Re: Making sense of the SQL trace file

Well,

HPUX has an operating system call with
a name like 'nano_sleep'; so they are clearly planning for a high-speed future.

I quess the fact that the timeslice is in hundredths doesn't preclude a hardware clock running in micro seconds, and being interrogated to that level of accuracy.

I did notice that on a fairly new fast NT box, the times I got back alway seemed to be in multiples of 291 microseconds - so perhaps that is the real granularity available at the operating system, despite the nominal capacity of the call.

--
Jonathan Lewis
http://www.jlcomp.demon.co.uk

Host to The Co-Operative Oracle Users' FAQ
http://www.jlcomp.demon.co.uk/faq/ind_faq.html

Author of:
Practical Oracle 8i: Building Efficient Databases

Screen saver or Life saver: http://www.ud.com
Use spare CPU to assist in cancer research.

K Gopalakrishnan wrote in message <0ac3v9.07d.ln_at_zstux.ita.pwr.wroc.pl>...


>Jonathan,

>

>I wonder how oracle gets time in millseconds. Except for Solaris and linux

>most of the Operating systems clocks in Centiseconds.  How oracle gets time

>in Milliseconds when the OS slices it in centiseconds?

>

>Best Regards,

>K Gopalakrishnan

>Bangalore, INDIA

>

>"Thomas Kyte" <tkyte_at_us.oracle.com> wrote in message

>news:9utfui014f8_at_drn.newsguy.com...

>> In article <1007825710.2791.0.nnrp-13.9e984b29_at_news.demon.co.uk>,

>"Jonathan

>> says...

>> >

>> >It's worth bearing in mind, though, that Oracle 9

>> >has started producing times in microseconds.

>> >(including tim=).  However there are a number of

>> >minor glitches and anomalies of use in the first

>> >releases.

>> >