From owner-freebsd-hardware Wed Nov 29 3:21:53 2000 Delivered-To: freebsd-hardware@freebsd.org Received: from mail-internal.nextra.de (mail.nextra.de [212.169.184.81]) by hub.freebsd.org (Postfix) with ESMTP id A286E37B401 for ; Wed, 29 Nov 2000 03:21:49 -0800 (PST) Received: from omnilinkw63 (f-euro.fw.nextra.de [212.169.184.9]) by mail-internal.nextra.de (8.9.3/8.9.3) with SMTP id MAA00964 for ; Wed, 29 Nov 2000 12:21:42 +0100 (CET) Message-ID: <000a01c059f6$94b32850$6560310a@omnilinkw63> From: "Oliver Blasnik" To: Subject: Again: Promise IDE RAID / Fasttrack100 Date: Wed, 29 Nov 2000 12:16:50 +0100 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2919.6700 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6700 Sender: owner-freebsd-hardware@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi there, we just got a very nice engineering sample from gigabyte (1U) with an embedded FastTrack100 Controller on-board. (For anyone who wants to know: 1U, two PPGA370 Sockets, two on-board ethernet (fxp), two hotswap ide ata100-drives, usb, serial, parallel. Work is on a next 1U with up to 3 drives and ipmi on board.). I gave 4.2release a try as far as I recognised that there was no support for this controller in previous versions. Detection was sucessfully, installing without any problem, looked really good to me. So, what is the first to check out with an mirrored raid? Take one of the drives out of the bay and look what happens. I did and looked... Hum... System stopped... Waiting... System still hangs... Put the drive back into bay. Nothing happens at all. Not the behaviour of an raid array I expected. Linux didn't recognise that controller at all, the shipped driver for w2k worked as expected -> system was running with an degraded raid. So, what about the state of the drivers for FreeBSD? Looking into source gave me the opinion that the two drives were handled as to independend ones, just mirroring/striping algos added. No special failure handling if connected via an FastTrack and mirrored. Using the striping really rocks - 53MB/sec write performance with two IDE 30GB drives. I asked Gigabyte for help and they told me that Promise will release their drivers for Linux and other Intel-Unixes in GPL at Dec., 20th. What a nice xmas present ;-) Hopefully anyone is going to implement that new drivers into FreeBSD or fix that problem. Cu, Oliver -- -- http://www.nextra.de - INTERNET@WORK ----- oliver.blasnik@nextra.de -- Nextra Deutschland | Oliver Blasnik Senior System Administrator GmbH & Co KG | Lyoner Strasse 26 D-60528 Frankfurt Region Mitte | tel +49-69-66441-0 fax +49-69-66441-199 ------------------------------------------------------------------------- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hardware" in the body of the message