Just add on, It looks the similar on 9i too. Why non-ASSM table is spending more pure CPU then ASM? It can be because of this totally specific test, but why? tkprof and Oracle v$ views showed that CPU time is just different as I stated, very different. This is telling me that ASSM is just much much faster. Regards, Zoran --- Martic Zoran <zoran_martic@xxxxxxxxx> wrote: > Thanks a lot Anjo, > > > ASSM will not do that many recursive SQL (data > > dictionary operations). > I understand that, but here I am not seeing them > with > the trace 10046. > So I said it is internal to Oracle algorithm how it > is > done. > > > difference is made there.You could probably get > > better results with non-ASSM > > by using freelists and using large extents. Your > > difference is purely in the > > recursive SQL part. > For me it is the same speed, whatever extent size. > I am not expecting that to be the problem either > because I did truncate reuse storage too before the > test. > > Did Oracle change something from 10g? > > I should go and do the test on 9i because now I am > curious. > > I am maybe stupid and do not get it. > > Regards, > Zoran > > > > --- Anjo Kolk <anjo.kolk@xxxxxxxxxxx> wrote: > > > > > ASSM will not do that many recursive SQL (data > > dictionary operations). The > > difference is made there.You could probably get > > better results with non-ASSM > > by using freelists and using large extents. Your > > difference is purely in the > > recursive SQL part. > > > > Anjo. > > . > > > > > __________________________________ > Do you Yahoo!? > Yahoo! Small Business - Try our new resources site! > http://smallbusiness.yahoo.com/resources/ > -- > //www.freelists.org/webpage/oracle-l > __________________________________ Do you Yahoo!? Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/ -- //www.freelists.org/webpage/oracle-l