From owner-freebsd-questions Wed May 13 10:38:28 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id KAA05186 for freebsd-questions-outgoing; Wed, 13 May 1998 10:38:28 -0700 (PDT) (envelope-from owner-freebsd-questions@FreeBSD.ORG) Received: from pau-amma.whistle.com (s205m64.whistle.com [207.76.205.64]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id KAA05162 for ; Wed, 13 May 1998 10:38:16 -0700 (PDT) (envelope-from dhw@whistle.com) Received: (from dhw@localhost) by pau-amma.whistle.com (8.8.7/8.8.7) id KAA27785 for freebsd-questions@freebsd.org; Wed, 13 May 1998 10:37:39 -0700 (PDT) (envelope-from dhw) Date: Wed, 13 May 1998 10:37:39 -0700 (PDT) From: David Wolfskill Message-Id: <199805131737.KAA27785@pau-amma.whistle.com> To: freebsd-questions@FreeBSD.ORG Subject: FreeBSD 2.2.6-RELEASE; SCSI device disappears??!? Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG This is rather long; sorry. I'm trying to be reasonably complete. Replaced a 0.5 GB Quantum disk with an 8.5 GB Micropolis; each at SCSI target 2 on SCSI bus 0 (ahc0): /kernel: (ahc0:2:0): "MICROP 4691WS T171" type 0 fixed SCSI 2 /kernel: sd2(ahc0:2:0): Direct-Access 8681MB (17780058 512 byte sectors) Hadn't had anything important (as it turns out) on the Quantum, so I was in no rush to move the data over. Had already set up some filesystems on the Micropolis drive (on a different FreeBSD box), where I had been exercising it (successfully). Device is Ultra-SCSI w/ an adapter (as is another drive on the same chain). Since the Quantum drive had been in the middle of the chain, I removed the termination jumper on the Micropolis drive: during my testing (on the other box), it had been the terminating device on the chain. I had had /dev/sd2b set up as a 1 GB swap on the test box; for the re- assignment, I thought it would be better to make it a filesystem, so I did a "newfs" to it, then tried to mount it: shrimp[55]% sudo newfs /dev/rsd2b /dev/rsd2b: 4194304 sectors in 1024 cylinders of 1 tracks, 4096 sectors 2048.0MB in 64 cyl groups (16 c/g, 32.00MB/g, 7680 i/g) super-block backups (for fsck -b #) at: 32, 65568, 131104, 196640, 262176, 327712, 393248, 458784, 524320, 589856, ... 3866656, 3932192, 3997728, 4063264, 4128800, shrimp[56]% sudo mount /dev/sd2b /shrimp/mntb /dev/sd2b on /shrimp/mntb: Device not configured Huh??!? Ran "sudo scsi -f /dev/rsd2.ctl -r" (alternated with several interspersed attempts to perform "sudo disklabel -r sd2"); finally got one that worked, finally(!) was able to do the mount, and fired up a Perl script I had cobbled up to create & destroy files... which ran for a while, then said: disk_tst-creator: Couldn't open /mntf/tmp/13055308 for writing: Input/output error I had been doing a "tail -f /var/log/messages" in another window all this time... no clues there. Tried doing "ls mnt?" (I had mounted the various filesystems on the disk at mount points that matched that expression)... and the "tmp" subdirectories weren't visible. :-( Then I tried again, and they were visible. [I didn't do anything. Honest!] So I fired up my Perl script again, which started up, then hung. Then I heard a voice from another cube call out, asking if the machine in question had hung. I tried firing up another window to take a look... and decided that the better part of valor was to go down to the server room & check. Just as I opened the door, I saw that the machine was completing the multi-user start-up, and it then prompted for a login. Still no clues that I can see in /var/log/messages or in the output of dmesg. So, since this is -questions, I have some: * Where might I look for clues as to why a filesystem (or the device) keeps disappearing? * What might I do to prevent that behavior (while retaining the use of the disk drive)? I look forward to suggestions.... :-( Thanks, david -- David Wolfskill dhw@whistle.com (650) 577-7158 pager: (650) 371-4621 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message