From: | Alice Lottini <alice_lottini(at)yahoo(dot)it> |
---|---|
To: | PostgreSQL Mailing List Hackers <pgsql-hackers(at)postgresql(dot)org> |
Subject: | inserting tuples into disk blocks |
Date: | 2003-05-02 15:36:39 |
Message-ID: | 20030502153639.7422.qmail@web13709.mail.yahoo.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
hi everybody,
we've noticed a seemingly strange behaviour when
inserting fixed-size tuples into disk blocks.
We created a new table (which obviously is empty) and
we inserted some fixed-size tuples by directly using
the interface function simple_heap_insert. All the
tuple attributes are not null and fixed-size and no
primary key has been defined on the table.
The strange thing is that the number of tuples stored
on a single block is different on each block: for
example, the first holds 272 tuples, the second 198,
the third 180 and so on. Note that the blocks are all
new as the table has been freshly created and no tuple
has ever been deleted or updated.
We repeated the process, with the same data, on
another newly created table but the blocks are filled
exactly in the same way, so it seems like this is a
behaviour based on some evaluation.
Does anybody know why this happens?
Thanks a lot!
alice and lorena
______________________________________________________________________
Yahoo! Cellulari: loghi, suonerie, picture message per il tuo telefonino
http://it.yahoo.com/mail_it/foot/?http://it.mobile.yahoo.com/index2002.html
From | Date | Subject | |
---|---|---|---|
Next Message | Jenny - | 2003-05-02 20:10:31 | Using GetLockStatusData in main |
Previous Message | Michael Wildpaner | 2003-05-02 13:09:28 | join over 'view ... union all' ignores indices |