Apart from backups, a nice RAID 1 setup will keep the service running during a disk failure and it is easy to repair when it happens, just change the faulty disk and re-sync. Some also use those as backups, by changing a spare disk regularly and keeping it somewhere safe.

If you do backups, don’t forget to test them. You don’t want to find yourself trying to restore some files and find out that they’re corrupt or you’re missing the encryption key. I remember trying to restore a database only to discover that the supposed production backups where from the testing environment.

For the noise: get a fanless PC and SSD storage. As these are virtually silent, you can keep the server somewhere open so it doesn’t overheat.