18 May 2013, 07:04

using duplicity to backup Marble Track 3

##07:04 Saturday 18 May 2013

10:46pm Friday 17 May 2013

I had been using Cyberduck, but it has to update the entire huge file each time so takes too long and cannot recover from network errors.

I found and even installed expandrive but it’s not free, so I kept looking

I looked at boar which would probably do just fine, but I did a couple more searches

I found git-bigfiles which would probably also do fine,

Then I found a quite recent article with several of my preferred keywords: Backing up to DreamObjects with Duplicity

Everything went pretty smoothly with a couple caveats.  Note: the following isn’t meant to be a tutorial so much as a reminder to myself how I installed Duplicity.

 brew install librsync python gpg ncftp

this failed until I ran

brew update

Now back on track, but you should use latest version of Duplicity

 tar xvzf duplicity-0.6.21.tar.gz

 pip install boto httplib2 oauth

 cd duplicity-0.6.21

 sudo python setup.py install

Now save the following into ~/.boto

[Credentials]
aws_access_key_id = xxxxxxxx your access key xxxxxxxxx
aws_secret_access_key = xxxxxxxxxxxx your secret key here xxxxxxxxxxxxxxx

Yay ready to go

 duplicity –allow-source-mismatch ~/Documents/duplicity/ s3://objects.dreamhost.com/duplicity

duplicity spooked me at first when it responded with this:

Local and Remote metadata are synchronized, no sync needed.
Last full backup date: none

I was like “gah! it didn’t backup!!!” but just give it some time(*) it should work.

(*) it took less time than for me to find another possible candidate: s3cmd