Want to play around with cluster filesystem but never to time to do so? Having a bunch of old CDs lying around and they should get digitalized, but it just takes for ever? Got a couple of spare computer? Scared of loosing the data in the process? Behold, there is a solution:

  1. Install you spare-computers with an OS.
  2. Deploy LizardFS in a cluster config on those
  3. Start ripping the shit out of your CDs on all spare computers.
  4. Collect all ripped files from the cluster.
  5. Done.

There are two annoying things about ripping CDs

  1. It takes time.
  2. If it doesn’t you need to collect the files from all the ripping nodes.

LizardFS is an open source distributed filesystem, the brother of MooseFS.

Installation and deployment is quite forward and done in less than halv an hour.

Adding the files on one node will distribute them over x-number of other nodes, preventing data from getting lost.

Regarding the ripping: I used abcde for quick results and put the output directory in the path of LizardFS. That worked beautifully.

Now I just gather the emotional strength to get rid of my CDs.