Re: Obtaining the exact size of the database.

From: Tom Lane
Subject: Re: Obtaining the exact size of the database.
Date: ,
Msg-id: 25289.1277048081@sss.pgh.pa.us
(view: Whole thread, Raw)
In response to: Re: Obtaining the exact size of the database.  (Dave Crooke)
List: pgsql-performance

Tree view

Obtaining the exact size of the database.  (venu madhav, )
 Re: Obtaining the exact size of the database.  (Dave Crooke, )
  Re: Obtaining the exact size of the database.  (Tom Lane, )
 Re: Obtaining the exact size of the database.  (Greg Smith, )
  Re: Obtaining the exact size of the database.  (Scott Marlowe, )

Dave Crooke <> writes:
> 4. If you're trying to figure out the net size of the table, i.e. how much
> free space is inside the table files for reuse by PG, then you need the
> pg_stat_tuple function ... this is built in to PG 8.4, and a plug-in
> activated by a script for PG 8.3, don't know if it exists in 8.1 or not.
> Like SELECT COUNT(*) this requires a full table scan.

I think what the OP actually wants is the number of live rows, so plain
old SELECT COUNT(*) would do it.  If that's too slow, a good alternative
is to ANALYZE the table and then look at its pg_class.reltuples entry
--- of course that will only be an approximate count.

            regards, tom lane


pgsql-performance by date:

From: Jesper Krogh
Date:
Subject: Aggressive autovacuuming ?
From: Greg Smith
Date:
Subject: Re: B-Heaps