RE: Limits on referential integrity

From: <Jared.Still_at_radisys.com>
Date: Mon, 21 Jan 2002 16:57:29 -0800
Message-ID: <F001.003F6900.20020121163023@fatcity.com>

Yeah, I've heard the performance arguments before.

I've never worked on a really large OLTP system, but I have worked on some of 20 gig or so with a few key tables having millions of rows ( 20 gig used to be big! ).

One in particular had quite a bit of RI. Running on a DG/UX system with 512M Ram and 4 cpus, with 3 mirrored pairs and 2 RAID 5 stripes, we were generally happy with the performance.

The transactions were very query intensive and performed upto about 20k transactions per day with an average transaction time of 1-2 seconds. These were very complex transactions.

There were also 10-20 operators at any one time entering manual transactions and doing customer service from the same database.

There was a CFO requesting all kinds of complex reports during the middle of the day. Did I forget to mention that Oracle apps 9 was on the same box?

RI was never to blame for performance problems.

In fact, our biggest performance problem was non-performance when the database died. ( Steve, I couldn't resist :)

This was on Oracle 7.0.16 to start with. RI was not the cause of any performance problems that I can recall.

One thing that kept the RI overhead low was that all transactions tables had surrogate keys generated from a sequence.

Jared

DENNIS WILLIAMS <DWILLIAMS_at_LIFETOUCH.COM> Sent by: root_at_fatcity.com
01/21/02 02:35 PM
Please respond to ORACLE-L

        To:     Multiple recipients of list ORACLE-L <ORACLE-L_at_fatcity.com>
        cc: 
        Subject:        RE: Limits on referential integrity

Jared - I wasn't clear, but then again it is Monday. I have a team of inexperienced developers starting a big, new Java application. They have a good, experienced data model consultant helping them create the data model.
They are eager to include referential integrity. So eager it has me a little
worried. My question: "Is there too much of a good thing?". In Oracle 7, sometimes sites would remove RI to ensure good performance (we are starting
this project on Oracle9i). Has anyone encountered problems with too many constraints? Any guidelines you use with developers? Thanks. Dennis Williams
DBA
Lifetouch, Inc.
dwilliams_at_lifetouch.com

-----Original Message-----
Sent: Monday, January 21, 2002 4:16 PM
To: Multiple recipients of list ORACLE-L

I would be you lunch that what they are implementing in their code is not actually RI. They may be implementing code to ensure things get inserted in the right order, and that child rows have a parent.

This is a very weak form of RI. Oracle is very good at implementing RI, and it is not dependent on an application. RI in the database is the route to choose unless there is some good reason not to.

RI in the database will prevent orphaned data created through updates, deletes or even ( gasp! ) bugs in the app.

Programmers tend to dislike RI in the database because it forces them to maintain data integrity in a transaction. This is not a bad thing, it just forces them to have a good understanding of their transactions.

Point out to them that it is less code to write as well. :)

Jared

DENNIS WILLIAMS <DWILLIAMS_at_LIFETOUCH.COM> Sent by: root_at_fatcity.com
01/21/02 01:35 PM
Please respond to ORACLE-L

To: Multiple recipients of list ORACLE-L <ORACLE-L_at_fatcity.com>

        cc: 
        Subject:        Limits on referential integrity

How much referential integrity should be implemented in Oracle? We are starting a large new Java project. Our current applications keep their referential integrity inside their own dictionary, so I haven't had to deal
much with referential integrity recently. Can there be too much of a good thing? What guidelines do you tend to use? At this point the developers are
designing the data model so they are busily linking all the little boxes. My
attitude at this point is "implement what you've got and if there are performance problems we'll deal with them when they arise". Can anyone give
me a better motto?
Thanks.
Dennis Williams
DBA
Lifetouch, Inc.
dwilliams_at_lifetouch.com

--

Please see the official ORACLE-L FAQ: http://www.orafaq.com
--

Author: DENNIS WILLIAMS
INET: DWILLIAMS_at_LIFETOUCH.COM

Fat City Network Services    -- (858) 538-5051  FAX: (858) 538-5051
San Diego, California        -- Public Internet access / Mailing Lists
--------------------------------------------------------------------

To REMOVE yourself from this mailing list, send an E-Mail message to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).

--

Please see the official ORACLE-L FAQ: http://www.orafaq.com
--

Author:
INET: Jared.Still_at_radisys.com

Fat City Network Services    -- (858) 538-5051  FAX: (858) 538-5051
San Diego, California        -- Public Internet access / Mailing Lists
--------------------------------------------------------------------

Fat City Network Services    -- (858) 538-5051  FAX: (858) 538-5051
San Diego, California        -- Public Internet access / Mailing Lists
--------------------------------------------------------------------

--

Please see the official ORACLE-L FAQ: http://www.orafaq.com
--

Author:
INET: Jared.Still_at_radisys.com

Fat City Network Services    -- (858) 538-5051  FAX: (858) 538-5051
San Diego, California        -- Public Internet access / Mailing Lists
--------------------------------------------------------------------