From owner-freebsd-scsi Sun Jan 30 21:24:19 2000 Delivered-To: freebsd-scsi@freebsd.org Received: from freebie.lemis.com (freebie.lemis.com [192.109.197.137]) by hub.freebsd.org (Postfix) with ESMTP id EB29114E51 for ; Sun, 30 Jan 2000 21:24:13 -0800 (PST) (envelope-from grog@freebie.lemis.com) Received: (from grog@localhost) by freebie.lemis.com (8.9.3/8.9.0) id PAA69044; Mon, 31 Jan 2000 15:53:56 +1030 (CST) Date: Mon, 31 Jan 2000 15:53:56 +1030 From: Greg Lehey To: Tom Cc: Marc Tardif , freebsd-scsi@FreeBSD.ORG Subject: Re: hardware vs software stripping Message-ID: <20000131155356.B68925@freebie.lemis.com> References: <20000131120326.D62824@freebie.lemis.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 1.0pre2i In-Reply-To: WWW-Home-Page: http://www.lemis.com/~grog X-PGP-Fingerprint: 6B 7B C3 8C 61 CD 54 AF 13 24 52 F8 6D A4 95 EF Organization: LEMIS, PO Box 460, Echunga SA 5153, Australia Phone: +61-8-8388-8286 Fax: +61-8-8388-8725 Mobile: +61-41-739-7062 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Sunday, 30 January 2000 at 21:17:10 -0800, Tom wrote: > > On Mon, 31 Jan 2000, Greg Lehey wrote: > >> On Sunday, 30 January 2000 at 20:04:18 +0000, Marc Tardif wrote: >>> On Mon, 31 Jan 2000, Greg Lehey wrote: >>> >>>> On Sunday, 30 January 2000 at 14:24:54 +0000, Marc Tardif wrote: >>>>> ... >>>>> using postgresql. The alternatives are either raid-1 which seems too >>>>> wasteful on disks or raid-5 which provides fault tolerance. This last >>>>> option could substituted for a tape backup and the possibility of a few >>>>> minutes down time in case of disk failure. >>>> >>>> I'm not sure I understand this sentence. Are you planning to forget >>>> RAID-5 after all and use a tape backup? For reasonably large disks, >>>> your downtime will be measured in hours, not minutes. For a RAID-5 >>>> array, you shouldn't get any down time. >>> >>> You understood correctly, but I guess you're right. From reading the >>> dpt.com website, hardware failures are caused by hard-drives 50% of the >>> time. Also, from one of Simon Shapiro's posting to this mailing list, I >>> could build a raid 0+5 array which would seem to be an optimal solution >>> for a database performing random reads and writes. Therefore, I should >>> probably forget simply using ccd or vinum for a production system. >> >> I don't know how you conclude that. First, the DPT probably won't buy >> you anything in terms of performance, and secondly it's out of >> production. > > Well, since I still can order the DPT 3334, I think its demise is > greatly exagerated... You can order them. Others have done that, but some have reported non-delivery. > as far as performance goes, it is not a new design and can't keep > the new disks busy. Right. > Second, a DPT can't do RAID 5 + 0. It does the RAID 5 in > hardware. The RAID 5 you'll have to do in software. Hmm. There's obviously a typo here, but I'm not sure how to correct it. I don't know the DPT implementation, but in Vinum RAID-5 implies striping. I'd be interested in hearing how DPT does it. >>> There is another problem though, which is that I can't really have a >>> general idea of the amount of space I'll require. Therefore, from my >>> understanding, if I need to expand an array containing an sql db, I'll >>> need to rebuild the whole thing after recreating a new filesystem on the >>> new array. I'd be very relieved if there was an easier way... >> >> I can't make qualified statements about SQL. > > Depends on the database. Many databases allow the additional of storage > on any device, so there shouldn't be a problem. > > An Infortrend (see below) can do a transparent copy-and-replace array > expansion. This however just leaves you with a bigger virtual disk. > FreeBSD has no way to grow a filesystem transparently. You can disklabel > the addtional space and make a new filesystem though. In a similar way, you can increase the size of a concatenated Vinum plex, or add another, larger plex to a volume. Both would increase the size of a volume. Some people also have tools for increasing the size of a ufs file system, but they still need work. >> That's my understanding. Even if you can get one, the performance is >> disappointing. In addition, I don't think you can't access the >> on-board management software from FreeBSD. > > I don't think you can access the on-board management of any of the HBA > RAID cards under FreeBSD. > > A SCSI-SCSI RAID controller (like a Infortrend or Mylex) is pretty nice. > You can manage them. The Infortrend support RAID 5 + 0 in hardware. > > As far as vinum goes, I don't see how I can use to make a mirrored > system disk (root and swap). I do :-) See http://www.lemis.com/vinum/wishlist.html for a discussion. > I don't know if that will even be possible on x86 architecture. Yes, it's possible. I'm looking at how to do it right now. Greg -- Finger grog@lemis.com for PGP public key See complete headers for address and phone numbers To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message