Re: Help interpreting TKPROF output

Thanx guys for your feedback,

   even if that didn't provide anything concrete so far. As for your suggestions to throw Siebel out the window, I perfectly agree with you, and have been dreaming about this day for about 4 years now (since I started my initial Siebel training), but I'm afraid it's not a practical choice career-wise at the moment. The reason there are so many outer-joins is that, for example, we still want to display an "activity" (table SIEBEL.S_EVT_ACT T1) even if there is no "opportunity" (table SIEBEL.S_OPTY T3) associated to it. That seems reasonable to me. I basically have to accept the data model as it is, with its strenghts (I'm still looking for them!) and its weaknesses.

    What I would like with respect to this statement is to be able to modify the "driving table" from SIEBEL.S_EVT_ACT (T1) to a smaller. I realize that the outer joins limit my ability to do this (a table "outer-joined" can't ever be chosen as the driving table by the RBO), but I thought I'd ask anyway in case anyone came through the same painful process as me.

Daniel

> wrote:
>
> >
> >Anyone has an idea as to what might be wrong? Or is there any other
> >tool (besides TKPROF) that they might use to better tune this
> >statement?
>
> 1 The datamodel
> 2 The datamodel
> 3 The datamodel
>
> The statement shows they are outer-joining about everything, so what
> would you expect?
> The best tool to tune this statement is an open window: take the
> statement and throw it out, and Siebel with it.
> I don't think anything can be done about the execution path, or you
> should provide full information on all the indexes on all the tables.
> Apart from throwing the app out of the window, your best bet is to
> throw more iron at the problem.
Received on Fri Nov 07 2003 - 20:53:56 CST