Method R Discussion GroupMethod R helps developers, DBAs, and decision-makers optimize Oracle-based software in every phase of the software life cycle. |
||
Discuss Method R products, techniques, and events. Ask new questions or post your own responses. |
I am trying to grasp the concepts laid out in this chapter, but the numbers in my trace files don't seem to work. Here are my questions:
1) First, do I have the correct understanding in the following points: - {c} is total cpu for a call and its recursive progeny - {e} is the elapsed time within a call and its recursive progeny - {e} includes {c}, so the difference between {e} and {c} is accounted for in elapsed time between calls - {wait ela} 2) If this is not true and the difference between {e} and {c} is not accounted for by waits, then what does account for the difference between the two? 3) Now for the question about the trace file. If {c} is accounted for within {e}, then how do I explain circumstances in the trace file where {c} is greater than {e}? Here is a line from my trace file: FETCH #1:c=150000,e=146737,p=0,cr=0,cu=0,mis=0,r=1,dep=0,og=4,tim=2734191821385 Thanks for your assistance. Regards, Jeremy Drobnick
Jeremy,
Your understanding of {c} and {e} are correct. Your third statement in question #1 is partly true, but it's missing the concept of unaccounted-for time, which is also key to answering your question #2. I think the key to your complete understanding lies in Chapter 7, where the unaccounted-for topic is convered in detail. Specifically, the answer to your question #3 is demonstrated in and around Figure 7-8 on page 164. There is another example of the problem you're asking about on page 166, in the example that discusses "FETCH #96". Hope this helps... |
Powered by FogBugz