Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Usenet -> c.d.o.misc -> Re: General question about fk constraints ...
FK constraints are an overhead but they are miniscule.
What is a *far* bigger overhead is the debugging and fixing of obscure
bugs in applications written for schemas which do not use FK
constraints.
The reason for this is simple - without FX constraints, applications
which break the relational model will leave the database in a mucky
state until years later when some poor sod (me) discovers the
inconsistency.
It is FAR better for applications to fall over at the point of failure rather than at the point of discovery!
In summary: FK constraints are good!
ETA.
"Jan Gelbrich" <j_gelbrich_at_westfalen-blatt.de> wrote in message news:<b27ql8$1a92b9$1_at_ID-152732.news.dfncis.de>...
> ... that is: about the drive of many software vendors to *avoid* them ...
>
> Hello !
>
> Just as a matter of interest, I would like to know *why* are so many data
> models implemented out there
> are neglecting fk constraints to a degree that it makes me almost mad -
> sorry ...
>
> When I was trained in ER design 1 and 1/2 year ago,
> we were told never to sacrifice fk constraints for the douptful sake
> of performance,
> and we set up easy-to-read ERMs with countable table numbers and so on ...
>
> Now, on my working place I see (and from many others I hear) that it seems
> very common
> *not* to use constraints, just for the sake of performance - all the guys
> around me say that, and they are
> just asking "R U kidding ?! *Why* for *** sake do You wanna use constraints
> ?! *You* as dba shoulda known better ... "
> which makes me almost speechless for a moment 8|
>
> On a recent post in the Oracle NGs I read that Peoplesoft and SAP are going
> just the same way, having
> thousands of tables in their apps. No way to Reverse Engineer because just
> by viewing the ERM, nothing can be
> understood.
>
> Am I too cynic when I come to think that all this is just to hide their apps
> logics behind a labyrith of tables,
> or does this general avoidance of constraints have a meaning that I still do
> not happen to understand ?
>
>
> Maybe this is a straight newbee question, but I really would appreciate any
> comments.
>
> TIA
>
> Jan
Received on Mon Feb 10 2003 - 09:08:02 CST
![]() |
![]() |