You are not logged in.
Hi all,
I feel Linux lacks a good and simple graphical tool to do desktop backups for regular users. I'm thinking of writing one myself (probably a front-end to a good command line tool), but before I get started, I want to make sure I'm not reinventing the wheel.
So:
- Is there a good graphical tool available already? It should be: simple (minimalist interface, sane defaults), fast and incremental.
- What do YOU currently use for your desktop backups?
EDIT: rsync alone is not great, because a good backup solution keeps multiple snapshots. Imagine you deleted a file by accident 15 days ago, and realize only today: rsync won't save you.
EDIT2 : I have tried backintime, I think it's a step in the right direction, but it's not that great. Maybe I should improve on it instead of starting something new, though.
EDIT3 : Has anyone tried PyBackPack (http://andrewprice.me.uk/projects/pybackpack/) ? I'm trying it at the moment and it seems to be getting it right. I'd like to hear some success stories before getting my hopes to high
Last edited by lardon (2010-05-11 13:33:55)
Autojump, the fastest way to navigate your filesystem from the command line!
Offline
Maybe Grsync? It is just a frontend to rsync, but it seems to be simple and powerful enough.
Anyway, take a look at this thread: http://bbs.archlinux.org/viewtopic.php?id=74058
Offline
rsync -a --link-dest=$OLDBACKUP * .[^.]* $NEWBACKUP
if I recall correctly... I have to consult the man page every time so it might be wrong
Offline
It's not GUI, but rdiff-backup is my current favourite.
Offline
extra/sbackup
A simple backup solution intended for desktop use
I use it to backup my home on an external hdd
For more info, see : http://sourceforge.net/projects/sbackup/
and https://help.ubuntu.com/community/Backu … ackupSuite
Last edited by xamaco (2010-05-11 12:50:47)
Offline
Offline
I experimented with several approaches to back up my local files but recently returned to sbackup despite its Gnome dependancy. While its GUI interface could be improved it is fairly simple and intuitive, does differential backups, can be run in batch mode (by cron), and needs fairly low amounts on memory space and run time when doing incremental backups.
Additionally I use rsync to back up my whole system on a separate hard disk.
Last edited by bernarcher (2010-05-11 13:12:24)
To know or not to know ...
... the questions remain forever.
Offline
Wrote my own in Bash... uses a mix of squashfs, aufs, and rsync to make incrementals.
Offline
Wrote my own in Bash... uses a mix of squashfs, aufs, and rsync to make incrementals.
Is it any better than rdiff-backup?
Autojump, the fastest way to navigate your filesystem from the command line!
Offline
DarGUI because it uses dar, http://dar.linux.free.fr/. Dar itself is very simple and can perform a variety of backups. I suggest you at least check it out.
Offline
I see you are much into a comprehensive GUI. Areca is my recommendation, I'm using it, it is not the fastest, but pretty straightforward, and easy to handle, yet feature full. Warning: java needed.
http://aur.archlinux.org/packages.php?ID=13637
zʇıɹɟʇıɹʞsuɐs AUR || Cycling in Budapest with a helmet camera || Revised log levels proposal: "FYI" "WTF" and "OMG" (John Barnette)
Offline
Is it any better than rdiff-backup?
Hard to say as I haven't used rdiff-backup. Squashfu compresses the seed (not the incrementals), but comes at a cost of some fairly wonky dependencies. Basically, I never wrote it with simplicity in mind. I just wanted to learn more about aufs (and I had a desire to learn more Bash as well).
I'm also a fan of the reporting feature I built into SquashFu...
:: SquashFu Usage Report
Bin ID Date Created Size
10 2010-05-11 02:10:01-04:00 544M
9 2010-05-10 02:10:01-04:00 552M
8 2010-05-09 02:10:01-04:00 492M
1 2010-05-08 02:10:01-04:00 491M
5 2010-05-07 02:10:01-04:00 507M
6 2010-05-06 02:10:01-04:00 476M
7 2010-05-05 02:10:01-04:00 539M
2 2010-05-04 02:10:01-04:00 495M
3 2010-05-03 02:10:01-04:00 614M
4 2010-05-02 02:10:01-04:00 982M
Incremental Total 5.6G
quake-seed.sfs 2.3G
Grand Total 7.9G
I'm starting to feel like Kiwi, plugging my own programming...
*ducks*
Last edited by falconindy (2010-05-11 14:19:59)
Offline
Why would one want to use something like rsync over something like git?
Offline
It doesn't snapshot -- I occasionally back-up my back-ups and don't need that -- but it does its job well.
#!/bin/bash
### System Backup ##############
# Version 0.5 by Scott Garrett #
# Wintervenom [(at)] gmail.com #
################################
backup_label="demonpit"
exclude="/dev/shm/rsync-excludes.rc"
excludes=(
'tmp/*' # Temp files.
'tmp/.*' # Hidden temp files.
'.mozilla/firefox/*/Cache' # Firefox caches...
'cookies.sqlite' # Mozilla-based cookies.
'formhistory.sqlite' # Mozilla-based form history.
'.thumbnails' # Thumbnail cache.
