| SNMP is very good at helping you know when faults occur in your network. For example, if an interface on your router is down or malfunctioning, you may still have network connectivity to the router, so you can use SNMP to further discover what the problem may be. Sometimes when something bad happens, however, you may only know that some network or system is unreachable. When this happens, it is good to know how to use a few tools of the trade to help diagnose and resolve the issue. The following points detail the concepts involved in network troubleshooting: The process of troubleshooting is systematic in that one must be methodical to properly solve the problem at hand. It requires skills and knowledge including an understanding of your environment, problem-solving skills, and an ability to accurately communicate with others. As you begin your troubleshooting task, be sure to change one thing at a time as you test to see whether the problem is resolved. Not doing so can possibly obscure the actual step that resolved the problem, making it difficult to accurately and quickly resolve the same problem when it happens again. Documentation is a key step in the process. If you document exactly what the symptoms were, as well as the resolution steps (including any tools that you used in the process), others can learn from this. The next time someone has to solve the same problem, they can do so more quickly and help minimize downtime. Don't assume that the problem is due to some convoluted set of circumstances. Always check the obvious first. I cannot count the times when I was a NOC engineer that the problem ended up being a bad cable or something simple. Keep in mind that things like firewalls and other security systems can hinder your ability to accurately troubleshoot. For example, a firewall may be configured to disallow packets to the host you are troubleshooting. It pays to understand the security architecture in your environment. It is always best to know a few tools well rather than a bunch of tools not so well. Testing after the problem is fixed is sometimes overlooked in the frenzy to get something back to operational status. Do not overlook this step.
The remainder of this appendix focuses on some general-purpose tools used in network troubleshooting. Covering these tools in detail is beyond the scope of this appendix. For an excellent treatment of network troubleshooting, see Network Troubleshooting Tools by Joseph Sloan (O'Reilly). |
