How would you design a distributed caching layer for a high-traffic content platform to minimize database load?
