Trapped in time. . . surrounded by evil. . . low on gas.
Sometimes, you don’t have time to understand the fine details: users and management are breathing down the back of your neck, and you absolutely must fix a problem immediately. Perhaps you’re the backup for a coworker who went out of town the day before ten users storm your shared office and demand that performance improve; perhaps you’re a contracted systems administrator who needs a quick checklist for things that might possibly be wrong with a system; perhaps you’ve read this book cover-to-cover and are looking for a concise reference so you can throw away that legal pad full of notes.
In this chapter, we’ll summarize procedures for approaching two common but not simple questions: where do I look in order to identify and solve common performance problems, and what tuning, in general, do I need to do on a fresh system in order to ensure performance is maximized?
Before we get into this material, however, a word of caution.
Warning! This chapter contains words of great peril! It illustrates solutions and approaches without explanation. As a result, it runs the risk of turning you into a Cargo Cult Performance Analyst (see the introduction in Chapter 2). Be careful!
These are the five most basic questions to ask when you see a system that is performing poorly. Note that all of the threshold guidelines here assume that the data ...