Catproc and Catalog Executed as non-sys user

Question

One of my friend executed the above packages from a non-sys user schema.
That schema is also important.
Now how to clean up these objects?
Just re-running the above packages as sys schema is enough?
Thanks

Answer 1

Excerpts from the note 1015902.4:
Shutdown your database with 'shutdown immediate'. If this hangs for an
extended period of time, then use 'shutdown abort'. This must be done
as 'internal'.
Start up your database with 'startup', or 'startup restrict'.
The following script should be run as 'sys' or 'internal' from SQL*Plus
(this is important for formatting the script that it creates). This script
will create another script called file.sql. If you do not run the following
script from SQL*Plus, you will need to edit the resulting script to make sure
file.sql will run properly.
----------------------------- <script start> --------------------------
set echo off
set pagesize 0
set feedback off
spool drop.sql
SELECT 'drop '||OBJECT_TYPE||' '||OWNER||'.'||object_name||';'
FROM all_objects
WHERE OWNER != 'SYS' AND OBJECT_TYPE != 'SYNONYM'
AND OBJECT_NAME IN
(SELECT OBJECT_NAME FROM ALL_OBJECTS
WHERE OWNER = 'SYS')
ORDER BY object_type, object_name;
spool off
set feedback on
set echo on
---------------------------- <script end> -----------------------------
Run the script created from above script as either internal or sys.
The script will contain sql commands like:
DROP PACKAGE SYSTEM.DBMS_SHARED_POOL;
DROP VIEW SYSTEM.DBA_OBJECTS;
SQL> @drop.sql
You may alternately find the user who owns 'sys' objects, and drop that
user.
SQL> drop user xxxx cascade;
This may hang. If it does, use shutdown abort, and rerun script. When the
script completes, 'shutdown immediate' your database. Startup and run
catalog.sql and catproc.sql as either 'sys' or 'internal'. You may then
run pupbld.sql as the 'system' user. If you have replication installed,
you may also need to run catrep.sql as 'sys'.
Hints for running the catalog, catproc and catrep scripts:
1. Check for available freespace in system tablespace, you will probably
want at least max(bytes) at least 4M, and sum(bytes) of 10M:
SQL> select max(bytes) from dba_free_space where tablespace_name = 'SYSTEM';
SQL> select sum(bytes) from dba_free_space where tablespace_name = 'SYSTEM';
2. Make system tablespace datafile autoextend:
SQL> alter database datafile 'path\filename' autoextend on
next 5m maxsize 100m;
View dba_data_files for file names
SQL> select * From dba_data_files where tablespace_name = 'SYSTEM';
3. Make rollback_data tablespace datafile autoextend:
SQL> alter database datafile 'path\filename' autoextend on
next 5m maxsize 200m;
4. Prior to running data dictionary scripts, spool to file
SQL> spool file.log
5. Check file.log and alert<SID>.log after execution for errors.
Solution Explanation:
=====================
When CATALOG.SQL or CATPROC.SQL were run as SYSTEM or another schema owner,
this created some valid and invalid objects owned by SYSTEM (in view
DBA_OBJECTS) that should be owned by SYS only. Comparing those object owned
by SYS with objects owned by other schemas and deleting the duplicate
objects removes the duplicate objects.