Re: auto-sizing wal_buffers - Mailing list pgsql-hackers

From Josh Berkus
Subject Re: auto-sizing wal_buffers
Date
Msg-id 4D3220E3.20804@agliodbs.com
Whole thread Raw
In response to Re: auto-sizing wal_buffers  (Greg Smith <greg@2ndquadrant.com>)
Responses Re: auto-sizing wal_buffers  (Tom Lane <tgl@sss.pgh.pa.us>)
Re: auto-sizing wal_buffers  (Marti Raudsepp <marti@juffo.org>)
Re: auto-sizing wal_buffers  (Fujii Masao <masao.fujii@gmail.com>)
Re: auto-sizing wal_buffers  (Jeff Janes <jeff.janes@gmail.com>)
Re: auto-sizing wal_buffers  (Greg Smith <greg@2ndquadrant.com>)
List pgsql-hackers
On 1/14/11 10:51 PM, Greg Smith wrote:
> 
> !         Since the data is written out to disk at every transaction
> commit,
> !         the setting many only need to be be large enough to hold the
> amount
> !         of WAL data generated by one typical transaction.  Larger values,
> !         typically at least a few megabytes, can improve write performance
> !         on a busy server where many clients are committing at once.
> !         Extremely large settings are unlikely to provide additional
> benefit.

I think we can be more specific on that last sentence; is there even any
*theoretical* benefit to settings above 16MB, the size of a WAL segment?Certainly there have been no test results to
showany.
 

If we don't know, keep it vague, but otherwise I suggest:

"Settings larger than the size of a single WAL segment (16MB by default)
are unlikely to produce any benefit."

--                                  -- Josh Berkus                                    PostgreSQL Experts Inc.
                        http://www.pgexperts.com
 


pgsql-hackers by date:

Previous
From: Dimitri Fontaine
Date:
Subject: Re: Include WAL in base backup
Next
From: Alex Hunsaker
Date:
Subject: Re: arrays as pl/perl input arguments [PATCH]