From: | "Ligia Pimentel" <pimentel_ligia(at)hotmail(dot)com> |
---|---|
To: | christoph(dot)dalitz(at)hs-niederrhein(dot)de |
Cc: | pgsql-general(at)postgresql(dot)org, MuK(dot)Rudolph(at)t-online(dot)de |
Subject: | Re: Table with 90 columns |
Date: | 2002-09-17 01:08:50 |
Message-ID: | F89TJby0LKhogUojTpe0000eec9@hotmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Hi.
You are right, if you are always using always all 90 fields of the record
and if it is normalized means you would have to make a lot of joins with
other tables, but, MY THEORY is that if the records are longer, the buffer
will read less records each time you access (physically) the disk, and (even
on an optimally indexed table, you will have more disk access, so it wil be
slower. Of course, this could be set up by fixing buffer sizes, but... I
would always suspect a table with 90 columns has a lot of redundancy on it,
and you will find other problems later, (again, I work by this rule...).
Anyway, On this case, I think that changing the database structure would
imply changes on the server side, not on the client side (unless all the
logic of the application is on the client-side, on user pages) which is not
desirable... again and this is MY THEORY, I would use server side components
(java servlets, ISAPIs, or something like that) which would make your
application more manageable,...
On the other side, remember, If you give the same problem to 10 software
engineers, you would surelly get 10 different solutions, so, I guess, if it
works for you...
Ligia
You wrote...
>As far as I understand normalization it is meant to avoid *redundance* and
>not to improve performance. Actually normalization in general decreases
>performance becaus a join over several tables is much less efficient than a
>select on a single table.
_________________________________________________________________
Join the worlds largest e-mail service with MSN Hotmail.
http://www.hotmail.com
From | Date | Subject | |
---|---|---|---|
Next Message | Bruce Momjian | 2002-09-17 01:56:37 | Re: pg_proc and pg_type |
Previous Message | Jason Earl | 2002-09-17 00:23:57 | Re: currval question |