smart_ plugin

smart_ is a wildcard plugin to monitor all SMART attributes of a specified hard drive.

Language	python
Known working OSes	Linux2.4 Linux 2.6 SunOS *BSD
written by	Nicolas STRANSKY

Configuration

You need to install python in order to run the plugin. You also need a valid smartmontools installation.

This plugin needs the following minimal configuration in plugin-conf.d/munin-node

[smart_*]
user root
group disk

To monitor a S.M.A.R.T device, link smart_<device> to the smart_ plugin file. E.g.

ln -s /usr/share/munin/plugins/smart_ /etc/munin/plugins/smart_hda

...will monitor /dev/hda.

Auto configuration

smart_ autoconf

returns "yes" if all required programs are found.

On Linux2.4 Linux 2.6 SunOS *BSD, the following command should list the autoprobed disks where the SMART informations are readable. Be aware that this command will actually try to get SMART informations from all the detected drives on the system.

smart_ suggest

Additional parameters

• smartpath - Specify path to smartctl program (Default: /usr/sbin/smartctl)
• smartargs - Override '-a' argument passed to smartctl with '-A -i'+smartargs

Parameters can be specified on a per-drive basis, eg:

[smart_hda]
user root
env.smartargs -H -c -l error -l selftest -l selective -d ata
env.smartpath /usr/local/sbin/smartctl

Use this to monitor drives pluged in a 3Ware RAID card:

[smart_twa0-1]
user root
env.smartargs -H -l error -d 3ware,1

[smart_twa0-2]
user root
env.smartargs -H -l error -d 3ware,2

# And so on...

SATA drives:

[smart_sda]
env.smartargs -d ata -a

Remarks

The smart_ plugin displays normalized VALUES for all SMART attributes. Subsequently, the value #194 (the temperature) is also a normalized VALUE and does not really reflect the temperature of the disk (the RAW_VALUE). The behaviour of the plugin is not going to change regarding the returned value of the temperature. If you want to monitor the actual temperature of your disks, use a separate plugin (hddtemp_smartctl) to monitor raw temperature values. But the smart_ plugin displays normalized VALUEs, not RAW_VALUEs. VALUEs are "normalized" by the disk firmware, so they can be compared against the given threshold (THRESH) - the higher, the better. Failure is indicated, if a particular value falls below the threshold. In most cases, the "RAW_VALUE" doesn't make much sense, except for temperature but again, displaying the temperature is out of the scope of the smart_ plugin.