From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from pigeon.gentoo.org ([208.92.234.80] helo=lists.gentoo.org) by finch.gentoo.org with esmtp (Exim 4.60) (envelope-from ) id 1Q7tMc-0000sQ-JZ for garchives@archives.gentoo.org; Thu, 07 Apr 2011 17:51:26 +0000 Received: from pigeon.gentoo.org (localhost [127.0.0.1]) by pigeon.gentoo.org (Postfix) with SMTP id A69641C004; Thu, 7 Apr 2011 17:50:00 +0000 (UTC) Received: from smtpq2.tb.mail.iss.as9143.net (smtpq2.tb.mail.iss.as9143.net [212.54.42.165]) by pigeon.gentoo.org (Postfix) with ESMTP id 52A5B1C004 for ; Thu, 7 Apr 2011 17:50:00 +0000 (UTC) Received: from [212.54.42.136] (helo=smtp5.tb.mail.iss.as9143.net) by smtpq2.tb.mail.iss.as9143.net with esmtp (Exim 4.71) (envelope-from ) id 1Q7tLD-0003xn-M0 for gentoo-user@lists.gentoo.org; Thu, 07 Apr 2011 19:49:59 +0200 Received: from 5353c7ed.cm-6-4d.dynamic.ziggo.nl ([83.83.199.237] helo=data.antarean.org) by smtp5.tb.mail.iss.as9143.net with esmtp (Exim 4.71) (envelope-from ) id 1Q7tLA-00024u-Dk for gentoo-user@lists.gentoo.org; Thu, 07 Apr 2011 19:49:56 +0200 Received: from localhost (localhost [127.0.0.1]) by data.antarean.org (Postfix) with ESMTP id DD58027AC for ; Thu, 7 Apr 2011 19:50:08 +0200 (CEST) X-Virus-Scanned: amavisd-new at antarean.org Received: from data.antarean.org ([127.0.0.1]) by localhost (data.antarean.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id uS9yuiYpLq3B for ; Thu, 7 Apr 2011 19:50:08 +0200 (CEST) Received: from www.antarean.org (net.antarean.org [10.10.11.5]) by data.antarean.org (Postfix) with ESMTP id C5D051391 for ; Thu, 7 Apr 2011 19:50:07 +0200 (CEST) Received: from 10.10.11.1 (SquirrelMail authenticated user joost) by www.antarean.org with HTTP; Thu, 7 Apr 2011 19:50:07 +0200 Message-ID: <319ad2ea44eec3e9f8794ef5c49ae633.squirrel@www.antarean.org> In-Reply-To: <646311.68321.qm@web39304.mail.mud.yahoo.com> References: <4D9D9071.2050504@gmail.com> <20110407133040.2B2D327AC@data.antarean.org> <109121.95859.qm@web39320.mail.mud.yahoo.com> <20110407141444.BCB771391@data.antarean.org> <646311.68321.qm@web39304.mail.mud.yahoo.com> Date: Thu, 7 Apr 2011 19:50:07 +0200 Subject: Re: [gentoo-user] LVM for data drives but not the OS From: "J. Roeleveld" To: gentoo-user@lists.gentoo.org User-Agent: SquirrelMail/1.4.21 Precedence: bulk List-Post: List-Help: List-Unsubscribe: List-Subscribe: List-Id: Gentoo Linux mail X-BeenThere: gentoo-user@lists.gentoo.org Reply-to: gentoo-user@lists.gentoo.org MIME-Version: 1.0 Content-Type: text/plain;charset=iso-8859-1 X-Priority: 3 (Normal) Importance: Normal Content-Transfer-Encoding: quoted-printable X-ZiggoSMTP-MailScanner-Information: Please contact the ISP for more information X-ZiggoSMTP-MailScanner-ID: 1Q7tLA-00024u-Dk X-ZiggoSMTP-MailScanner: Found to be clean X-ZiggoSMTP-MailScanner-SpamCheck: geen spam, SpamAssassin (niet cached, score=0.472, vereist 5, BAYES_05 -0.50, RDNS_DYNAMIC 0.98, T_RP_MATCHES_RCVD -0.01) X-ZiggoSMTP-MailScanner-From: joost@antarean.org X-Spam-Status: No X-Archives-Salt: X-Archives-Hash: f83795bd62074124917ca03133cedc8f On Thu, April 7, 2011 7:31 pm, BRM wrote: > ----- Original Message ---- > >> From: Joost Roeleveld >> On Thursday 07 April 2011 06:52:26 BRM wrote: >> > ----- Original Message ---- >> > >> > > From: Joost Roeleveld >> > > >> > > On Thursday 07 April 2011 06:20:55 BRM wrote: >> > > > ----- Original Message ---- >> > > > >> > > > > From: Neil Bothwick >> > > > > >> > > > > On Thu, 07 Apr 2011 05:22:41 -0500, Dale wrote: >> > > > > > I want to do it this way because I don't trust LVM enough >> > > > > > to put my >> > > > > > >> > > > > > OS >> > > > > > on. Just my personal opinion on LVM. >> > > > > >> > > > > This doesn't make sense. Your OS can be reinstalled in an >> hour >> > > > > or two, your photos etc. are irreplaceable. >> > > > >> > > > Makes perfect sense to me as well. >> > > > >> > > > Having installed LVM - and then removed it due to issues; namel= y, >> > > > the fact that one of the hard drives died taking out the whole >> LVM >> > > > group, leaving the OS unbootable, and not easily fixable. Ther= e >> > > > was a thread on that (started by me) a while back (over a >> year). >> > > > >> > > > So, perhaps if I had a RAID to underly so I could mirror drive= s >> > > > under LVM >> > > > >> > > > for recovery I'd move to it again. But otherwise it is just a >> PITA >> > > > waiting >> > > > >> > > > to happen. >> > > > >> > > > Ben >> > > >> > > Unfortunately, any method that spreads a filesystem over multipl= e >> disks >> > > can be >> > > >> > > affected if one of those disks dies unless there is some mechani= sm >> in >> > > place that can handle the loss of a disk. >> > > For that, RAID (with the exception of striping, eg. RAID-0) >> provides >> > > that. >> > > >> > > Just out of curiousity, as I never had the need to look into thi= s, >> I >> > > think that, in theory, it should be possible to recover data fro= m >> LVs >> > > that were not >> > > >> > > using the failed drive. Is this assumption correct or wrong? >> > >> > If you have the LV configuration information, then yes. Since I >> managed to >> > find the configuration information, I was able to remove the affect= ed >> PVs >> > from the VG, and get it back up. >> > I might still have it running, but I'll back it out on the next >> rebuild - > or >> > if I have a drive large enough to do so with in the future. I was >> wanting >> > to use LVM as a bit of a software RAID, but never quite got >> > that far in the configuration before it failed. It does do a good j= ob >> at >> > what it's designed for, but I would not trust the OS to it either >> since the >> > LVM configuration is very important to keep around. >> > >> > If not, good luck as far as I can tell. >> > >> > Ben >> >> LVM isn't actually RAID. Not in the sense that one gets redundancy. I= f >> you >> consider it to be a flexible partitioning method, that can span >> multiple >>disks, >> >> then yes. >> But when spanning multiple disks, it will simply act like JBOD or >> RAID0. >> Neither protects someone from a single disk failure. >> >> On critical systems, I tend to use: >> DISK <-> RAID <-> LVM <-> Filesystem >> >> The disks are as reliable as Google says they are. They fail or they >> don't. >> RAID protects against single disk-failure >> LVM makes the partitioning flexible >> Filesystems are picked depending on what I use the partition for >> > > The attraction to LVM for me was that from what I could tell it support= ed > and > implemented a software-RAID > so that I could help protect from disk-failure. I never got around to > configuring that side of it, but that was my goal. > Or are you saying I was misunderstanding and LVM _does not_ contain > software-RAID support? Unless I am mistaken, LVM does not provide redundancy. It provides disk-spanning (JBOD) and basic striping (RAID-0). For redundancy, I would use a proper RAID (either hardware or software). On top of this, you can then decide to have a single filesystem, LVM or even partition this. I think the confusion might have come from the fact that both LVM and Linux Software Raid use the "Device Mapper" interface in the kernel confi= g and they are in the same part. Also, part of the problem is that striping is also called RAID-0. That, t= o people who don't fully understand it yet, makes it sound like it is a RAID. It actually isn't as it doesn't provide any redundancy. I do hope you didn't loose too much important data when you had this issu= e. -- Joost