… or they should be, anyway.
I think that one of the more popular excuses around for not having backups it’s “I haven’t gotten to it”; usually because you don’t have the time to try that fantastic tutorial you found for encrypted-incremental-automatic-deduplicated-control-versioned-backups on Amazon S3.
The thing it’s… it’s ok if you don’t have time for it, because it means you’re doing your job… which very likely isn’t Chief Backups Officer. What it’s not ok it’s that you keep postponing your backups!
That’s why I think that when you’re first configuring your server you should immediately configure some sort of backup that:
- It’s very quick to setup, so you actually do it
- It’s easy to restore from, so it’s actually useful
And since I’m assuming you’re not an idiot, I know you’ll do your best to keep them safe; which doesn’t mean creating some new fancy encryption scheme but using existing tools to do the job (for instance, ssh and rsync are both encrypted, so they’re good enough for transmitting the data to another server).
I’m sure there are plenty cool alternatives to keep your data safe, but the truth it’s that if they don’t comply with these two basic requirements you should wonder if there’s a better, simpler way.