After successfully using Amanda to backup Linux and Windows clients to disk I want to try backing up some data offsite using Amazon's S3. I have only found one document on Amanda and S3 in the How to collection. I also found the sample S3 configuration file. From the Amanda-release file it says I have version 3.2a, but I downloaded and installed 3.3.0, I don't know if that is part of the issue.

I am totally confused. How do I connect Amanda to S3?

First off in the sample amanda.conf file it defines an application app_amgtar then defines a dumptype gui-base using the application as the program. To me it looks the same as using program "GNUTAR" that I used in my other configurations. Is there a difference?

Next what is the significance of the dumptype gui-base? I don't see anything special in the dumptype. Why GUI?

Is auth "bsdtcp" the only authorization that can be used with S3?

Now looking at the How To. It looks incomplete should I add the lines to the sample file?

Do I have to create a S3 bucket before I even begin all of this fun?

ANY help would be appreciated