From owner-freebsd-stable Fri Oct 30 08:57:26 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id IAA00267 for freebsd-stable-outgoing; Fri, 30 Oct 1998 08:57:26 -0800 (PST) (envelope-from owner-freebsd-stable@FreeBSD.ORG) Received: from pop.uniserve.com (pop.uniserve.com [204.244.156.3]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id IAA00261 for ; Fri, 30 Oct 1998 08:57:24 -0800 (PST) (envelope-from tom@uniserve.com) Received: from shell.uniserve.ca [204.244.186.218] by pop.uniserve.com with smtp (Exim 1.82 #4) id 0zZHrU-0006UL-00; Fri, 30 Oct 1998 08:57:20 -0800 Date: Fri, 30 Oct 1998 08:57:19 -0800 (PST) From: Tom X-Sender: tom@shell.uniserve.ca To: "Milliken, Scott" cc: "'stable@freebsd.org'" Subject: RE: RAID support in FBSD? In-Reply-To: <7B62F9E0DD56D111AADB006097A52FCC0465EA@STIUSATLCX1.salestech.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Fri, 30 Oct 1998, Milliken, Scott wrote: > > Besides, a host based RAID controler takes RAID processor "offline" > > too. > > A DPT PM334 has a dedicated CPU and RAM, it just happens to be on a > > PCI > > card. > > > You're also limited by the speed of the PCI bus. If you're > running a brand-spanking new motherboard with 100 MHz PCI then you > *might* be able to come close to the performance of a lower end offline > RAID (or SCSI-SCSI). Another advantage of offline RAID that wasn't And how are SCSI-SCS boxes not "limited" by PCI? It seems that SCSI-SCSI units are more limited, because they are connected via a 40MB/s (or 80MB/s with ultra2) bus, but host based cards are right on the PCI bus (132MB/s). A low end SCSI-SCSI unit will not exceed the performance of a host base one. Not with the extra overhead. > mentioned and that an in-box RAID has a liability for is power loss. If > your power supply goes out in the CPU then you can have corrupted data > on the RAID, while if you lose power on the CPU with an offline box, it > has a separate power supply. Most quality offline RAIDs have built in > battery backup units and redundant power supplies to make sure that this Two power supplies are worse than one. It only takes one to fail to take down the system, so system MTBF is the MTBF of one power supply divided by two. Besides, if you want the system to be any good at all, you'll put redundant power supplies everywhere. > doesn't happen - if the power is lost to the unit it flushes the cache > and parks the heads on the drives. This has come in quite handy on > more than one occasion (like when a tornado passed within 100 yards of > the data center and ripped out all the power lines) when the UPS > couldn't keep going indefinitely. Then you server should shutdown if it has no power. That's what smart UPSes are for. > I'm not trying to "slam" in-box RAID solutions, but if you're > going to go - go all out. (To steal from the Hardee's commercial). If > you can't afford an offline RAID solution, in-box is definitely better > than no RAID at all. A SCSI-SCSI also makes no sense for small 3 or 5 drive arrays either. > Scott A. Milliken > IMS Health Strategic Technologies > Systems Integration Group > Atlanta, GA Tom To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message