In addition to the usual request to Telecommunication asking to place twisted pair lines for the 74 drops, I needed to consider the logical layout and design of the concentrator. In order to support 100Base T as well as future growth, I decided to use the Bay 5000 concentrator. Additionally, new fiber optics had to be pulled within the building to support these new subnets. As discussed with Alan Ford, the MS Computer Lab Manager, the transition to twisted pair is a rolling conversion which is well under way.
This project was to determine the reason why the UNIX based DHCP server could not assign an IP address to Windows 95 clients. Using NetXRay, I performed packet capture and decode to determine the source of the DHCP error. This demonstrated that the exchange from the DHCP server to the DHCP client consistently had a bad CRC. It was later determined that the version of the software running on the router was old, and had a know problem with generating CRC errors when redirecting DHCP packets.
This project was to determine why users of the Hamaliton LAN were getting disconnected from the network. It appeared this was some how related to usage. Using the Netscout Probe, I determined that there was a jabbering network interface card. I then proceeded to track down the source of the jabbering by using the Netscout software’s expert visualizer. This graphed in real time the number of errors each device on the network was producing. By correlating this with when the jabbering occurred, I determined the MAC address of the jabbering NIC. The computer was then tracked down through the management of the concentrator. The MAC address was then correlated to a port on the concentrator, which then led directly to the computer that was producing the jabbers.
Federal Hall recently underwent a renovation. This included gutting the existing wiring. In addition to replacing that wiring, Network Services also decided to make two data ports in each room active instead of only one. The number of data drops, which were now doubled, was then augmented by the installation of a new computer lab and computer study room. It was determined that the old plan was not adequate. The existing Bay 3000 concentrator could not support that many connection. The decisions was made to order a Bay 5000 concentrator, which was subsequently built and configured. In addition, a four inch conduit had to be run from the computer lab to the wiring closet one floor up. This allowed Telecommunications to run the necessary twisted pair wiring and place data drops were they were requested. This project was managed right down to the deadline. Difficulties of working with many different departments, as well, as the UPS strike, and politics all had to be over come.
Since the creator of the network, Tom Southall left, there have been many gaps of knowledge. One such gap was the lack of understanding of how the network was designed. It became my task to understand how the fiber backbone was run and which routers served what departments. After becoming a topology expert, I wanted to document this information. I saw the world wide web as the perfect medium. I set up a password protected web server on a Windows NT server to host my pages. Then using Java Script and HTML, I created wire frame images. When the ports of these images were highlighted, a text box would produce information such as the IP address of the port and what hardware was on the other side. When you clicked on the port it would take you to a page that shows a wire frame image of the associated hardware as well as some important notes about the location and function of the device.