Articles

The cache scheduled the outage: stampedes, avalanches and how to survive midnight

A backend interview classic — 10,000 rps in front of a 1,000 rps database, one cache, one shared TTL, and a database that dies at 00:00:00. What the failure is actually called, why it does not recover on its own, and the full ladder of fixes: TTL jitter, single-flight coalescing, stale-while-revalidate, probabilistic early expiry and load shedding.