Description of problem: Almost a hundred lines of output in /var/log/messages during each boot. Obviously, the config file is not correct for smartd. A quick review of /etc/smartd.conf shows uncommented lines for devices that do not exist on my system, email to "admin", and other undesired settings. Recommend replacing DEVICESCAN and current contents with a generated file based on output from the hardware detection software. For reference, on my system, /dev/hda is my hard disk, /dev/hdd is the CD ROM. Version-Release number of selected component (if applicable): kernel-utils-2.4-8.31 How reproducible: Boot your system Steps to Reproduce: 1. Boot your system 2. grep smartd /var/log/messages 3. Actual results: See below for sample. Expected results: A few lines showing the devices I have, not what I don't have. Should also change if / when I add or replace hardware. Additional info: Oct 3 08:27:17 dws77 smartd[2448]: smartd version 5.1-11 Copyright (C) 2002-3 Bruce Allen Oct 3 08:27:17 dws77 smartd[2448]: Home page is http://smartmontools.sourceforge.net/ Oct 3 08:27:17 dws77 smartd[2448]: Using configuration file /etc/smartd.conf Oct 3 08:27:17 dws77 smartd[2448]: Scan Directive DEVICESCAN found in /etc/smartd.conf. Will scan for devices. Oct 3 08:27:17 dws77 smartd[2448]: Drive: DEVICESCAN, implied '-a' Directive on line 19 of file /etc/smartd.conf Oct 3 08:27:17 dws77 smartd[2448]: smartd: Scanning for devices. Oct 3 08:27:17 dws77 smartd[2448]: Device: /dev/hda, opened Oct 3 08:27:17 dws77 smartd[2448]: Device: /dev/hda, not found in smartd database. Oct 3 08:27:18 dws77 smartd[2448]: Device: /dev/hda, is SMART capable. Adding to "monitor" list. Oct 3 08:27:18 dws77 smartd[2448]: Device: /dev/hdb, opened Oct 3 08:27:18 dws77 smartd[2448]: Device: /dev/hdb, unable to read Device Identity Structure Oct 3 08:27:18 dws77 smartd[2448]: Unable to register ATA device /dev/hdb at line 19 of file /etc/smartd.conf Oct 3 08:27:18 dws77 smartd[2448]: Device: /dev/hdc, opened Oct 3 08:27:18 dws77 smartd[2448]: Device: /dev/hdc, unable to read Device Identity Structure Oct 3 08:27:18 dws77 smartd[2448]: Unable to register ATA device /dev/hdc at line 19 of file /etc/smartd.conf Oct 3 08:27:18 dws77 smartd[2448]: Device: /dev/hdd, opened Oct 3 08:27:18 dws77 smartd[2448]: Device: /dev/hdd, not found in smartd database. Oct 3 08:27:18 dws77 smartd[2448]: Device: /dev/hdd, appears to lack SMART, use '-T permissive' Directive to try anyway. Oct 3 08:27:18 dws77 smartd[2448]: Unable to register ATA device /dev/hdd at line 19 of file /etc/smartd.conf Oct 3 08:27:18 dws77 smartd[2448]: Device: /dev/hde, No such device or address, open() failed Oct 3 08:27:18 dws77 smartd[2448]: Unable to register ATA device /dev/hde at line 19 of file /etc/smartd.conf ... repeats until /dev/hdl and /dev/sda through /dev/sdz Oct 3 08:27:18 dws77 smartd[2448]: Device: /dev/sdz, No such device or address, open() failed Oct 3 08:27:18 dws77 smartd[2448]: Unable to register SCSI device /dev/sdz at line 19 of file /etc/smartd.conf Oct 3 08:27:18 dws77 smartd[2448]: Started monitoring 1 ATA and 0 SCSI devices Oct 3 08:27:19 dws77 smartd: smartd startup succeeded
Mark, Obviously, for your system you'll want /etc/smartd.conf to look something like: # /etc/smartd.conf /dev/hda -a -m mark_h_johnson and since you obviously keep an eye on syslog, you might also want to add -I 194 to the above, (assuming that your disk's temperature Attribute is #194). Now to address the "real" point of your comments... It would be sensible to (on Red Hat) parse /etc/sysconfig/hwconf to identify disk devices. A python script that did a competent job of this could probably be done in minutes - and could be dropped right into kickstart to create a custom /etc/smartd.conf. I'd also be happy to include such a script in the smartmontools package, if it existed. Unfortunately each different distribution has their "own" way of creating such a database, so this does need to be redone for each different distribution. And unfortunately I don't know python. Perhaps someone at Red Hat will take this up.
we do have such a script these days, and hopefully it does the right thing. I'm going to close this bug as its been around for over a year without comment. Please reopen if the current release doesn't do what you think it should.