From mboxrd@z Thu Jan 1 00:00:00 1970 From: "mel" Subject: Promise UltraTrak RM15000 + Adaptec 39160 Date: Fri, 8 Aug 2003 12:36:21 +0200 Sender: linux-scsi-owner@vger.kernel.org Message-ID: <20030808103621.M63039@mmtab.se> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Return-path: Received: from [212.209.150.83] ([212.209.150.83]:2956 "EHLO mail.mmtab.se") by vger.kernel.org with ESMTP id S271116AbTHHKi7 (ORCPT ); Fri, 8 Aug 2003 06:38:59 -0400 Received: from mmtab.se (localhost.localdomain [127.0.0.1]) by mail.mmtab.se (8.12.9/8.12.9) with ESMTP id h78AaLMw028336 for ; Fri, 8 Aug 2003 12:36:21 +0200 List-Id: linux-scsi@vger.kernel.org To: linux-scsi@vger.kernel.org The story so far: I've been running an out of the box RH 7.3 with 2.4.18-3 kernel together with the 39160 card and a UltraTrak RM8000 for over 12 months. The only change made to the kernel was that I compiled the aic7xxx driver into the kernel. The server is running nfs and samba only. The RM8000 is equipped with eight 80GB IBM drives and exports about 0.5TB. The mobo is a Kinetiz 7E-A with a 1GHz AMD and 512MB of RAM. When we run out of disk i invested in a RM15000 and 15 120GB IBM-drives. I figured it was time to upgrade kernel and drivers and therefore I got the 2.4.21 kernel together with the latest driver ( aic79xx-linux-2.4-20030603- tar.gz ) from Justin Gibbs site http://people.freebsd.org/~gibbs/linux/ After installation I tried to partition and makefs the new drive. I got a lot of parity errors and "trying to queue an ABORT message" from the card. I did some 'Googling' but could not find the answer. As the server is in production I assembled a lab machine and hooked the RM15000 to that machine. I did a RH7.3 install and had it to autodetect my gear ( aic7xxx as a module ). As a surprise everything worked without a glitch with this setup. I tried to use the 2.4.21 kernel with the new driver ( as a module ) on this machine but with this it malfunctioned in the same way as before. Is there anyone that have a clue to this behaviour? Shouldn't the never version be better ;) /Mel -- This sentence is false.