This talk from Joyent engineer
at the 2013 Southern California Linux Expo (SCaLE) covers Linux performance analysis and tools--specifically tools for observability and methodologies to use them. Gregg presents some analysis tools for Linux like DTrace and discusses use cases in which they may come in handy. He also presents a summary of the utilization saturation and errors (USE) method (which you can
read more about here
), and explains some areas of a Linux system that some developers may not know to check for performance problems, and which are at least worthwhile to know about.
Watch Gregg's presentation and follow along with his slides below.