Sharding

1. The Incident

The replicas worked. Reads spread across three machines, product pages are fast, and the site survived its first primary failover. Eighteen months pass. The store grows.

And now the primary is failing, in a way no replica can help. Writes have grown from 300 per second to 2,500 per second: orders, cart updates, inventory changes, reviews. Remember the last lesson's trap: every replica must apply every write to stay current