Re: [Duplicity-talk] Using as a more efficient sync mechanism

[Mogliii]

It's not quite what I need. What I'm dealing with aren't git trees, by
and large, and creating a git tree out of several terabytes of content
and a few million files would be an interesting exercise in "let's see
how much memory a git process can eat." :)

(As a comparison, running any git operation on linux-stable.git, which
is only 600MB and tens of thousands of files, eats up about 700MB of ram.)

Best,

There is also git-annex
http://git-annex.branchable.com/

It only tracks the hash and location of files in git. The actual files are in the file system.

The author of git-annex is currently working on a deamon
http://www.kickstarter.com/projects/joeyh/git-annex-assistant-like-dropbox-but-with-your-own

Not sure if this helps you in any way