Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Mailing Lists -> Oracle-L -> RE: Index-Organized Table experiences
Performance, etc, was excellent. The only problem was with using the Oracle
development tools (Designer and Developer) with the IOT's. The tools didn't
differentiate between an IOT and a "regular" table, and wanted to generate
code that used ROWID's; IOT's don't have ROWID's.
-----Original Message-----
From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org]On Behalf Of Daniel Fink
Sent: Wednesday, April 28, 2004 10:15 AM
To: oracle-l_at_freelists.org
Subject: Index-Organized Table experiences
We are looking to implement IOTs for a couple of intersection entities in a 10g db. I would like to hear from those brave enough to actually use IOTs what is the good, the bad and the ugly.
example:
Employee (heap table)
Project (heap table)
There is a many-to-many relationship between the tables (1 employee can be on many projects and 1 project can have many employees).
The emp_project table is the intersection entity containing emp_id and project_id as the only columns. There are FK constraints on each of the columns. The combination of emp_id and project_id is unique.
This situation *sounds* like the right one for an IOT, otherwise we would have 1 table and 2 indexes (1 on each column).
My main concerns are:
1) Integrity/performance
2) Locking behavior (do I need to adhere to the traditional
"index all foreign keys" rule to prevent excessive locking?)
3) Any especially nasty gotchas
Thanks,
Daniel
![]() |
![]() |