The goal of troubleshooting is to identify and diagnose problems, and then determine and execute the necessary repair. There are many troubleshooting methods, and they vary by the type of problem that you are trying to resolve. Implementing a repeatable troubleshooting process is important so that you can quickly resolve problems. A common troubleshooting method is as follows:
- Clearly define the problem. Obtain an accurate description of the problem by verifying the reported problem, including when it was noticed, and how it can be reproduced. The more clearly defined the problem statement, the easier it will be to complete the remaining steps.
- Gather information related to the problem. Turn up logging, review event logs, and try to reproduce the problem. In many cases, you will have an idea about what the problem is after completing your problem statement. However, be sure to gather as much accurate information as possible, without coming to conclusions and making premature decisions about the nature of the problem.
- List the potential cause of the problem. With the problem statement and gathered data, you can enumerate all potential problem causes. This step requires a little creativity to come up with all of the components related to the issue. It is important to be thorough, and to explore all possible options. Search from product knowledge base, product support documentation, and the Internet for information about possible causes.
- Rank the possible causes by probability, and define their solutions. Create a list of either solutions or additional troubleshooting that is required to address each potential cause. Search the product knowledge base, product support documentation, and the Internet for information about possible resolutions.
- Rank solutions by ease of resolution and impact to complete. It would be obvious to try the most likely solutions first, one at a time, until you discover the solution. In some cases, however, the solutions are invasive and require long outages or more resources to complete, in which case you might want to first try the less probable but less invasive solutions.
- Try the most probable and easily implemented resolutions first. Work through the list of solutions, one at a time, until you resolve the issue, or gather additional information that changes the definition of the problem.
- Reduce logging to normal. To reduce server loads, be sure to return all settings back to normal.
- Document resolution and root cause for future reference. Although you may remember details of the solution later, documenting the root cause and the resolution will reduce resolution times in the future.