Jonathan, >You can also play around with v$sql_workarea to get some idea >of what big operations have used a lot of CPU (rather than >hammering v$sql - which is a bit labour intensive and latch- >unfriendly in a highly concurrent system). I thought that queries on v$sqlarea were to be avoided in favour of v$sql since the former requires a group by on x$kglcursor while the latter is a simple select on x$kglcursorr? Would that not prolong the query and increase the chances of latch contention when v$sqlarea is queried instead of v$sql? [I checked this on 8.1.7, 9.2.0.4 and 10.1.0.20.] Thanks for all your (very valuable) inputs to this list! John Kanagaraj ---------------------------------------------------------------- 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 http://www.freelists.org/archives/oracle-l/ FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html -----------------------------------------------------------------