The problem: if I have 1TB of data, I need to have 1TB of "local" storage to make <i>incremental</i> backups with most tools. That consumes space and disables me to store my incremental backups encrypted/signed/etc<p>There is obviously the not-so-space-technology of backup catalogs. For windows systems there are: backup4all and Genie-Soft Backup Manager that use those.<p>For linux I could not find one. The closest alternative I could come with is 'git add ./*', 'git-gc' and then storing the packfile offsite, but it's a bit unstraight method, I suppose.<p>Are there any free/OSS alternatives for both linux and windows, which you can tell me about?
I know this thread is old, but I ran across it in a search.
I use a tool called dar (<a href="http://dar.linux.free.fr/" rel="nofollow">http://dar.linux.free.fr/</a>) that is along the lines of tar, but with extra features for dealing with multi-volume backup sets and diffs easily.<p>It lets you extract a catalog from the backup set that you can use locally as a checkpoint for doing diff/incremental backups. The catalog is really small, mine is ~10Mb for hundreds of thousands of files.<p><pre><code> ## here is the initial backup: backing up /nas to
## an archive set /mnt/usb/nas on a usb device
dar -R / -c /mnt/usb/nas -g nas
## extracting the catalog from the backup and
## saving it under my home directory
dar -A /mnt/usb/nas -C /home/dan/backup/nas-catalog
## I use the local copy of the catalog in my home folder
## when doing diff backups to another mounted usb device
dar -R / -c /mnt/usb/nas-diff -g nas -A /home/dan/backup/nas-catalog</code></pre>
I don't quite understand the use case. Do you want to backup data and then delete the local copy? I think that's more like archiving than backup, and it's not clear to me what "incremental" even means in that case. Or do you intend to keep your 1 TB of live data, but you don't want the backup tool to use any <i>additional</i> disk space?<p>Speaking of catalogs, I wonder if Duplicity is what you want.