The actual problem was with their DB, not the repo storage so I'm not sure what the actual fallout was, but people lost things like issues/merge requests, so I can see how workflow would be interrupted.