Introduction
Smartmontools is a set of applications that can test hard drives and read their hardware SMART statistics. Note: SMART data may not accurately predict future drive failure, however abnormal error rates may be an indication of possible hardware failure or data inconsistency.
This how to will help you to configure Smartmontools to do actions such as shut down the computer or send an e-mail when the disk is going to fail.
Prerequisites
- A modern S.M.A.R.T. capable hard disk
Setting up
Installation
You can install the smartmontools package from the Synaptic Package Manager (see SynapticHowto), or by typing the following into the terminal:
sudo apt-get install smartmontools
Checking a drive for SMART Capability
To ensure that your drive supports SMART, type:
sudo smartctl -i /dev/sda
where /dev/sda is your hard drive. This will give you brief information about your drive. The last two lines may look something like this:
SMART support is: Available - device has SMART capability. SMART support is: Enabled
Enabling SMART
In the case that SMART is not enabled for your drive, you can enable it by typing:
sudo smartctl -s on /dev/sda
Testing a Drive
You may run any type of test while the drive is mounted although there may be some drop in performance. There are three types of test that can be conducted on a drive:
- Short
- Extended (Long)
- Conveyance
To find an estimate of the time it takes to conduct each test, type:
sudo smartctl -c /dev/sda
The most useful test is the extended test (long). You can initiate the test by typing:
sudo smartctl -t long /dev/sda
Results
You can view a drive's test statistics by typing:
sudo smartctl -l selftest /dev/sda
To display detailed SMART information for an IDE drive, type:
sudo smartctl -a /dev/sda
To display detailed SMART information for a SATA drive, type:
sudo smartctl -a -d ata /dev/sda
Note: This also works for IDE drives in new kernels that are being run through the SCSI stack and show up as /dev/sdX
Suggested application: GSmartControl
Take a look at GSmartControl. It's a nice graphical frontend to smartctl; it shows all SMART values, and highlights those that indicate old age or impending failure, plus you may run tests on demand:
As usual, you may install it from Synaptic or running sudo apt-get install gsmartcontrol.
Advanced: Running as Smartmontools as a Daemon
You can run Smartmontools in the background and have it check drives and email when there are issues:
Open the file /etc/default/smartmontools with your favourite text editor. For example (using vim): sudo vim /etc/default/smartmontools. Uncomment the line start_smartd=yes.
How smartd is going to scan the disks and what it will do in case of errors is controlled by the daemon configuration file, /etc/smartd.conf. Again, use your favourite text editor to open this file. There should be one uncommented line, similar to:
DEVICESCAN -m root -M exec /usr/share/smartmontools/smartd-runner
In this example (which is the default for Karmic), smartd will:
scan for all ATA/SCSI devices (DEVICESCAN). The rest of the file will be ignored;
mail a report to the 'root' account in case of trouble (-m);
but instead of the mail command, it will execute /usr/share/smartmontools/smartd-runner and feed the report to it (-M exec program).
/usr/share/smartmontools/smartd-runner is a script that basically saves the report to a temporary file, and then runs anything it finds in /etc/smartmontools/run.d/; take a look there to understand what you already have (there should be a script that mails the report).
There are several -M directives that change when and how often reports are sent. You need to specify (-m something) in order to use them, even if you're not sending any mail.
You may include some useful options:
DEVICESCAN -H -l error -l selftest -f -s (O/../../5/11|L/../../5/13|C/../../5/15) -m root -M exec /usr/share/smartmontools/smartd-runner
In this example, smartd will:
check the SMART health status (-H);
report increases in both SMART error logs (-l);
check for failure of any Usage Attributes (-f);
schedule an Offline Immediate Test every Friday at 11 am, a Long Self-Test every Friday at 1 pm, and a Conveyance Self-Test every Friday at 3 pm (-s) -- see the smartd manual page for what these tests do so you can choose what suits you.
You may also replace DEVICESCAN with the path of the device which you'd like to be monitored (e.g. /dev/sda), and the daemon will only monitor this drive. You'll need one such line for each device.
Actions in case of trouble
You'll want to configure the actions smartd will take in case of trouble. If all you want is a notification shown on your desktop, skip to "Personal computer" below.
Most of the time, you only need to place a script in /etc/smartmontools/run.d/. Whenever smartd wants to send a report, it will execute smart-runner and the latter will run your script.
You have several variables available to your script (again, see the smartd manpage). These come from a test run:
SMARTD_MAILER=/usr/share/smartmontools/smartd-runner SMARTD_SUBJECT=SMART error (EmailTest) detected on host: XXXXX SMARTD_ADDRESS=root SMARTD_TFIRSTEPOCH=1267409738 SMARTD_FAILTYPE=EmailTest SMARTD_TFIRST=Sun Feb 28 21:45:38 2010 VET SMARTD_DEVICE=/dev/sda SMARTD_DEVICETYPE=sat SMARTD_DEVICESTRING=/dev/sda SMARTD_FULLMESSAGE=This email was generated by the smartd daemon running on: SMARTD_MESSAGE=TEST EMAIL from smartd for device: /dev/sda
Your script also has a temporary copy of the report available as "$1". It will be deleted after you finish but the same content is written to /var/log/syslog.
