From owner-freebsd-questions Tue Oct 24 01:28:02 1995 Return-Path: owner-questions Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id BAA27025 for questions-outgoing; Tue, 24 Oct 1995 01:28:02 -0700 Received: from ns.dknet.dk (uucp@ns.dknet.dk [193.88.44.42]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id BAA27001 for ; Tue, 24 Oct 1995 01:27:49 -0700 Received: from pingnet (uucp@localhost) by ns.dknet.dk (8.6.12/8.6.12) with UUCP id JAA15807 for freebsd-questions@freefall.cdrom.com; Tue, 24 Oct 1995 09:27:22 +0100 Received: from terry by ic1.ic.dk with UUCP id AA00327 (5.65c8/IDA-1.4.4j for freebsd-questions@freefall.cdrom.com); Tue, 24 Oct 1995 09:13:41 +0100 Received: from hda.com (fax@localhost) by terry.terry.ping.dk (8.6.12/8.6.12) with UUCP id JAA14212; Tue, 24 Oct 1995 09:01:29 +0100 Received: from dkuug by ic1.ic.dk with UUCP id AA22220 (5.65c8/IDA-1.4.4j); Mon, 23 Oct 1995 22:33:32 +0100 Received: from hda.com (hda.com [199.232.40.182]) by ns.dknet.dk (8.6.12/8.6.12) with ESMTP id VAA09650; Mon, 23 Oct 1995 21:19:21 +0100 Received: (from dufault@localhost) by hda.com (8.6.11/8.6.9) id MAA27114; Mon, 23 Oct 1995 12:24:18 -0400 From: Peter Dufault Message-Id: <199510231624.MAA27114@hda.com> Subject: Re: Troubles dumping SCSI-disk To: hvd@terry.ping.dk (Henrik Vestergaard Draboel) Date: Mon, 23 Oct 1995 12:24:17 -0400 (EDT) Cc: freebsd-questions@terry.ping.dk In-Reply-To: <199510221951.UAA01267@terry.terry.ping.dk> from "Henrik Vestergaard Draboel" at Oct 22, 95 08:51:51 pm X-Mailer: ELM [version 2.4 PL24] Content-Type: text Content-Length: 1622 X-Charset: Latin1 X-Char-Esc: 29 Sender: owner-questions@FreeBSD.org Precedence: bulk > I have 3 scsi disks in my system. One of them is a IBM-spitfire, 1gb > fast-scsi-2, connected to my ahc-2940w. > > When I try to dump(8) this disk it always fails, with a lot of > messages logged to the console (log/messages attached). > > My other 2 scsi- and 3 ide-disks dump(8) with out problems! > > "dd if=/dev/rsd1a of=/dev/null" and "iozone" run ok, while "tar cf > /dev/null /usr2" also fails, when the disk is close-to full. > > Can anybody help me with this? ... > Oct 22 20:31:50 terry /kernel: sd1(ahc0:4:0): NOT READY asc:4,1 > Oct 22 20:31:50 terry /kernel: sd1(ahc0:4:0): Logical unit is in process of becoming ready > Oct 22 20:31:50 terry /kernel: , retries:2 This is a problem in the current SCSI driver where it doesn't delay at all between the detection of the error and the retry. I don't know why your disk is "in the process of becoming ready", but we should probably give it at least a few milliseconds to come ready. In general, if something is "becoming ready" we should probably keep retrying for a long time. As a hack, and to at least let you do dumps and verify that this is the problem, if you have the source and can rebuild the kernel it isn't hard to make "asc:4,1" infinitely retried. Unfortunately, it is a lot harder to do something more intelligent. Note that this will allow your system to hang forever doing disk retries. If you want those really stupid patches ask me and I'll send them to you. -- Peter Dufault Real Time Machine Control and Simulation HD Associates, Inc. Voice: 508 433 6936 dufault@hda.com Fax: 508 433 5267