Re: Too many IO?

From: Alvaro Herrera <alvherre(at)commandprompt(dot)com>
To: Tatsuo Ishii <ishii(at)postgresql(dot)org>
Cc: Pg Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Too many IO?
Date: 2012-03-14 19:07:32
Message-ID: 1331750583-sup-3408@alvh.no-ip.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers


Excerpts from Tatsuo Ishii's message of mar mar 13 23:29:44 -0300 2012:

> As you can see, this query generated 1255+1250 = 2505 times block read
> either from the buffer or the disk. In my understanding the query
> accesses an index tuple, which will need access to root page and
> several number of meta pages (I mean index pages they are not either
> root or leaf pages)

To clarify terminology a bit -- we call "metapage" of a btree index the
page number zero, which among other things contains the page number for
the root page. That is, since the root page can move around due to
splitting, its number can always be found by reading the metapage.

Pages that aren't root nor leaf pages are typically called "internal
pages", though there are other names for them ("non-leaf").

--
Álvaro Herrera <alvherre(at)commandprompt(dot)com>
The PostgreSQL Company - Command Prompt, Inc.
PostgreSQL Replication, Consulting, Custom Development, 24x7 support

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Jeff Janes 2012-03-14 19:29:37 Re: wal_buffers, redux
Previous Message ktm@rice.edu 2012-03-14 18:13:27 Re: Faster compression, again