As the standalone Observium server grows with more and more devices, rules, checkers, etc. The possibility of losing that all due to hardware failure would make me mad. Is there any best practices of making the backups of Observium data to make the emergency restore more faster than from zero ?

Thanks