Date: Thu, 18 Oct 2007 02:15:31 -0700 From: Jeremy Chadwick <koitsu@FreeBSD.org> To: Stefan Lambrev <stefan.lambrev@moneybookers.com> Cc: freebsd-current@freebsd.org Subject: Re: Very slow HP 8 Internal Port SAS HBA with RAID 0, 1 on Hot Plug SAS/SATA Models (LSI SAS1068 - mpt) Message-ID: <20071018091531.GA8737@eos.sc1.parodius.com> In-Reply-To: <471722BE.20107@moneybookers.com> References: <47161A42.2050306@moneybookers.com> <471636E4.90807@thekeelecentre.com> <47163ACC.1080807@moneybookers.com> <47163B9B.4010901@thekeelecentre.com> <471715AD.3080102@moneybookers.com> <47171750.8050906@thekeelecentre.com> <20071018085812.GA8481@eos.sc1.parodius.com> <471722BE.20107@moneybookers.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Oct 18, 2007 at 12:09:18PM +0300, Stefan Lambrev wrote: > May be I have wrong impression that mirror should increase read speed > almost twice compared to > single disk, but I forget that bonnie++ was run as single thread. > May be if I start it with concurrency=2 then I'll see the real performance > of the mirror ? > Or should I move to the thread to performance@ :) The easiest way for me to describe it would be to refer you to the Wikipedia article. See the section "Other benefits of mirroring": http://en.wikipedia.org/wiki/Disk_mirror Generally speaking, RAID 1 sequential read across 2 disks should be slightly higher than the same disk in a JBOD configuration -- keyword is "slightly". RAID 0 (striping) is what gives you the immense improvement, but I'm sure you know how risky RAID 0 is... -- | Jeremy Chadwick jdc at parodius.com | | Parodius Networking http://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, USA | | Making life hard for others since 1977. PGP: 4BD6C0CB |
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20071018091531.GA8737>