From owner-freebsd-questions Thu Apr 22 14:16: 1 1999 Delivered-To: freebsd-questions@freebsd.org Received: from cygnus.rush.net (cygnus.rush.net [209.45.245.133]) by hub.freebsd.org (Postfix) with ESMTP id 3A51E14DDA for ; Thu, 22 Apr 1999 14:15:52 -0700 (PDT) (envelope-from bright@rush.net) Received: from localhost (bright@localhost) by cygnus.rush.net (8.9.3/8.9.3) with SMTP id QAA19497; Thu, 22 Apr 1999 16:29:49 -0500 (EST) Date: Thu, 22 Apr 1999 16:29:48 -0500 (EST) From: Alfred Perlstein To: mi@aldan.algebra.com Cc: questions@FreeBSD.ORG Subject: Re: writing slower with SoftUpdates In-Reply-To: <199904212008.QAA57336@misha.cisco.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Wed, 21 Apr 1999, Mikhail Teterin wrote: > According to iozone, writing a 256M file to this empty 512M > fs works at > > 18247338 bytes/second without SoftUpdates > 16687585 bytes/second with SoftUpdates > > The results are repeatable with very little deviations. Reading speed > is about the same regardless of SoftUpdates and is around 13.9Mb/s. > > The disk is a 9Gb Cheetah LVD, attached to the LVD outlet of Adaptec's > 2940U2W. > > Another, identical Cheetah on the same SCSI cable, with another 512M > empty filesystem gives > > 18179755 b/s with SoftUpdates > 18315425 b/s without SoftUpdates > > Again, reading is not affected, but averages higher at around 18Mb/s. > > Are this results what one should expect? The only thing I can think of is that softupdates introduces longer code paths in the filesystem code for block accounting and meta data operations, for I/O intesive stuff it works great, but if your processor isn't that fast or you have benches that specifically mess with the softupdates code you'll probably get performance hits. -Alfred To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message