Re: unable to generate trace on HP-UX

It's a very nice concept, I've seen Tanel use it to show trace output in a live sql*plus session, and just slightly "abused" it for this, been using it a couple times. It also works if you don't have access to SYS, or if you want to automate the process.

However, as Gints Plivna has pointed out, there's always another way to do it :-)
But if you're not inside that session you're trying to trace, that won't work, as unfortunately, there's no
dbms_system.set_varchar2_param_in_session. And it sounded like the session was still active from his description.

Stefan

On 8/14/06, Kurt Van Meerbeeck <kurtvm_at_pandora.be> wrote:
>
>
> The named pipe solution is quite nice...
>
> Maybe this also helps - if the fd is still open, but the file was removed,
> you
> can do the following :
> SQL> oradebug setospid <pid>
> Statement processed.
> SQL> oradebug close_trace
> Statement processed.
>
> And then enable the trace again
>
> Cheers,
> Kurt
>
> At 10:38 14/08/2006 +0200, Stefan Knecht wrote:
> >
> > Perhaps you have enabled the trace in a certain session, analyzed (and
> > removed) the tracefile and disabled the event.
> > If you then, later on re-enable the trace for the same session, no trace
> file
> > will be written.
> >
> > The reason is, that Oracle keeps the file descriptor open, pointing to
> the
> > "old" file still, which isn't there anymore. No new file will be
> created.
> >
> > A workaround is, BEFORE you enable your event, figure out the name of
> the
> > tracefile that will be written to (ORacle uses uniform formats, can
> easily be
> > predicted) , and put a named pipe in its place. Then, when you enable
> the
> > trace, you start reading from that pipe (using i.e. tail -f ), and write
> that
> > output to a file you want to analyze. Using this technique, you can
> easily
> > enable and disable the trace, and it'll work just fine, as long as the
> named
> > pipe stays in place.
> >
> > Stefan
> >
> >
> > On 8/14/06, Syed Jaffar Hussain
> > <<mailto:sjaffarhussain_at_gmail.com>sjaffarhussain_at_gmail.com> wrote:
> >>
> >> Hello List,
> >>
> >> Recent past, we were able to generate trace files (setting events) on
> >> one of our database (9i) on HP-UX.
> >> Today, surprisingly, when we set trace, there is no trace file
> >> generated at the udump?
> >> What could cause this problem? There is another db running on the same
> >> server, were trace files are generating.
> >>
> >> --
> >> Best Regards,
> >> Syed Jaffar Hussain
> >> 8i,9i & 10g OCP DBA
> >> Banque Saudi Fransi,
> >> Saudi Arabia
> >>
> >> I blog at :<http://jaffardba.blogspot.com/>
> http://jaffardba.blogspot.com/
> >> http://www.oracle.com/technology/community/oracle_ace/ace1.html#hussain
> >>
> -------------------------------------------------------------------------
> >> ---------
> >> "Winners don't do different things. They do things differently."
> >> --
> >>
> >> <http://www.freelists.org/webpage/oracle-l>
> http://www.freelists.org/webpa
> >> ge/oracle-l
> >>
> >
>
>
>
> --
> Kurt Van Meerbeeck
> kurtvm_at_pandora.be
> kurt_van_meerbeeck_at_axi.be
> dude_at_ora600.org
> http://www.ora600.org
>
> Imagination is more important than knowledge...
>
> The instance has crashed.
> I am the ORA 600.
> No one hears your screams.
>
>
>
>
>

--
http://www.freelists.org/webpage/oracle-l