From owner-freebsd-questions@FreeBSD.ORG Sun Apr 25 22:32:05 2004 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 109D516A4CE; Sun, 25 Apr 2004 22:32:05 -0700 (PDT) Received: from depot.depot.axxiscorp.com (ip-64-32-234-4.dsl.iad.megapath.net [64.32.234.4]) by mx1.FreeBSD.org (Postfix) with ESMTP id AF19243D48; Sun, 25 Apr 2004 22:32:01 -0700 (PDT) (envelope-from RJV@WEBLINKMO.COM) content-class: urn:content-classes:message MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable X-MimeOLE: Produced By Microsoft Exchange V6.0.6487.1 Date: Mon, 26 Apr 2004 00:34:55 -0500 Message-ID: <5EEBE9C3C61D1142994C6B620C51E8470CC842@ip-64-32-234-4.dsl.iad.megapath.net> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: odd 4.9 scsi problem Thread-Index: AcQrUDPHDMUQahn2Td6n4ipX3JBYNw== From: "Richard J. Valenta" To: , Subject: odd 4.9 scsi problem X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 26 Apr 2004 05:32:05 -0000 I have an odd scsi problem - this is a DS20 with 3 Qlogic scsi cards, 0 has only the external TZ89 DLT drive, 1 has the hard drives, 2 has the internal DAT (TLZ,T1Z10)and the cdrom. If I power down all devices completely including the server (not just reboot) and then power everything up, this is the result: worx# camcontrol devlist at scbus0 target 3 lun 0 (pass0,sa0) at scbus1 target 0 lun 0 (pass1,da0) at scbus1 target 1 lun 0 (pass2,da1) at scbus1 target 2 lun 0 (pass3,da2) at scbus2 target 4 lun 0 (pass4,sa1) at scbus2 target 5 lun 0 (pass5,cd0) But an odd thing happens to the DAT after I manually reset and rescan - it changes names, worx# camcontrol reset 2 Reset of bus 2 was successful worx# camcontrol rescan 2 Re-scan of bus 2 was successful worx# camcontrol devlist at scbus0 target 3 lun 0 (pass0,sa0) at scbus1 target 0 lun 0 (pass1,da0) at scbus1 target 1 lun 0 (pass2,da1) at scbus1 target 2 lun 0 (pass3,da2) at scbus2 target 4 lun 0 (sa1,pass4) at scbus2 target 5 lun 0 (pass5,cd0) But this isn't the main problem - I just discovered this playing around after the main problem - when I try to use sa0 (the DLT), I often get an error like below: archive/backup/images/Videos/rich.mov tar (child): /dev/sa0: Wrote only 0 of 10240 bytes tar (child): Error is not recoverable: exiting now Broken pipe worx# when I try to perform the backup again it hangs and the tail of /var/log/messages shows: Apr 25 18:07:11 worx /kernel: (sa0:isp0:0:3:0): failed to write terminating filemark(s) Apr 25 18:07:11 worx /kernel: (sa0:isp0:0:3:0): tape is now frozen- use an OFFLINE, REWIND or MTEOM command to clear this state. And when I try to do so: worx# mt -f /dev/sa0 rewind mt: /dev/sa0: Device not configured and when I reboot... worx# camcontrol devlist at scbus1 target 0 lun 0 (pass0,da0) at scbus1 target 1 lun 0 (pass1,da1) at scbus1 target 2 lun 0 (pass2,da2) at scbus2 target 4 lun 0 (pass3,sa0) at scbus2 target 5 lun 0 (pass4,cd0) Odd change of CAPS for the DAT, and the DLT has disappeared... System info: single 500mhz alpha processor DS20, with 3 Qlogic scsi cards and a brand new kernel, which only includes Qlogic cards in its config, and all scsi device types as well. I had 5.0 on this machine and ran backups fine. Then changed to 4.9 trying to get sound to work, that was unsuccessful, and now I have this tape problem, I don't know if they are related. At times, if I simply let the machine sit, the DLT will just fall off the chain and a rescan/reset will not bring it back. At the same time the other tape gets renamed. I'd like to blame a card to the tape drive, but I've moved things around and had the same problems, and would think that if the DLT was fried it just wouldn't work - not work some of the time. Love to hear anyone's opinions or thoughts on repairing this. Richard=20