Mohamed, My sincere apologies, you were a respondent yourself, I had mistaken you as the original poster. My sincere apologies for the reproof, no matter how gently intended; I can only plead lingering jetlag as an excuse. Thanks! -Tim On 10/18/2012 2:28 AM, Tim Gorman wrote: > Mohamed, > > A gentle observation... > > You have asked for help in this forum. Respondents (who cared enough > to respond promptly rather than just deleting your message) have > requested more information in order to help you. Yet in your next > response you choose to argue and to not provide the information? > > Please be assured that this list best helps those who help > themselves. Think about it. > > Good luck, > > -Tim > > > On 10/17/2012 9:08 AM, mohamed houri wrote: >> It is always good to post the oracle version information. But I am still >> unable to figure out how the absence of this information could halve the >> answer given by a real execution plan of a very simple query taking 300 >> seconds to complete. There are of course changes in the CBO from >> version to >> version that could explain why a CBO did choose a path here and not >> there. >> But, when confronted to a localized performance problem (a query for >> example) I am inclined to start by getting its execution plan first. If >> this reveals itself of no help then I will trace this query with the >> 10046 >> events to see what is happening behind the scene (SQL recursive >> calls, VPD >> policy,etc...) >> Best regards >> Mohamed Houri >> www.hourim.wordpress.com -- //www.freelists.org/webpage/oracle-l