Red Hat Bugzilla – Bug 113432
smartd dead but subsys locked
Last modified: 2015-01-04 17:04:16 EST
From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4.1)
Description of problem:
smartd fails to start automatically during bootup. Checking the
Service Configuration browser reports the error given in the summary.
This is an "out-of-the-box" installation with no smartd.conf.
Hardware is a Dell Inspiron 1100 with a Hitachi_DK23EA-30 hard drive.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1.Boot the system (Kernel is 2.4.22-1.2140.nptl
2. Open redhat-config-services; check state of smartd
Actual Results: Status textbox shows the message.
Expected Results: Message such as "not running" or "stopped".
Arjan, can you please comment on this bug report? I'm not sure what's
going on here. Could this just be a stale lock file in
/var/lock/subsys? Note that smartd does have a '-p' == '--pidfile'
option that you can
use to write the PID into /var/run.
I just recieved this error using Fedora Core 2 Test 3.
To fix this problem I edited the /etc/smarted.conf file to reflect the
SMART enabled drive that are present in the system.
It appears that the installer does not configure this file per machine
but uses a default config file that gusses that /dev/hda is in fact a
hard drive (on my machine it is not).
You are correct that Fedora ships with /etc/smartd.conf configured to
monitor /dev/hda. If that drive does not exist, then errors result.
See related bug reports:
The best solution might be a python script that parses the hardware
database and/or /etc/fstab and then constructs a list of devices to
monitor, and puts this into /etc/smartd.conf.
in future, smartd won't be enabled by default. At some stage, if such a script
were to be written, we could look into reenabling it by default again.
Dave: This will work around the issue, to be sure. But for the record,
/dev/hda is a real disk on all my systems. Do I have to specify by partition
(for instance /dev/hda3 = /boot)?
no, it needs to talk to the whole disk, not a partition.
Is this still a problem with the latest errata packages ?
We've done the script mentioned in #3 for some time, so this should be working.
This seems to be working now. In my last FC3 installation smartd seems to be
OK. I haven't been able to test this on the computer which displayed the issue,
but it's no longer a problem for me. Erik