How did they even manage to generate a terabyte sized repo, that's crazy. Do they have something written up on how it's structured and why they'd even go that route?