|
When a performance bottleneck or problem occurs, system tracing tools can help you identify the source of the problem. When tracing a system, you need to know not only which tools to use, but when and where to use them. Generally, system tracing involves multiple steps: With each step, you delve deeper into the problem to ultimately create a solution or optimize the system or the appropriate application. By taking a system-wide view of the problem first, and then zeroing in on a specific application or series of events, you can more reliably determine the true cause of the performance problem. For example, just because the system slows down after a particular application is started does not necessarily mean the application is the cause of the slowdown. The application might be a catalyst that exposes a problem within another application on the system. A number of Linux tracing tools are available in the open-source community. This chapter discusses several of the most common:
This chapter also discusses requirements that need to be met before you begin tracing a system. Note that given the dynamic nature of Linux, new system tracing tools are likely to be developed, and existing tools, including the ones discussed in this chapter, will continue to be updated. |
|