From owner-freebsd-questions Tue Jan 31 14:50:09 1995 Return-Path: questions-owner Received: (from root@localhost) by freefall.cdrom.com (8.6.9/8.6.6) id OAA04333 for questions-outgoing; Tue, 31 Jan 1995 14:50:09 -0800 Received: from Root.COM (implode.Root.COM [198.145.90.1]) by freefall.cdrom.com (8.6.9/8.6.6) with ESMTP id OAA04327 for ; Tue, 31 Jan 1995 14:50:06 -0800 Received: from corbin.Root.COM (corbin.Root.COM [198.145.90.18]) by Root.COM (8.6.8/8.6.5) with ESMTP id OAA11861; Tue, 31 Jan 1995 14:46:10 -0800 Received: from localhost (localhost [127.0.0.1]) by corbin.Root.COM (8.6.9/8.6.5) with SMTP id OAA00250; Tue, 31 Jan 1995 14:46:10 -0800 Message-Id: <199501312246.OAA00250@corbin.Root.COM> X-Authentication-Warning: corbin.Root.COM: Host localhost didn't use HELO protocol To: terry@cs.weber.edu (Terry Lambert) cc: alan@picard.isocor.ie (Alan Byrne), questions@FreeBSD.org, alan@buster.internet-eireann.ie, mckusick@cs.berkeley.edu Subject: Re: disk optimization messages from kernel In-reply-to: Your message of "Tue, 31 Jan 95 14:54:54 MST." <9501312154.AA28765@cs.weber.edu> From: David Greenman Reply-To: davidg@Root.COM Date: Tue, 31 Jan 1995 14:45:58 -0800 Sender: questions-owner@FreeBSD.org Precedence: bulk >> I've just recentlly upgraded to 2.0R.snap-190128. >> I upgraded due to problems with 1.1.5.1 + P90's with PCI cards (long story) >> >> While I was installing the src distribution in /usr/src I got the >> following messages on the console, should I worry about them. >> >> /usr: optimization changed from TIME to SPACE >> /usr: optimization changed from SPACE to TIME >> /usr: optimization changed from TIME to SPACE >> /usr: optimization changed from SPACE to TIME > >UFS by default will time-optimize placement of blocks to speed read access. > >When the disk goes over a certain threshold, it switches from time to space >optimization, on the theory that anything that works is better than anything >that doesn't -- that is, it is better you don't run out of disk than you be >able to read from it as fast as possible. > >When you drop back below the threshold, plus a watermark, it switches back. Actually, there is a bug in the 4.4 code that triggers the change almost gratuitously when the filesystem is far from reaching the threshold. It then flops back and forth a few times as more files are added. In other words, it's quite broken and was added to my bugs-to-fix whiteboard about a month ago. -DG