From: | Edipo Elder Fernandes de Melo <edipoelder(at)ig(dot)com(dot)br> |
---|---|
To: | pgsql-sql(at)postgresql(dot)org |
Subject: | Large or Small tables? |
Date: | 2002-04-15 17:38:05 |
Message-ID: | 20020415173808.1190E4776F5@postgresql.org |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-sql |
Hi all,
I'm developing a selective system for selection of students for a
university. My system has tables for candidate, score, answers, exams, and
others.
The candidate table has about 40 columns, of that, 35 are candidate
information (name, address, etc), 10 are for score, status, etc, and use 2
columns as primary key (a idCandidate and idSelection).
The answers table has five columns: idSelection, idExam, idQuestion,
idCandidate, and answer. Better explained: for each answer, it has a record
(it's a lot of tuples - about 25000 candidates x 8 exams x 15 questions for
exams = 3 milions!)
The score table has idCandidate, idSelection, idExam, and score of each
candidate and exam.
Basicaly, to process the selection, I count how much questions the
candidate got correct, make some calc, and find a score.
My question is if is better have a unique candidate table with all
information of a candidate, or broke the table in two tables (one to
candidate information, and the other for score and status) for process of
find score. Or better, Having a large table in a middle of a process, where
i use only 10% of information on it, slow it down?
Thanks for all,
Edipo Elder
[edipoelder(at)ig(dot)com(dot)br]
_________________________________________________________________________
Voc podia estar baixando sua musica predileta, enquanto lia esse e-mail.
No perca tempo, tenha acesso rpido a internet com o Super iG.
http://registro.ig.com.br/superig
From | Date | Subject | |
---|---|---|---|
Next Message | Stephan Szabo | 2002-04-15 17:45:09 | Re: data order problem |
Previous Message | Josh Berkus | 2002-04-15 15:46:46 | Re: data order problem |