Re: Tracefiles not generated

I get this all the time on various versions of oracle running on solaris :(

I think it has something to do with the trace file exceeding the

max_dump_file_size

and thus the trace file doesn't grow beyond the set size. Even after changing it.

exec sys.dbms_system.set_int_param_in_session(sid, serial#, 'max_dump_file_size', 10000000);

it doesn't appear to fix the problem for me :(

Sometimes it doesn't seem to trace at all.

I now exclusively use the following method which I find works every single time.

• Try this
• get the spid from the v$process table for the session that
• you want to trace
• Try the relevant commands, from below, from sqlplus,

oradebug help

oradebug setospid 21700
oradebug UNLIMIT
oradebug Event 10046 trace name context forever, level 12 oradebug Event 10046 trace name context forever, level 8 oradebug tracefile_name
oradebug Event 10046 trace name context off oradebug procstat
oradebug close_trace
oradebug flush

nospam_kurt.laugesen_at_gmail.com wrote:
> I am trying to trace my sql (setting event 10046 either in my own
> session or using dbms_system.set_ev).
> I have done this many times before, but I am now working on some new
> instances, that I have not created, and the funny thing is that nothing
> happens. No tracefiles are generated and no errors reported.
> Only when i set tracefile_identfier=<something> do the files get
> generated.
> This means I CAN in fact do my job, but it still puzzles me, I have
> never seen this behaviour before.
>
> It also makes it hard to trace another session, since I have to assure
> that they set tracefile_identier in their session before I start
> tracing a part of a long batchrun.
>
> Kind regards
> Kurt Laugesen
>
Received on Wed Sep 20 2006 - 19:30:00 CDT