Week 10 Assignment 3: Business Intelligence and Data Warehouses
CIS175008VA016-1154-001 Intro to Networking
Professor Obi Ndubuisi
June 14, 2015
For the company DesignIT, I would use a LAN type of network (for local-area network). This networks several computers, a mail (exchange) server, a database server and a file share server in one building and interconnects with each switch or backbone. I chose this for this small space because it is less than 100 square feet and is all within a single dwelling. This way the employees can all share their resources with one another, an intranet, database and printers.
This network only needs a single modem with wireless ...view middle of the document...
Many ISPs are linked with cable television and landline packages. DesignIT would only need a basic television package with a single receiver in their conference room and a 10-15 line telephone service.
Point-to-point topology would not be useful because each computer is connected directly. Mesh topology requires many cables and can handle a high-volume traffic. If the wireless access point or router fails it would only affect the computers and devices in the particular area, same with switches but not connected by a cable directly from the wall. In DesignIT’s case, hopefully the failure is just in the access point so that at least the employees would have a medium-weak to weak wifi connection. I would suggest a central backbone installed in ceiling above the reception area. CAT5 or CAT6 cables can support up to 100 meters before signal degradation. These connections can be routed form the reception area ceiling coming out like a star and then down the walls or separations and out to a port jack. Each workstation’s PC would connect to their corresponding wall Ethernet jack. In many workstation cases, the cable goes form the wall jack to their phone on their desk and back down to their CPU which usually sits on the florr or a lower-sitting platform.
For a small company, like the last company I worked for (121 computers, 14 servers) or DeignIT, client-server networking would be the best to implement. All the computers are set up like desk workstations and connected to the same network with hubs.
A server or several servers can allow users to share resources located on these servers. This is better than peer-to-peer because the servers are dedicated. If there's a failure troubleshooting the server can be done- singled out and fixed. With peer-to-peer the scability disallows central management on one server or several servers which would increase a burden of managing multiple clients rather than the servers themselves.
Peer-to-peer networks may cost less but the company would have to dedicate IT support for the desktops assigned to employees. The enduser support would need one or two people to manage those individual computers and files/applications are shared on single computers rather than access through the servers. Administration is easier with client-server networks becasue the resources are shared in a central location or locations.
Licenses are cheaper than a whole person's salary yearly. An IT personnel makes $30,000-50,000 yearly while purchasing extra licenses for client-server is $2,000-15,000. A company that small can have 2-4 people in their IT department: a network engineer, a software ingineer and an IT manager- maybe one person for enduser support. All of which could be outsourced and handled by a third party.
With peer-to-peer, because of the software and applications are individualized and not in a central location there would be a need to have personnel for enduser support and management. The worry is spread out rather...