Automate snapshots of "grupper"

[henrist]

From previous setup:
/fbs/grupper.snapshots/snapshot.sh

Cron:
*/30 * * * * root /fbs/grupper.snapshots/snapshot.sh 1>/dev/null

The previous setup is using LVM snapshots to ensure a consistent state of the disk while copying files to create a snapshot. We haven't set up LVM partitions in CoreOS yet, and it will require us to keep free space for the LVM snapshots. So to do this the same way we need to migrate the data volume on the host that will have this data.

We might consider instead to add a second data volume to the instance to hold the "grupper" data. This will also help ensuring that full disks does not affect other services.