Problem context and objectives
Lead systems engineer · Feed Cache Frontier
Design caching for a personalized news feed where reads dominate writes, celebrities have millions of followers, and freshness matters.
The design must survive feed reads 300K/s and posts 5K/s. Every box must earn its place by satisfying a requirement or containing a failure.
How you win
- 1Trace one user action from request to durable outcome
- 2Connect every component to a stated requirement
- 3Explain the losing tradeoff, not only the chosen technology
Rules and pressure
- feed reads: 300K/s
- posts: 5K/s
- Low-latency first page
- Bounded staleness
New words in this mission
Open a term for a plain-language explanation.latency+
The waiting time from starting an action until its result is available. Lower latency makes interactions feel more immediate.
cache+
A fast copy of frequently needed data. It reduces repeated work, but the system must decide how stale that copy may become.
Live architecture trace
Feed Cache Frontier: request to outcome
Reader sends “read” to Feed service. Cache-aside gives a clear miss path and a TTL bounds staleness when invalidation events are delayed.
Design caching for a personalized news feed where reads dominate writes, celebrities have millions of followers, and freshness matters. Follow one real user action through the complete architecture; a component is useful only when you can explain the request, state, or failure it handles.