Personal computer
For a visual notification, you may just install smart-notifier. You will see a large popup with the report:
Alternatively, you may create a custom notification (bubble) as seen in other GNOME programs.
You will need to install the libnotify-bin package:
sudo aptitude install libnotify-bin
Now create a text file called 60notify in /etc/smartmontools/run.d:
sudo vi /etc/smartmontools/run.d/60notify
and add the following to the file:
DISPLAY=:0.0 notify-send --icon=important "Possible disk failure" "$SMARTD_DEVICE may have a problem"
(The DISPLAY=:0.0 part is a variable assignment that helps programs to locate your X server. It's already set for your terminal, but the script lacks it since it is being run inside a different session).
Now give it execute permissions:
sudo chmod +x /etc/smartmontools/run.d/60notify
This will produce a nice libnotify bubble with a warning icon:
You may also experiment with Zenity:
DISPLAY=:0.0 zenity --text-info --filename="$1" --title="smartd: $SMARTD_DEVICE may have a problem"
Notice: Be very careful with these scripts as they are run under the root account.
Server
Here, you may wish to handle things differently. In this example we want to mail an admin and shut down the server. Comment out the line that contains DEVICESCAN, by adding # to the beginning of the line. Then, add this to the end of the file:
/dev/hda -H -l error -l selftest -f -s (O/../../5/11|L/../../5/13|C/../../5/15) \ -m admin@somewhere.com -M exec /usr/share/smartmontools/smartd-runner
(Be sure not to add any whitespace after the "\")
Now, we are going to make the script which is going to shut down the computer *after* we mail the admin. Create a text file called 99shutdown in /etc/smartmontools/run.d and add the following to the file:
sleep 40 shutdown -h now
The number 99 at the start of the filename is to ensure that it is called last when smartd-runner runs. It will wait 40 seconds and then shut down the computer. Of course, you may customize this at will; you may not wish to turn off the server.
Now, it is time to start the daemon:
sudo service smartmontools start
Testing
If you want to test all these actions, add -M test after exec /usr/share/smartmontools/smartd-runner and restart the daemon (sudo service smartmontools restart). When the daemon comes up, it will execute the script immediately with a test message. Notice: If you included the shutdown -h line, the script will shut down the computer as soon as the service starts. To fix this, you will have to start the computer in recovery mode and remove the -M test option from /etc/smartd.conf.
Based on Gentoo Wiki: HOWTO Monitor your hard disk(s)withsmartmontools.
Note
Before running this, be sure to check that you have a "mail" command, and do a test first to your address. On my default Fiesty:
jim@beorn:~$ mail
The program 'mail' can be found in the following packages:
- mailx
- mailutils
Try: sudo apt-get install <selected package>
Make sure you have the 'universe' component enabled
bash: mail: command not found
Utility: Checking all disks at once
Note: Following the Gentoo Wiki I made a modified script which checks all the disk in /dev/disk/by-id/ Just invoke the script below as follows:
./smart.sh short|long|offline
The script creates a directory named smart-logs and stores all the files there.
# Script by Meliorator. irc://irc.freenode.net/Meliorator
# modified by Ranpha
[ ! "$@" ] && echo "Usage: $0 type [type] [type]"
[ ! -e smart-logs ] && mkdir smart-logs
[ ! -d smart-logs ] && Can not create smart-logs dir && exit 1
a=0
for t in "$@"; do
case "$t" in
offline) l=error;;
short|long) l=selftest;;
*) echo $t is an unrecognised test type. Skipping... && continue
esac
for hd in /dev/disk/by-id/ata*; do
r=$(( $(smartctl -t $t -d ata $hd | grep 'Please wait' | awk '{print $3}') ))
echo Check $hd - $t test in $r minutes
[ $r -gt $a ] && a=$r
done
echo "Waiting $a minutes for all tests to complete"
sleep $(($a))m
for hd in /dev/disk/by-id/ata*; do
smartctl -l $l -d ata $hd 2>&1 >> smart-logs/smart-${t}-${hd##*/}.log
done
done
for i in {1..10}; do
sleep .01
echo -n -e \\a
done
echo "All tests have completed"(Remember to give execute permissions to the script with chmod +x smart.sh).
Testing drives behind MegaRAID
If /dev/sda is a MegaRAID device then straightforward execution of smartctl on it is not effective. It just returns an empty report for the controller itself. To get S.M.A.R.T. attributes of a drive behind the RAID controller you need to use the following command:
for i in {0..23}; do sudo smartctl -x /dev/sda -d megaraid,$i >>./OUTPUT; done
Examples of SMART reports
See large collection of smartctl reports for various hard drives here.