Cache me if you can: Boosting performance with Data Caching in Python
summary
Shows concrete wins: a simple dict cache makes repeated API fetches drop from 1.14s to effectively 0s per request, and Python's @lru_cache cuts fib(40) from ~21s to near 0s. Covers eviction tradeoffs like LRU eviction, invalidation strategies (manual, event-driven, TTL), and metrics to track cache hits/misses and latency. Recommends external stores redis/pymemcache for multi-server caches and warns caching must not mask bad indexes, poor data modeling, or wrong DB choices.
AI-written summary · about
chapters (10)
- 0:00Introduction and speaker intro
- 1:26What is caching and latency
- 4:13Key considerations before caching
- 7:05When to use caching
- 8:47Cache invalidation strategies
- 10:12Choosing caching approach
- 11:18Implementing caching in Python
- 14:08Local vs distributed caching
- 15:32Measuring cache effectiveness
- 16:15Closing and Q&A
related talks
32mCache me if you can: Boosted application performance with Redis and client-side caching
45mCaching: A trip down the rabbit hole
35mCaching everywhere
06mCache Is the King! Create Cache in AWS Lambdas
05mLightning Talk: HTTP Caching in Python
42mCaching for Cash
01h 04mCaching Best Practices
21mHigh Performance APIs through Caching