Re: posix_fadvise v22

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Greg Smith <gsmith(at)gregsmith(dot)com>
Cc: Robert Haas <robertmhaas(at)gmail(dot)com>, Gregory Stark <stark(at)enterprisedb(dot)com>, ITAGAKI Takahiro <itagaki(dot)takahiro(at)oss(dot)ntt(dot)co(dot)jp>, Postgres <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: posix_fadvise v22
Date: 2009-01-02 15:33:25
Message-ID: 24481.1230910405@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Greg Smith <gsmith(at)gregsmith(dot)com> writes:
> On Thu, 1 Jan 2009, Robert Haas wrote:
>> The only thing I haven't been able to do is demonstrate that this change
>> actually produces a performance improvement. Either I'm testing the
>> wrong thing, or it just doesn't provide any benefit on a single-spindle
>> system.

> When I did a round of testing on the earlier prefetch test program Greg
> Stark put together, one of my single-spindle Linux system didn't show any
> real benefit. So as long as you didn't see performance degrade, your not
> seeing any improvement isn't bad news.

ISTM that you *should* be able to see an improvement on even
single-spindle systems, due to better overlapping of CPU and I/O effort.
If the test case is either 100% CPU-bound or 100% I/O-bound then no,
but for anything in between there ought to be improvement.

regards, tom lane

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Kevin Grittner 2009-01-02 15:38:52 Re: Documenting serializable vs snapshot isolation levels
Previous Message Tom Lane 2009-01-02 15:20:20 Re: Including kerberos realm