Re: Missing Grant on Package Body

Strange - my sample was from 19.11

I wouldn't expect Oracle to remove a detail like that. Possibly, though, a change of text to "about to raise" MIGHT have happened. If you search for the table name (perhaps something more easily findable than "T" you might find a variation on the message.

An alternative from the level 1 trace file is to look for the "ERROR SIGNALED" lines which report "yes" and component "NULL" because the relevant SQL seems to be in the stack frame dump of one of those calls (and will probably be a few hundred lines away). Sample line, followed by a piece of the SQL that referenced my missing object:

2785 :CALL TYPE: call ERROR SIGNALED: yes COMPONENT: (null) 3140 :7F7A2B9C7160 2E317520 6C696166 00000031 00000000 [ u1.fail1.......]

Do you get anything similar in your level 1 trace?

Regards
Jonathan Lewis

P.S. Just before posting I did another check of the trace file from my call to "alter package pack1 compile body;" and realised that the trace file shows two very similar sets of trace information, and reports the two lines I've quoted at (after?) the end of the first dump but not at the end of the second. A few lines here around line 5,000 of 10,000:

• VKTM Time Drifts Circular Buffer ----- about to signal 942 Name: FAIL1
  • 2024-07-11T15:29:15.476508+01:00 (ORCLPDB(3)) dbkedDefDump(): Starting a non-incident diagnostic dump (flags=0x0, level=1, mask=0x0)
• Error Stack Dump -----

Do you also get two dumps in one?

On Thu, 11 Jul 2024 at 15:04, Mikhail Velikikh <mvelikikh_at_gmail.com> wrote:

> It can be seen from the output I provided previously that I searched for
> the word signal too.
>
>> egrep -iw 'signal|tc_table'
>> /u01/app/oracle/diag/rdbms/orcl/orcl/trace/orcl_ora_58350.trc
>
>
> The `-w` flag was used because there were multiple "ERROR SIGNALLED" lines
> such as one below:
>
>> CALL TYPE: call ERROR SIGNALED: no COMPONENT: PLSQL_Infrastructure
>>
> Even after removing the `-w` flag or using the exact command you were
> using I am not getting any hits on 19.23 with my specific test case.
> Which means that there are no such lines in my environment.
>
>>
>>>>>>

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