Richard Freeman on 2 Dec 2009 08:33:04 -0800


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: [PLUG] Self-hosted online backups?


JP Vossen wrote:
> Obviously, I could script something using tar, GPG, rsync, and/or other 
> tools, but I can't be the only person out there who wants this, and why 
> reinvent the wheel?
> 

At various points I've used a few different approaches.

I don't have servers at multiple geographic locations, which limits my 
options if I want offsite backup.

Right now I'm using this cron script and AWS:

#!/bin/sh
# run weekly backup
BACKDIR='/sstorage3/sarab-back'
UPDIR='/sstorage3/sarab-up'
BUCKET='s3://<bucket-name>'

<insert command here to export your databases to a file in the backup path>

ionice -n 7 nice -n 20 /usr/bin/sarab.sh

cd $BACKDIR
CURBACK=`ls --sort=time -1 | head -n 1`
cd "$BACKDIR/$CURBACK"
rm -r "$UPDIR/$CURBACK"
mkdir "$UPDIR/$CURBACK"

for encfile in * ; do nice -n 20 gpg -o "$UPDIR/$CURBACK/$encfile" -r 
<keyid> --encrypt $encfile ; done

cd "$UPDIR"

ionice -c 3 nice -n 20 s3cmd --no-delete-removed sync . "$BUCKET/" || exit
ionice -c 3 nice -n 20 s3cmd --delete-removed sync . "$BUCKET/"

So, sarab is doing most of the work, and then I encrypt the output and 
ship it to S3.  No trust needed on my part.  All the stuff in this 
script can be found via goodle and is available in Gentoo portage.

Previously I used to use the backuppc package, which might work for you. 
  It runs on a server and connects to a series of hosts (with rules 
defined at various levels from default to per-host), and the backups are 
analyzed and stored in a sparse manner using hard links.  To get it to 
work you'd probably need a VPN connection to all of your hosts.

Backuppc also has some advantages in that it has a nice web-based 
console and you can even give users access to do their own restores 
(either by downloading any saved version of the file via the web 
interface, or by having the server just put the file back on the host 
where it came from).

As others have indicated, there are a ton of other solutions as well. 
The S3-based solution is pretty cheap if you aren't imaging your entire 
OS and you don't need to try to backup multiple systems with sparse 
files.  I pay a few dollars a month for it - and that is with all of my 
photos being protected (but I do burn them to DVD as they accumulate and 
store them offsite and exclude them from backups as that happens - so 
that greatly reduces my S3 bandwidth and space use).
___________________________________________________________________________
Philadelphia Linux Users Group         --        http://www.phillylinux.org
Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce
General Discussion  --   http://lists.phillylinux.org/mailman/listinfo/plug