> Ok, I see you have a distinct preference for symptom fighting, and
> following your own interpretations, instead of accepting advice from
> others.
> One might want to ask why you posted this at all, as you already know
> the solution.
> Yet I must warn you: you are using wrong criteria. It doesn't tell
> anything the top 5 sql executes in less than 20 seconds, what matters
> is the number of buffers it uses.
> Changing cursor_sharing to force also seldomly fixes anything, it makes
> sure the patient suffers less from terminal cancer, yet he still
> suffers from cancer.
> So if you want to continue to follow your own path, I won't block you,
> I just won't respond to any of your future posts.
>
> --
> Sybrand Bakker
> Senior Oracle DBA

Do not take me wrong. I really appreciate your help and took your
advice to look for the TOP SQL from day one. And I will let you know
the result after we examine all the expensive SQLs.

Thanks