• Choose a category
• All
• Classic-Fiction

If she has good connectivity to the overall network but can’t ping the Web server, that could point to a different problem altogether. Figure 23-47 shows the OSI seven-layer model graphically.

Figure 23-47 OSI

The only drawback to the OSI seven-layer model, in my view, is that it’s too complex. I like to conceptualize network issues into fewer layers—four to be precise. Let’s take a look.

Mike’s Four-Layer Model

Network problems, by the very nature of the complexity of a network, usually make for more complex problems. Given that, I have created a four-step process that I modestly call “Mike’s Four-Layer Model.” These four things go through my mind every time I have a problem. I think about four distinct categories to help me isolate the symptoms and make the right fix.

Hardware

Hardware is probably the most self-explanatory of the four categories. This covers the many ways data can be moved from one PC to another. Does the system have a good connection? How’s the cabling? This also covers network cards: Are they installed properly and tested? Plus, the Hardware category hits on all of those interesting boxes, such as hubs, switches, and repeaters, among which all of the wires in the network run. If you can see it, it’s under this category.

Protocols

This category covers the protocols, such as TCP/IP or NetBEUI. Is the protocol installed? Is it configured properly? Does any particular system’s configuration prevent it from working with another system?

Network

The network category has two parts: servers and clients. Network operating systems must differentiate systems that act as server from those that do not. If a system is a server, some process must take place to tell it to share resources. Additionally, if a system is intended to share, it must be given a name. This category also includes defining and verifying users and groups; does your system need them? Do the right accounts exist, and are they working properly?

Shared Resources

Once all of the systems, users, and groups are working properly, you need to identify the resources they will share. If a drive or folder is to be shared, the OS must provide a way to identify that drive or folder as available for sharing. The rules for naming shared resources are called naming conventions. A great example would be a system that offers its D:\FRED directory for sharing. This D:\FRED directory needs a network name, such as FRED_FILES. This network name is displayed to all of the devices on the network.

Sharing a resource is only half the battle. Individual systems need to be able to access the shared resources. The network needs a process whereby a PC can look out on the network and see what is available. Having found those available resources, the PC then needs to make them look and act as though they were local resources. A network also needs to control access to resources. A laser printer, for example, might be available for sharing, but only for the accounting department, excluding other departments.

Chapter Review

Questions

1. To provide a computer with a physical and electronic connection to a network, what must be installed?

A. A hub

B. A router

C. A NIC

D. A bridge

2. Which of the following is needed to configure a PnP NIC in a Windows XP system?

A. CMOS

B. Configuration software

C. Device driver

D. DMA

3. How far apart can two PCs that share the same 100BaseT switch be placed?

A. 100 meters

B. 200 meters

C. 330 meters

D. 1000 meters

4. What is the minimum specification of cable types for 100BaseT networks?

A. CAT 2

B. CAT 3

C. CAT 4

D. CAT 5

5. Joe needs to network two computers in his office using an Ethernet peer-to-peer connection. What kind of cable does he need?

A. CAT-5

B. Crossover

C. UTP

D. STP

6. What are the two TIA/EIA standards for connecting an RJ-45 connector to UTP cable?

A. 10BaseT/100BaseT

B. CAT5/CAT5e

C. RG-58/RG-59

D. 568A/568B

7. Steven’s Windows XP system can