Re: A 154 GB table swelled to 527 GB on the Slony slave. How to compact it? - Mailing list pgsql-general

From Aleksey Tsalolikhin
Subject Re: A 154 GB table swelled to 527 GB on the Slony slave. How to compact it?
Date
Msg-id CA+jMWofc2FtRYMX0883dZmrBNabnG7CzfB54NcfsVD8JxQAtNA@mail.gmail.com
Whole thread Raw
In response to Re: A 154 GB table swelled to 527 GB on the Slony slave. How to compact it?  (Vick Khera <vivek@khera.org>)
Responses Re: A 154 GB table swelled to 527 GB on the Slony slave. How to compact it?
List pgsql-general
On Tue, Mar 13, 2012 at 5:14 AM, Vick Khera <vivek@khera.org> wrote:
> I'll bet what happened was postgres re-wrote your table for you,
> effectively doing a compaction.  You can get similar effect by doing
> an alter table and "changing" an INTEGER field to be INTEGER.
> Postgres does not optimize that do a no-op, so you get the re-writing
> effect.

How does table rewriting work?  Does it happen a row at a time or all at once?

In other words, how much free disk space is needed on an 800 TB filesystem
to rewrite a 550 TB table?  (Have I got enough space?)

Aleksey

pgsql-general by date:

Previous
From: Scott Marlowe
Date:
Subject: Re: copy in date string "00-00-00 00:00:00"
Next
From: Rob Sargent
Date:
Subject: Did xpath_table get dropped.