Theories on corruptioon cause

The DNS I manage failed badly earlier this week. It was running one RAID 1 and one RAID 1+0. Both arrays developed massive corruption across most files literally overnight. The only solution was a fresh install of the OS (Windows Server 2003). It seems OK right now, and all hardware seems fine or is at least working now.

What could have caused such sudden and mass corruption across multiple arrays? This is an important question because I really want to prevent this from happening again. Any and all wild speculation is welcome.

 

Greetings, Yog-Sothoth.

Common to all arrays + causes such instant corruption + cures itself (if you discount the OS reinstall) = ...

Power?

 

Good theory, thanks. I live in Florida, the lightning capital of the world, so I keep the server plugged into a decent battery-backup surge protector. Perhaps some careful monitoring and diagnostics of the power supply are in order. Could anyone recommend some good free software for that purpose?

So far I'm thinking that the 2 most likely culprits are the power supply and the RAID controller. Can anyone think of a purely software issue that could cause this?