Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
Home -> Community -> Usenet -> c.d.o.server -> Re: Performance problem on inserting
"Niall Litchfield" <n-litchfield_at_audit-commission.gov.uk> wrote in message
news:3e37cfdd$0$230$ed9e5944_at_reading.news.pipex.net...
> "Telemachus" <telemachus_at_ulysseswillreturn.net> wrote in message
> news:RKPZ9.6140$V6.8296_at_news.indigo.ie...
> > Do alter session set sql_trace=true for sql*plus sessions on both
> machines,
> > and do a 10046 trace and rerun the scripts.
> >
> > then run tkprof with the wait option and post the results. let's see
> what's
> > being waited on.
> >
> > the 8i tk can't do the 10046 wait analysis but it would be useful to
have
> > too.
>
> AFAIK You can run the 9i tkprof against an 8i trace file.
>
> <disclaimer>
> the script tells you not to do this but gives no reasons
> </disclaimer>
I believe the reason is (apart from anything that is peculiar in the depths of the code) that version 9 calculates to the millisecond, and version 8 to the centisecond, so all durations etc reported are out by a factor of 10.
Not difficult to adjust for, I agree.
Regards
HJR
Received on Wed Jan 29 2003 - 13:45:42 CST