Note in-line Regards Jonathan Lewis http://www.jlcomp.demon.co.uk http://www.jlcomp.demon.co.uk/faq/ind_faq.html The Co-operative Oracle Users' FAQ http://www.jlcomp.demon.co.uk/seminar.html Optimising Oracle Seminar - schedule updated May 1st :----- Original Message ----- :From: "Cary Millsap" <cary.millsap@xxxxxxxxxx> :To: <oracle-l@xxxxxxxxxxxxx> :Sent: Monday, June 07, 2004 9:08 PM :Subject: RE: CBO irregularity : :It's what CBO is built to do: take into account a tremendous number of :variables, each of which has a right to influence the optimizer's choice for :best plan. CBO tends to work pretty well when (a) you give it the :information it needs, and (b) you let it do its job (okay, and maybe also :"(c) you wait 'til 9.2 to use it"). I think (a) you give it the information it needs (b) you don't lie to it (c) you don't try to mislead it (d) you don't try to cheat it (e) you don't try to decieve it (f) you don't mess with hints you don't understand (g) you don't just fiddle until something works (h) ..... (z) you give it a little help to work around a few teensy, weensy buggettes But the optimizer in 10g is so much better than the ones that went before - it's going create havoc when people upgrade because so many bugs have been fixed. The optimizer keeps working out the right answers - which means any work arounds you have in place now may cause problems on your 10g upgrade. ---------------------------------------------------------------- Please see the official ORACLE-L FAQ: http://www.orafaq.com ---------------------------------------------------------------- To unsubscribe send email to: oracle-l-request@xxxxxxxxxxxxx put 'unsubscribe' in the subject line. -- Archives are at //www.freelists.org/archives/oracle-l/ FAQ is at //www.freelists.org/help/fom-serve/cache/1.html -----------------------------------------------------------------