Date: Wed, 7 Dec 2005 19:08:21 +0100 From: Pawel Jakub Dawidek <pjd@FreeBSD.org> To: OxY <oxy@field.hu> Cc: freebsd-geom@freebsd.org Subject: Re: geom mirror write performance problem Message-ID: <20051207180821.GL53288@garage.freebsd.pl> In-Reply-To: <003a01c5fb58$3faca4e0$0201a8c0@oxy> References: <20051207165538.GJ53288@garage.freebsd.pl> <002201c5fb56$bff37590$0201a8c0@oxy> <20051207175359.GK53288@garage.freebsd.pl> <003a01c5fb58$3faca4e0$0201a8c0@oxy>
next in thread | previous in thread | raw e-mail | index | archive | help
[-- Attachment #1 --] On Wed, Dec 07, 2005 at 07:01:29PM +0100, OxY wrote: +> it is not possible to test it..but i think it's ok.. +> it worked great, the 5.4->6.0 upgrade is the source of the problem.. Actually it is possible... You said, when only one component is connected performance is fine. You can try something like this: # gmirror remove gm0 ad2 # swapoff /dev/mirror/gm0s1b (if you have one) and then run this script: #!/bin/sh dd if=/dev/zero of=/dev/ad2s1b bs=1m count=1000 & dd if=/dev/zero of=/dev/mirror/gm0s1b bs=1m count=1000 & wait -- Pawel Jakub Dawidek http://www.wheel.pl pjd@FreeBSD.org http://www.FreeBSD.org FreeBSD committer Am I Evil? Yes, I Am! [-- Attachment #2 --] -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (FreeBSD) iD8DBQFDlyUVForvXbEpPzQRAnaJAJ4/PzZxvU47hjHOLcT/7Ns4xeR1wACff/RM clS+G7sP+s4Hq0/AlpdEmik= =1XtC -----END PGP SIGNATURE-----
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20051207180821.GL53288>