You may also want to take a look at the v$event_name view, which lists all the wait events available for a specific Oracle release. It is also a great quick reference for the parameter values and the event classifications.

I like to start my 3-circle analysis by focusing on the Oracle system. Central to the Oracle response-time analysis is a good understanding of what Oracle processes are waiting for, such as a block, a buffer, a latch, or an enqueue. From an all-session or a system-level perspective, the central wait event view is v$system_event. When looking at v$system_event over an interval of time, you get a broad understanding of the Oracle wait situation.

Figure 2-8 is a 30-second interval snapshot based on v$system_event. The report swpctx.sql is part of OraPub's System Monitor (OSM) tool kit, which is a free and easy-to-use performance diagnostic tool kit available from OraPub's web site. Notice there are no session-level references. Session-level information is available in v$session_event and v$session_wait. The top wait event is clearly db file scattered read. In fact, sessions are spending 53% of their Oracle wait time on this single wait event.