Date: Fri, 16 Nov 2001 15:20:57 +0100 From: Daniel Lang <dl@leo.org> To: Bill Moran <wmoran@potentialtech.com> Cc: freebsd-stable@FreeBSD.ORG Subject: Re: dirpref benefit on virtual disks Message-ID: <20011116152056.F44535@atrbg11.informatik.tu-muenchen.de> In-Reply-To: <20011113121837.A34946@atrbg11.informatik.tu-muenchen.de>; from dl@leo.org on Tue, Nov 13, 2001 at 12:18:37PM %2B0100 References: <20011112152726.A10505@atrbg11.informatik.tu-muenchen.de> <3BF013B8.8070009@potentialtech.com> <20011113121837.A34946@atrbg11.informatik.tu-muenchen.de>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi Again, Daniel Lang wrote on Tue, Nov 13, 2001 at 12:18:37PM +0100: [..] > I just did it. I've copied the FreeBSD mirror (which should be > complicated enough) onto a free partition, and then ran a find > on the original location and then on the new one. > > The find command was: > find . -print > /dev/null > > the result seemed convincing > > old dirpref: 2.97s user 67.45s system 6% cpu 17:36.22 total > new dirpref: 2.09s user 5.94s system 5% cpu 2:39.32 total > > But I cannot rule our side-effects. The system is not quiet, > and the array with the old code is much more busy than the spare > one, which is not busy at all. Anyway, I will recreate the tree on > the busy array und do the test again. I did that now, and still result is convincing: 2.44s user 7.35s system 5% cpu 3:04.66 total this is just slightly longer than on the quiet array, and still much faster than before. So I guess one can say it definitely improves things on virtual disks, too. I can only recommend to recreate the directories to anyone. Best regards, Daniel -- IRCnet: Mr-Spock - ceterum censeo Microsoftinem esse delendam - *Daniel Lang * dl@leo.org * +49 89 289 25735 * http://www.leo.org/~dl/* To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20011116152056.F44535>