8.12 Troubleshooting the Network
8.12.2
Distinguish whether a problem is attributable  to the operator or the system
In order to troubleshoot the network for operator or system problems; first the issue needs to be identified. Is the problem protocol-based or a network issue? Second, identify what parts of the network are affected and determine if problem exists on the cabling or the workstations.

When identifying the exact issue, begin with a broad view -- say the entire network -- and as the research is conducted, the problem should become more isolated. When dealing with network problems, recreating the problem can provide assistance in learning the events that have occurred during the error. If the problem is complex and is possible to occur in the future, recreating the problem may be beneficial to the troubleshooter for future reference. However, if the problem is simply to replace a NIC or a piece of hardware, recreating the problem may not be necessary.

Isolating the cause of the problem has two benefits. (1) If the problem is isolated to a specific area or number of users, then the rest of the network can continue to be functional. (2) By isolating the issue, it is easier to diagnose the problem between three to five workstations than it is on 500 workstations.

After identifying the problem, correcting the problem may be even more complicated. There may also be more than one way to fix the problem. First, determine the various methods to correct the problem; sometimes a problem can be temporarily patched or a software patch can also provide a temporary fix. There is also a possibility that when the problem is fixed, another problem will arise because of the fix.

Proper documentation (journals, equipment logs) and feedback, such as methods used to contain the problem and additional comments, can be helpful should the problem arise in the future.