I use tarsnap for all my backups, but that's a service as well as a tool.

This is my daily crontab: http://pastie.org/8982530

Same here. Just did a full backup of my VPS:

               Total size Compressed size
  All archives 3.7 GB     1.5 GB

Deduplication ftw.

That's just showing the compression (deflate). Deduplication is why "(unique data)" line which you didn't paste will be lower.

CrashPlan. Works well on all major OSes but it does use a lot of memory since it's Java. You can use it to backup to their own online service or you can do peer-to-peer between systems with CP installed.

rsync.net lets you do things like this:

  mysqldump -u mysql db | ssh [email protected] \ 

"dd of=db_dump"

and this:

  ssh [email protected] s3cmd get s3://rsynctest/mscdex.exe

and this:

  ssh [email protected] "git clone git://github.com/freebsd/freebsd.git freebsd"

... and has had a "HN readers" discount for new users for many years now.

Yeah, would have loved to have something for all platforms... Haven't used Linux in a long time so I'll admit I'm unfamiliar with it all. Did try to do some research but couldn't find a solution that would cover all distros.

Next year, with more time, I'll try and add something there!

> but couldn't find a solution that would cover all distros.

For your purposes, it would be fine to post something that works on Ubuntu as long as you can verify that it's not Ubuntu-only (if it's open source and works on at least two distros, chances are it works on any distro)[0].

[0] Heck, even most things that are made for Ubuntu work on other distros with minimal modification - it's just a matter of convenience.

Updated the site a couple of hours ago with Ubuntu instructions, thanks to our German translator

rsnapshot (http://www.rsnapshot.org/) which is a wrapper around rsync.

Backups go to a local machine with a three drive linux software raid-5 array.

Occasional offsite sync to another remote Linux box.

There are a lot:

http://skeptu.com/online-backup-service+has-client:linux-app...

I'm using spideroak. The Linux client is actually functional.

duplicity to S3 here. I'd like to eventually figure out how to point duplicity to Glacier. And of course it'd be nice if GPG supported some more modern cyphers and modes, but it's Good Enough(tm).

I don't think you can point it directly at Glacier (if I'm wrong someone please let me know) but what you can do is to set "Lifecycle" options on your S3 bucket.

I have a folder in one of my buckets, for example, that I call "archive". I put in there anything I want to store on Glacier long-time. With my lifecycle settings, objects in that folder are automatically moved to Glacier after one day.

backuppc, incremental backups, de-duplication, supports Windows/Linux with smb/ssh/nfs etc