Date: Tue, 18 Jul 2006 14:46:57 -0700 (PDT) From: "R. B. Riddick" <arne_woerner@yahoo.com> To: Alexandre Biancalana <biancalana@gmail.com>, freebsd-geom@freebsd.org Subject: Re: geom_stripe error Message-ID: <20060718214657.63152.qmail@web30303.mail.mud.yahoo.com> In-Reply-To: <8e10486b0607181351u1319483w87d40792d6c1cce8@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--- Alexandre Biancalana <biancalana@gmail.com> wrote: > Really is happening what you told.... the partition concat/concat1 have the > number 31, but in gstripe list it appears at position 21 position.... (see > my first email) > Hmm... Oh... I just had a closer look at the funtion "list_one_geom" in /usr/src/sbin/geom/core/geom.c and found out, that the order of the disks in the "gstripe list" output does not say so much... It is possibly the order in which the disks are added to the device... Somehow the consumers (disks) are stored in a list and in an array (just in this array the position corresponds with the disk number, but in the list no order is necessary)... So my theory was wrong... Did you try an easier and more efficient configuration in the meantime? E.g.: gconcat(gstripe(4 * 200GB), gstripe(4 * 73GB)) or so... Possibly you would have to use bsdlabel in order to re-partition your disks... And you would loose your data... But when u r done it might be much faster, because striping over partitions of the same disk is quite slow, I think... The advantage of striping (RAID0) is (if I recall correctly) that we can have higher parallelism (e. g. a RAID0 on 10 disks allows (max) 10 concurrent requests as if there was only 1 request (looking at the processing time))... -Arne __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060718214657.63152.qmail>