Yes I know there are some scripts that I can use with telegraf. But I don't like to modify my host so much if I need to reinstall it.
Any thoughts on adding this to the external metric?
currently our daemon for the external metric server gathers and sends this data every 10 seconds,
i would rather not add there a smartctl execution for all disks
(as that daemon does already many things in each loop)
but by default the smartd daemon runs already and sends email on errors, i guess this could be configured to call a script (see 'man smartd' for this)