Re: data on devel code perf dip - Mailing list pgsql-hackers

From Bruce Momjian
Subject Re: data on devel code perf dip
Date
Msg-id 200508120201.j7C21LX00128@candle.pha.pa.us
Whole thread Raw
In response to Re: data on devel code perf dip  (Andrew - Supernews <andrew+nonews@supernews.com>)
List pgsql-hackers
Andrew - Supernews wrote:
> On 2005-08-12, Alvaro Herrera <alvherre@alvh.no-ip.org> wrote:
> > In light of this, may I ask whether it makes sense to compare the
> > performance of two runs with similar shared_buffer settings?  With
> > O_DIRECT, I understand from this manpage that the OS is going to do
> > little or no page caching, so shared_buffers should be increased to
> > account for this fact.
> >
> > Am I missing something?
> 
> O_DIRECT should only be being used for the WAL, not for buffer I/O.

And only when open_sync or open_datasync are used.  But the group write
patch is used in all cases, but again only WAL.

--  Bruce Momjian                        |  http://candle.pha.pa.us pgman@candle.pha.pa.us               |  (610)
359-1001+  If your life is a hard drive,     |  13 Roberts Road +  Christ can be your backup.        |  Newtown Square,
Pennsylvania19073
 


pgsql-hackers by date:

Previous
From: ITAGAKI Takahiro
Date:
Subject: Why do index access methods use LP_DELETE?
Next
From: Tom Lane
Date:
Subject: Re: data on devel code perf dip