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. |
If possible, I would like to as a clarifying question to the presenter concerning the time values in the STAT lines. If my notes are correct, the presenter said that the value of time is the cumulative duration of the row source operation and its descendants. Using the example STAT lines below, the first line has a time of 52 and second line has time of 43. The second line is a descendant of the of the first line. Given the description, the TABLE ACCESS BY INDEX ROWID has a cumulative time of 52 and an individual time of 9. Is my understanding correct?
STAT #5 id=1 cnt=0 pid=0 pos=1 obj=255 op='TABLE ACCESS BY INDEX ROWID HIST_HEAD$ (cr=2 pr=0 pw=0 time=52 us)' STAT #5 id=2 cnt=0 pid=1 pos=1 obj=257 op='INDEX RANGE SCAN I_HH_OBJ#_INTCOL# (cr=2 pr=0 pw=0 time=43 us)'
Your understanding is correct.
Sometimes when you do that math, however, you'll come up with negative numbers. ...Which are clearly not representative of reality. We're still trying to figure out all the factors that can cause that. One factor we've identified as at least a contributor is when you don't have STATISTICS_LEVEL=ALL, the numbers seem more often than not to be provably incorrect. |
Powered by FogBugz