Managing SD Card Images

Having multiple Raspberry Pis doing a variety of tasks with some aspect of new development constantly underway means a lot of different SD Card images.   I got tired of manually managing them so developed a little Python script to help.

What I wanted was a script that would run under Python on a Linux workstation and do the following:

1. Save an image to the hard drive of the Linux box or burn a previously saved image to an SD Card (--save and --burn).
2. Optionally compress the saved image (and decompress as part of the burn process) as the hard drive on my Linux box is kinda small (--zip).
3. Organize my images into three folders with one for original Distro images, one for date stamped Backups, and one for Checkpoints.   An example of my use of Checkpoints:   I have an Arch image that is built just to the point of the network running, one that includes LXDE, and one that is my working Robot (--checkpoint and --distros with backup being default).   
4. List images available for burn in the above folders (--list).

As I developed the script I also added some more options to:

1. Provide password on the command line so it can be fed to sudo for the execution of the dd command 
2. Enable a save or restore at the partition level.   Note that only two partitions are supported and that a burn destination will be reformatted and partitioned to fit the saved image
3. Alter output from the script with one displaying everything (--verbose) and one nothing (--quiet).  The default is to display a minimum of status but to prompt for confirmation of command line options.
4. Added an image number to the --list option such that the user can provide that number for a burn request rather than an entire image name.

Here is the --help for the script:

Options:
  -h, --help            show this help message and exit
  -b, --burn            Burn an image to an SD card
  -c, --checkpoint      Checkpoint type image, default is Backup
  -d, --distro          Distro type image
  -e DEVICE, --device=DEVICE
                        Device name (e.g. sdc)
  -l, --list            List images availble for burning
  -p, --part            Save at partition level
  -q, --quiet           Display nothing (including confirmation request) other
                        than an error
  -r ROOT, --root=ROOT  Root password for sudo
  -s, --save            Save image to computer
  -v, --verbose         Display everything
  -z, --zip             Source (/Target) is (/to be) compressed using zip

Some examples of the script command line invocation:

python mi.py --burn --distro --device=sdd distro-name
python mi.py --burn --distro --zip --device=sdd distro-name
python mi.py --burn --checkpoint --zip --device=sdd checkpoint-name
python mi.py --burn --zip --device=sdd backup-name
python mi.py --save --checkpoint --zip --device=sdd checkpoint-name
python mi.py --save --zip --device=sdd --root=secret backup-name

python mi.py --burn --distro --device=sdd 3
--or--
python mi.py -bde=sdd 3  

Output from the --list command showing image numbers:

will@UbuntuMini:~/Images$ ./mi --list
Images available in Distros
  ( 1) 2013-02-09-wheezy-raspbian:
       1.94gb/z    saved on 2013-04-11 07:06:08.736134
  ( 2) archlinux-hf-2013-02-11:
       1.94gb/z    saved on 2013-04-11 09:10:53.009246

Images available in Checkpoints
  ( 1) arch-1:
       3.97gb/z    saved on 2013-04-13 22:41:33.510442
  ( 2) arch-2-LXDE:
       3.97gb/z    saved on 2013-04-13 23:00:23.804046
  ( 3) tvhe:
       3.96gb/p/z  saved on 2013-04-18 15:15:52.762579

Images available in Backups
  ( 1) 2013-04-12-CritterCam:
       16.04gb/z    saved on 2013-04-12 09:44:34.467685
  ( 2) 2013-04-14-arch-webcam:
       3.97gb/z    saved on 2013-04-14 17:12:50.241074  

I had made the source code for this script available but then pulled it when I a) realized that there were still features that I wanted to add and b) that as I added those features I was also introducing bugs.

I have finally finished development and a fair amount of testing on the three Ubuntu machines that I have available.  The code is available here.  I would really appreciate some feedback and any reports of problems.