Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 30 Jun 2005 10:21:38 +1000 (EST)
From:      Bruce Evans <bde@zeta.org.au>
To:        Jeremie Le Hen <jeremie@le-hen.org>
Cc:        Michael Vince <mv@roq.com>, Steve Roome <steve@pepcross.com>, performance@freebsd.org
Subject:   Re: FreeBSD MySQL still WAY slower than Linux
Message-ID:  <20050630100814.O45160@delplex.bde.org>
In-Reply-To: <20050629094533.GA49933@obiwan.tataz.chchile.org>
References:  <20050623145041.GC64879@bibipentium.lonres.com> <42BD64F1.4080001@roq.com> <20050627134146.GA626@bibipentium.lonres.com> <42C1FA31.3050806@roq.com> <20050629094533.GA49933@obiwan.tataz.chchile.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 29 Jun 2005, Jeremie Le Hen wrote:

> Could you try mounting the filesystem where the database lives with
> the noatime option, and re-run your tests ?  IIRC from previous threads
> on this subject, Linux doesn't really honor this while FreeBSD does,
> which pulls down the performances.

I think you mean the async option.  Neither would make much differerce.

atimes are cached in ffs, so mounting with noatime only makes much
difference if lots of different files are read, which is exactly what
doesn't happen when data is in a database instead of in separate files.

Mounting with async can make a large difference, but normally makes
no difference with ffs, since ffs is normally configured with soft
updates and the async option is silently ignored in this configuration.

You could try mounting with the sync option for all OSes.  This should
make them all very slow, but the results might be interesting since it
is the only sync-related mount option that gives similar behaviour at
the fs level.  It might also show if the problem is related to syncing.

Bruce



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050630100814.O45160>