Re: diskspace usage recovered on db rebuild - Mailing list pgsql-general

From Lonni J Friedman
Subject Re: diskspace usage recovered on db rebuild
Date
Msg-id 7c1574a90504010928592f7f9@mail.gmail.com
Whole thread Raw
In response to diskspace usage recovered on db rebuild  (David Link <dlink@soundscan.com>)
List pgsql-general
On Apr 1, 2005 9:21 AM, David Link <dlink@soundscan.com> wrote:
> Greetings worthymen.
>
> I have a question regarding filesystem disk space usage.
>
> We have a production database containing 5 years of sales data.
> Linux 2.6.5; Postgresq 7.4.7.  VACUUM ANALZYE the entire database
> everynight (about 40min).
> It's size, @SUM(pg_class.relpages) * 8192K, is ...
>
>   About 66 Gigabytes on disk.
>
> When I rebuild the database (using pg_dump and pgsql ), the new
> resultant database is ..
>
>   About 48 Gigabytes on disk.
>
> A 27% space savings.
>
> Can someone tell me why that is?

Are you doing a vacuum full each night?  What is the specific
command(s) that you are using for vacuum, pg_dump and the import?


--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
L. Friedman                                    netllama@gmail.com
LlamaLand                       http://netllama.linux-sxs.org

pgsql-general by date:

Previous
From: David Link
Date:
Subject: diskspace usage recovered on db rebuild
Next
From: "Karl O. Pinc"
Date:
Subject: Re: Best practices for (plpgsql ?) trigger optimization?