Re: Quick estimate of num of rows & table size

From: Lonni J Friedman <netllama(at)gmail(dot)com>
To: Thalis Kalfigkopoulos <tkalfigo(at)gmail(dot)com>
Cc: "pgsql-general(at)postgresql(dot)org" <pgsql-general(at)postgresql(dot)org>
Subject: Re: Quick estimate of num of rows & table size
Date: 2012-11-05 22:14:07
Message-ID: CAP=oouHENneb0Jq2ADDmVqHGj=oKx76s3RNvkSvE7qnCW7s9ag@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

On Mon, Nov 5, 2012 at 2:02 PM, Thalis Kalfigkopoulos
<tkalfigo(at)gmail(dot)com> wrote:
> Hi all,
>
> I read somewhere that the following query gives a quick estimate of the # of
> rows in a table regardless of the table's size (which would matter in a
> simple SELECT count(*)?):
>
> SELECT (CASE WHEN reltuples > 0 THEN
> pg_relation_size('mytable')/(8192*relpages/reltuples)
> ELSE 0
> END)::bigint AS estimated_row_count
> FROM pg_class
> WHERE oid = 'mytable'::regclass;
>
> If relpages & reltuples are recorded accurately each time VACUUM is run,
> wouldn't it be the same to just grab directly the value of reltuples like:
>
> SELECT reltuples FROM pg_class WHERE oid='mytable'::regclass;
>
> In the same manner, are pg_relation_size('mytable') and 8192*relpages the
> same?
>
> I run both assumptions against a freshly VACUUMed table and they seem
> correct.

This doesn't seem to work for me. I get an estimated row_count of 0
on a table that I know has millions of rows.

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Thalis Kalfigkopoulos 2012-11-05 23:56:46 Re: Quick estimate of num of rows & table size
Previous Message Thalis Kalfigkopoulos 2012-11-05 22:02:46 Quick estimate of num of rows & table size