Since I first came to mint in 2016, I’ve been a fan of redo. Compared to alternatives like clonezilla, it is a simple program to use, especially for a newbie, but:
- It hasn’t been updated since 2012.
- The redo website, redobackup.org, has disappeared - the domain is available to buy if you want
- The iso is based on ubuntu 12.04 and running a 3.13 kernel so a long way out of date.
- Looking under the hood, it sort of doesn’t understand GPT. MBR = 512 bytes at the start of a disk, GPT = 33x512 bytes at the start and copied at the end of the disk. Redo uses dd to copy the first 64x512 bytes, so okay but it misses the copy at the end.
Time to look at alternatives, or since I’m retired, maybe redo redo.
- Clonezilla - I’m not a fan, clunky UI and confusing, especially for a newbie.
- Fsarchiver - a CLI utility, there are GUIs for it, I tried installing from ppa dieterbaum/qt-fsarchiver, it complained it wanted v0.8.5-8 for fsarchiver, the version from the repos is v0.8.4. Then ran, but woudn’t show any drives/partitions. Think there is a python front end but not tried. Also not impressed that NTFS support is still stated to be ‘experimental’.
- Mondo Rescue - a CLI UI and doesn’t save images.
- Partimage - CLI with a GUI, but no ext4 support.
- Doclone - I want an iso image not a tar.gz to build.
- Partclone - CLI, it’s what redo and clonezilla are based on.
- It is windows based, I have an instinctive aversion to this
- It out-performs redo on speed, but under-performs on image size (test - 45G used NTFS GPT drive on SSD to 5400rpm HDD, macrium reflect 8:48mins/20.8G image file, redo 13:54mins/18.4G)
- It is very slow to boot (USB3), macrium = 1:31mins, redo = 22 secs.
- Reasonably simple to use - I’d argue that redo is simpler.
- It is a front end for partclone.
- It is a perl script.
- It uses glade to generate the GUI (I’d never heard of it).
- It uses gzip to compress the output from partclone - no choices on compression level, set at 3.
- Output then ‘split’ into 2048M files - to meet the 2TB limit on legacy filesystems - no choices.
Moving forward, it would seem that any priorities for an updated redo should be:
- Replace gzip with pigz
- Move to a modern, small distro as the base - it would be nice to be able to boot from CD. Open to suggestions here, but I’d like it to be debian/ubuntu based (apt). antiX????
- Pascal (lazarus - reasonably easy to implement a GUI), ~ 20k lines of code.
- C++ (arduino), but I'd have to marry that with a GUI (glade/qt5 designer? = zero experience), ~ 2k lines of code.
- Python, < 500 lines of code.
View this as the equivalent of Catweazel’s aims to produce something akin to synctoy for linux - a project with no defined end date (at least for me - ‘her indoors’ has a long list of things to do in the house and garden) .