disk drives. RPC - allows a program to call 'procedures' on a machine across a network in a similar way that it would call normal functions. XDR - a way of describing how various types of data (e.g. integers, arrays, structures, etc) should be transmitted across the network. IP - defines the datagrams (packets of data) that all data on the Internet is carried within. TCP - provides a means of exchanging a continuous, error free stream of bytes between computers. TCP 'packets' are normally carried across networks within IP datagrams. UDP- slightly more sophisticated than raw IP datagrams, but not much. ICMP- 'network status enquiries and reports', carried by IP.
Linux - is a clone, written entirely from scratch, of the Unix operating system. Linux is a 32-bit multitasking operating system, which is robust and capable enough to be used in organizations ranging from universities to large corporations.
Linux supports many different networking protocols: TCP/IP networking has been present in Linux since its beginnings. IPX/SPX (Internet Packet Exchange/Sequenced Packet Exchange) is a proprietary protocol stack developed by Novell and based on the Xerox Network Systems (XNS) protocol.
Linux provides full Appletalk networking that is the name of Apple's internetworking stack. It allows a peer-to-peer network model, which provides basic functionality such as file and printer sharing. Each machine can simultaneously act as a client and a server, and the software and hardware necessary are included with every Apple computer. The Linux kernel has built-in ISDN capabilities. Isdn4linux controls ISDN PC cards and can emulate a modem with the Hayes command
Question 6
Briefly outline features offered by Java and C, which may be used in network programming. [10 marks]
C was developed in the 1970s, and is a middle-level language. It combines the control structures of high-level languages with the ability to manipulate bits. C gives the complete control over the machine. Therefore C is used in many areas of network programming, such as in UNIX. We can programme sockets in UNIX using the C programming language. A socket is a software device that allows processes to communicate with one another over the Internet, within local area networks or even on single computers. Sockets provide the programmer with a flexible communications interface that allows reliable transfer of large amounts of data between processes.
Java is not like other programming languages, it is designed to work within the World Wide Web. It is platform-independent, this means that it is independent of the kind of computer you are using. Java was built to access the network, it has extensive features for network programming at various levels from connecting via URLs or sockets to accessing remote methods (RMI). Java makes connecting easy by using one of these classes, java.net, java.rmi and java.jdbc. So there are a variety of ways of connecting to the network, from URL connections to resources, to ports and sockets, database connectivity and remote objects. Each is supported by special classes and libraries and are all part of the standard Java language.
Bibliography
These are the websites and books I used to assist me with this coursework.
- www.itep.co.ae/itportal/english/Content/EducationalCenter/InternetConcepts/networking.asp
- Bishop, Judy, ‘java Gently’, Addison Wesley, 1998
- Palmer, MJ, ‘Hands on networking’, International Thomson Publishing Company, 1998
Taranjit Singh Matharu page of