Re: 9iDB Security Hole?

There are physical datastructures in the database which are not backwards compatible - for example if you try to take a database with unlimited extents back from 8 to 7 (I think those are the right numbers) by setting the compatibility, you find that the code recognises that the markers for unlimited extents exist in segment headers, but the version you are pretending do be isn't supposed to know how to use them.

Sometimes you can clean out the physical problem sometimes you can't.

The idea was good though - I tried the same with

    optimizer_features_enable=8.1.7
just in case - but it didn't work.

Latest news - I raised an iTAR with metalink pointing out that I couldn't use a 9i database in production with this bug in place. This is now with developers looking at a backport, but with no date for the backport currently predictable.

--
Jonathan Lewis
http://www.jlcomp.demon.co.uk

Author of:
Practical Oracle 8i: Building Efficient Databases

Next Seminar - Australia - July/August
http://www.jlcomp.demon.co.uk/seminar.html

Host to The Co-Operative Oracle Users' FAQ
http://www.jlcomp.demon.co.uk/faq/ind_faq.html



Niall Litchfield wrote in message
<3cbd8484$0$227$ed9e5944_at_reading.news.pipex.net>...


>I get the below when setting compatible to 8.1.7.0.0. Now the error does

>point to my having done something 9i specific but I'm not clear what this

>is - the db is a clean 9i install.

>

>SQL> startup pfile=c:\oracle\admin\<snip>\pfile\init.ora

>ORACLE instance started.

>

>Total System Global Area  118255568 bytes

>Fixed Size                   282576 bytes

>Variable Size              83886080 bytes

>Database Buffers           33554432 bytes

>Redo Buffers                 532480 bytes

>Database mounted.

>ORA-00402: database changes by release 9.0.0.0.0 cannot be used by release

>8.1.7.0.0

>ORA-00405: compatibility type "PDML ITL invariant"

>

>

>--

>Niall Litchfield

>Oracle DBA

>Audit Commission UK

>*****************************************

>Please include version and platform

>and SQL where applicable

>It makes life easier and increases the

>likelihood of a good answer

>

>******************************************

>

>

>"Todd M. Helfter" <tmh_at_jumpgate.cc.purdue.edu> wrote in message

>news:a9jk9h$66b$1_at_mozo.cc.purdue.edu...

>>

>> Can anyone think of an init.ora parameter to disable all users from

having


>access

>> to the security hole?  Something like "TURN_OFF_SQL92_BLAH_BLAH" or would

>setting

>> compatibility to 8.1.7 have the same effect?

>>

>> --

>> Todd M. Helfter

>> Database Analyst/Programmer

>> Purdue University

>> tmh_at_purdue.edu

>

>