'.recently-used.xbel' # Recent-open history.
'.ccache' # Compiler caches
'sessionstore.js' # Firefox session-saves.
'.gvfs' # GNOME virtual filesystem.
'.local/share/Trash' # XDG trash.
'.local/share/user-places*' # Recent-open history.
'.purple/logs' # Pidgin/Finch logs.
'.cache/Thunar/thumbnailers.cache' # Thumbnailer cache.
'.cache/chromium' # Chrome cache.
'*~' # Backup/temp files.
)
out () {
echo ">> $*"
}
stat_done () {
echo ">> ...done."
}
stat_fail () {
echo ">> ...failed!"
}
if [ $UID != '0' ]; then
echo "Must be executed as root user."
exit 1
fi
out "Mouting backup device ('$backup_label')..."
backup_dev=`readlink -f "/dev/disk/by-label/$backup_label"`
if mount | grep -qF "$backup_dev on"; then
backup_root=`mount | grep -F "$backup_dev on" | cut -d' ' -f3`
echo " - This device seems to already be mounted at '$backup_root'."
echo ' Do you want to use this, instead (y/N)?'
echo -n ' << '
read -n1 confirm
echo
[ "$confirm" != 'y' ] && exit 1
else
backup_root="/mnt/$backup_label"
mkdir -p "$backup_root"
if ! mount -o noatime,user "$backup_dev" "$backup_root"; then
echo ' - Failed to mount device.'
exit 1
fi
echo ' - Device mounted successfully.'
fi
out "Backing up system..."
echo > /tmp/rsync-excludes.rc
for file in ${excludes[@]}; do
echo "$file" >> "$exclude"
done
rsync -axl -h --progress --delete --delete-excluded --exclude-from="$exclude" /{home,usr,opt,var,*bin,lib*,etc,boot,root,srv} "$backup_root/$HOSTNAME"
mkdir -p "$backup_root/$HOSTNAME/"{dev,sys,proc,tmp,mnt,media}
mkdir -p "$backup_root/$HOSTNAME/var/"{tmp,lock}
chmod 1777 "$backup_root/$HOSTNAME/tmp" "$backup_root/$HOSTNAME/var/"{tmp,lock}
stat_done
out "Backing up package list..."
pacman -Qqe | grep -v "$(pacman -Qmq)" > "$backup_root/$HOSTNAME/pacman.list"
pacman -Qmq > "$backup_root/$HOSTNAME/aur.list"
stat_done
out "Unmouting backup partition..."
if umount "$backup_dev"; then
rmdir "$backup_root/$HOSTNAME"
rmdir "$backup_root"
else
echo ' - Could not unmount.';
fi
Offline
Git won't keep permissions or ownership.
Well at least for me that isnt something I miss.
Combine git with btrfs or the like and you have the ultimate backup solution imo.
Offline
Why would one want to use something like rsync over something like git?
Git is terrible with big backups and big files. It was never created with backups in mind. Flyback uses git as its backend. I've tested it on a 60GB backup, and it took hours to go through it.
Autojump, the fastest way to navigate your filesystem from the command line!
Offline
dmz wrote:Why would one want to use something like rsync over something like git?
Git is terrible with big backups and big files. It was never created with backups in mind. Flyback uses git as its backend. I've tested it on a 60GB backup, and it took hours to go through it.
Why don't you use git for your files except for the big files, and backup all of it using rsync.
Last edited by soulicia (2010-05-12 02:36:28)
Offline
Offline
I've used Flyback and Grsync and both did the job with a minimum of fuss/effort...
Offline
Not GUI, but I have used bacula as my backup system for years. At the office, I have an old machine running archlinux with some big drives that is my bacula backup server, all it does is backup the entire network to disk, every night, and has done so reliably for a couple of years. It's not easy to get set up, but once it is, you can pretty much forget it's even running. It has saved my butt from stupid mistakes a couple of times.
Offline
i use unison (backup as sync). it does also have gui (unison-gtk2), which is not needed actually, once you have configured which directories should be synced.
Pinky, are you pondering what I'm pondering?
Offline
For me unison doesn't qualify as a good backup because it doesn't allow you to keep several snapshots. So if you accidentally delete a file and only notice after doing a unison sync, your file is still lost.
Autojump, the fastest way to navigate your filesystem from the command line!
Offline
Not really what I use, but something I saw a while back (no idea if it still exists), but there was Time Slider in OpenSolaris back in 2008.
http://java.dzone.com/news/killer-featu … olaris-200
Not sure if it was limited to the ZFS filesystem that (I believe) OpenSolaris uses. But I assume that's something along the lines of what you were looking for?
Lswest <- the first letter of my username is a lowercase "L".
"...the Linux philosophy is "laugh in the face of danger". Oops. Wrong one. "Do it yourself". That's it." - Linus Torvalds
Offline
Lots of great stuff here, but I still miss backintime - it is in the AUR and works a bit like apple's timevault or time machine or whatever it is called...
never trust a toad...
::Grateful ArchDonor::
::Grateful Wikipedia Donor::
Offline