If you're in 10g, could you use the ASH data for this troubleshooting? I can't remember the exact frequency, but it's very frequent as well and stored at this raw level for an hour. Then rolled up and stored for 7 days. Thanks, Finn On 2/1/08, Zhu,Chao <zhuchao@xxxxxxxxx> wrote: > > hi, all > We have a homegrown script that pulls different views to log many > different system level statistics for troubleshooting, and one of the view > we query is v$system_event/x$kslei. As we does the pull of data every 10 > seconds, it is very CPU consuming. The target of querying the view, is to > give a break down of system wide wait events in the past 10 seconds (that > is, say, db-file-sequential-wait, 1000ms, 30%,, enqueue, 300ms, 20%...) > Is there any possible replacement for this view? We want to reduce the > CPU usage of this script; > > -- > Regards > Zhu Chao > www.cnoug.org >