Performance

Software caching explained: why caches speed up systems, where they fail, and how to think about freshness, consistency, and cache misses.

End-to-end software performance fundamentals: latency, throughput, percentiles, bottlenecks, and measuring user-perceived speed without breaking reliability.

Thundering herd: when many clients do the same work at once and overload a dependency. Understand why it happens, what it looks like, and how to reduce risk.

Backpressure: a system’s way of saying “slow down” before overload turns into timeouts and retries. Understand why it matters and what signals it uses.

Load shedding rejects work during overload so systems stay usable. Learn why it matters, what it looks like, and how it prevents retry storms.