When a Scale testbed encounters a test failure, it'd be better if the LogCabin servers would stop taking snapshots, as they might be discarding valuable information. This issue is to create a command to tell the servers to stop snapshotting altogether.

As a safeguard in case this was ever accidentally executed against a customer's cluster, we probably want it to disable snapshots for some period of time, e.g. a week or until the next reboot, whichever comes first.