From owner-freebsd-hackers Thu Mar 5 06:57:01 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id GAA23375 for freebsd-hackers-outgoing; Thu, 5 Mar 1998 06:57:01 -0800 (PST) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from Kitten.mcs.com (Kitten.mcs.com [192.160.127.90]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id GAA23361 for ; Thu, 5 Mar 1998 06:56:55 -0800 (PST) (envelope-from karl@Mars.mcs.net) Received: from Mars.mcs.net (karl@Mars.mcs.net [192.160.127.85]) by Kitten.mcs.com (8.8.7/8.8.2) with ESMTP id IAA25404; Thu, 5 Mar 1998 08:56:55 -0600 (CST) Received: (from karl@localhost) by Mars.mcs.net (8.8.7/8.8.2) id IAA06363; Thu, 5 Mar 1998 08:56:54 -0600 (CST) Message-ID: <19980305085654.20326@mcs.net> Date: Thu, 5 Mar 1998 08:56:54 -0600 From: Karl Denninger To: "Justin T. Gibbs" Cc: hackers@FreeBSD.ORG Subject: Re: SCSI Bus redundancy... References: <19980303200652.07366@mcs.net> <199803050710.AAA17706@narnia.plutotech.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.84 In-Reply-To: <199803050710.AAA17706@narnia.plutotech.com>; from Justin T. Gibbs on Thu, Mar 05, 1998 at 12:10:14AM -0700 Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Thu, Mar 05, 1998 at 12:10:14AM -0700, Justin T. Gibbs wrote: > > The best I've seen off our RAID systems right now is about 11MB/sec (that's > > megaBYTES, not bits). That's on an Ultra bus, with 2 ultra busses going to > > the RAID disks. > > > > Neither the disk buses nor the RAID controller CPU are saturated. I > > believe this is pretty much the wall on one SCSI channel, at least with > > 16 SCBs. I'm going to try it with SCBPAGING turned on and see if that > > helps, but for sequential reads it probably won't matter much. > > The problem is that the largest I/O you can send to your RAID box is > 64k which it will stripe over N disks. As soon as you drop to 16 or > 8k a disk per transaction, you will never be able to saturate the system. > This is one of the reasons CCD performs so much better than a RAID 0 > external box. CCD can perform up to a 64k transaction per disk. > We really need to get buffer chaining into the kernel so we can get past > this silly 64k I/O barrier. > > > I suspect the bottleneck is in the AIC code at this point, or the bus > > itself, or the interrupt latency on the DMA completion is killing me. > > There is no appreciable difference between running at 40MB/sec (ultra > > full-bore) and 20MB/sec, indicating that perhaps the hold-up is in the > > Adaptec microcode, driver, and/or the Adaptec/PCI bus interface. > > You need to up the number of transactions handled in parallel. The ahc > driver has no problem saturating the SCSI bus, if you feed it enough > to do. Will going to paged SCBs do this? Some disks have had trouble with this in the past, which is why I run it disabled right now. But with the CMD controllers in the loop now, that's no longer a factor. -- -- Karl Denninger (karl@MCS.Net)| MCSNet - Serving Chicagoland and Wisconsin http://www.mcs.net/ | T1's from $600 monthly to FULL DS-3 Service | NEW! K56Flex support on ALL modems Voice: [+1 312 803-MCS1 x219]| EXCLUSIVE NEW FEATURE ON ALL PERSONAL ACCOUNTS Fax: [+1 312 803-4929] | *SPAMBLOCK* Technology now included at no cost To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message