From owner-freebsd-questions Sun Oct 29 18:01:59 1995 Return-Path: owner-questions Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id SAA23146 for questions-outgoing; Sun, 29 Oct 1995 18:01:59 -0800 Received: from fountain.village.org (root@fountain.village.org [198.137.146.37]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id SAA23141 for ; Sun, 29 Oct 1995 18:01:55 -0800 Received: from localhost (localhost [127.0.0.1]) by fountain.village.org (8.6.11/8.6.6) with SMTP id TAA25336 for ; Sun, 29 Oct 1995 19:01:43 -0700 Message-Id: <199510300201.TAA25336@fountain.village.org> To: freebsd-questions@freebsd.org Subject: IDE bad block handling in FreeBSD 1.1.5 Date: Sun, 29 Oct 1995 19:01:43 -0700 From: Dieter Dworkin Muller Sender: owner-questions@freebsd.org Precedence: bulk What is the true state of affairs for bad block handling on IDE disks in 1.1.5? The bad144 and badsect man pages talk about various DEC drives, which I have, but not on my FreeBSD system ;-) The symptom I'm seeing is the system locking up in a relatively tight loop, repeatedly complaining about a timeout or error (it varies) from one of the disks (wd0 and wd1). Once this happens, nothing short of a reset will get it back. It looks like it's hitting a bad block, and refusing to believe it's bad. All kernel-based things seem to work, but nothing happens at user level (i.e., telnet gets a connection, but no prompt, caps lock light toggles, ping works, stuff like that). FWIW, I first saw this problem whenever I tried to use anything connected to an Ultrastor SCSI board while running off of the built-in IDE controller (it's a Packard Bell, and yes, I know better now). Once I took the Ultrastor out, the problem went away until about a month ago. Nothing's changed in the hardware config in at least four months, and it's been stable for at least eight months. Of course, since I have to use reset, I don't get anything useful in terms of log messages. Any suggestions would be welcome, although upgrading to 2.x would be less so (other than for this occasional difficulty, it's a stable production system). Thanks. Dworkin