It's Not Always User Error
When users experience problems on a network, the root cause of those problems will almost always boil down to three possibilities: user error, software problems (including operating systems), and physical connectivity problems. Although it is a convenient way to dismiss a problem outright and return to surfing of the Web (okay, so not all network administrators sit and surf the Web all day), not all network glitches can be blamed on user error. Software issues on client machines and servers alike can cause access and connectivity problems on the network. A missing or corrupt driver for a printer or network interface card can cause printing or connectivity failure. Physical connectivity problems such as shorted-out cables, an unplugged hub, or a router with a bad Ethernet interface can also cause problems on the network, and sometimes physical connectivity snafus can be some of the hardest to track down or troubleshoot.
Documentation that the users can actually read and refer to as they use the network can also greatly cut down on user error. A brief and concise manual that explains basic network logon and resource access can be a real help to your users. This also gives you an opportunity to establish written policies for network use; for example, you might establish rules such as no software downloads from the Internet and no floppy disks or CDs from home. Both of these rules could actually help reduce the risk of virus infection, and no downloads mean that local hard drives remain pretty clean (users can fill up their hard drives with junk pretty fast if they have a high-speed Internet connection and some free time on their hands).
Establishing user policies for your company's network not only provides a set of rules for the network users, but it allows you to assume (although users are not always going to follow all the rules) that network client machines will only be running software that is appropriately licensed and correctly configured and that important user files are stored in the appropriate place so that they can be periodically backed up.
Although everyone is aware of the often overused anecdote that accompanies the word assume when an assumption proves to be wrong, at least assuming that computers have been configured in a particular "standard" way allows you to concentrate on other issues when attempting to troubleshoot network problems.
Most networking operating systems, including NetWare and Windows Server 2003, provide you with ways to control the user environment and protect local machine configurations from being radically changed by the user or users that sit at that computer. An example of an environment that provides a system of policies for protecting local software installations and configurations is Windows Group Policy. We discuss Group Policy in Chapter 9, "Networking with Microsoft Windows Server 2003."
So, let's assume that you have done everything you can to lock down local computer configurations. You are still going to have problems on the network (we are ruling out user error) and these are going to be caused by either software or hardware. Let's take at look at some approaches to troubleshooting software and hardware problems on the network.