Ask HN: Your Amazon Glacier strategy with versioning and de-duplication on *nix?

Command line, or GUI (preferred).

Something that just sits in the background, monitors change and keeps backing up the change (while taking care of the de-duplication).

What do you use? (I know Arq which I believe has been covered more than once here)

And how do you use it? What is your backup frequency? How do keep the cost minimal? Is it too frequent, or seldom - as in few snapshots a week?

Have you every had to restore? If yes, could you find a way to minimise the price of retrieval?

I've around 700GB of personal data that I just want to be sitting there in Glacier for the worst case scenario.

(PS. Yes, I've looked at Tarsnap and it doesn't work for me).