In reply to annak:
We will either go this route or via Cloud servers in the future but that will be a big architecture change for us.
There's redundancy in the storage system and obviously regular backups. We could be back online in a day or so with minimal data loss if anything major happened in the current setup. As Remus said, the stack is very stable and we've always used hosting where you're paying a bit more to ensure you have good, well maintained, hardware and good tech support.