Our solution to ORA-00600: internal error, arguments: [kkmmctbf:bad intcoln]

We were facing a problem in one of our databases with jobs that stopped immediately after attempting to start, and then aborting their respecitive processes immediately.

Error:

2024-08-20T10:09:51.641274+02:00
Errors in file /orasw/rdbms/diag/rdbms/cdb/cdb/trace/cdb_j000_3187052.trc:
ORA-00600: intern feilkode, argumenter: [kkmmctbf:bad intcoln], [0], [], [], [], [], [], [], [], [], [], []
2024-08-20T10:09:51.645053+02:00
opidrv aborting process J000 ospid (3187052) as a result of ORA-600

Cause:

Right prior to the incident we ran out of free space in the SYSTEM tablespace, and a new datafile was promptly added.
The internal table SYS.SOURCE$ contains all the PL/SQL code for the database ( Oracle provided code and your own code )
Our schemas are editioned, and for some reason we seem to be hitting bug number 14163397: "Trigger on editioning view can get ORA-600 [kkmmctbf:bad intcoln] (Doc ID 14163397.8)"

Solution:

alter system set "_ignore_fg_deps"=ALL container=all scope=both;

The above parameter is used to ignore fine grain dependencies during invalidation. Flush the shared pool and buffer cache if its holding up in memory the old versions:

alter system flush shared_pool;
alter system flush buffer_cache;

Wait for a little while, then the jobs started to run automatically again.

The above solution was found with assistance from Oracle Support and should not be attempted in a production environment without prior investigation and only upon recommendation from Oracle Support.