Date: Mon, 12 Feb 2007 21:29:47 +1030 From: "Daniel O'Connor" <doconnor@gsoft.com.au> To: Peter Jeremy <peterjeremy@optushome.com.au> Cc: freebsd-stable@freebsd.org Subject: Re: What is a good choice of sata-ii raid controller for freebsd? Message-ID: <200702122129.48268.doconnor@gsoft.com.au> In-Reply-To: <20070212083748.GA837@turion.vk2pj.dyndns.org> References: <00ad01c74b65$79db1710$0c00a8c0@Artem> <200702121607.05427.doconnor@gsoft.com.au> <20070212083748.GA837@turion.vk2pj.dyndns.org>
next in thread | previous in thread | raw e-mail | index | archive | help
--nextPart5094056.ED1RJJEulh Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline On Monday 12 February 2007 19:07, Peter Jeremy wrote: > On 2007-Feb-12 16:07:03 +1030, Daniel O'Connor <doconnor@gsoft.com.au>=20 wrote: > >I regularly ship systems overseas where the power fails frequently. The > >inability to boot because one disk got hosed is Bad News (tm). > > A decent UPS can help here. We ship good UPSs but our customers aren't made of gold so the battery runs= =20 out :) > >It depends on your exact situation, I was just pointing out that SW RAID > >doesn't cover all the bases HW RAID does. > > If the disk is dead then the BIOS will skip it and the system should > boot normally (I've tested this by pulling a disk since I didn't have > a suitable dead disk to hand). A hard error in the 2nd stage boot > loader, ficl or the kernel is definitely the worst case - I agree that > this is very difficult for software raid to recover from. It's not so much SW RAID per se, it's the BIOS that's the weak link. Also, SW RAID would prevent you (safely) writing from the loader, however t= his=20 is rare/nonexistant so it isn't a big problem. Unfortunately disks dropping totally dead is reasonably rare (in my=20 experience). What is much more like is a progressive failure where more and= =20 more of the disk is rendered unreadable. > Note that even with hardware raid, there are still lots of failure > points. The least reliable parts of a current computer are the CPU > and PSU fans, not the disks. Amazing when you think the CPU has no moving parts ;) =2D-=20 Daniel O'Connor software and network engineer for Genesis Software - http://www.gsoft.com.au "The nice thing about standards is that there are so many of them to choose from." -- Andrew Tanenbaum GPG Fingerprint - 5596 B766 97C0 0E94 4347 295E E593 DC20 7B3F CE8C --nextPart5094056.ED1RJJEulh Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (FreeBSD) iD8DBQBF0Eik5ZPcIHs/zowRAqfXAJ9US1LzTyuUk18qXGI1E8vnEqINCgCfRrgb 618E4PpKL7qBE3pBQcKK9mQ= =vAIs -----END PGP SIGNATURE----- --nextPart5094056.ED1RJJEulh--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200702122129.48268.doconnor>