Cache=Cash!
summary
Benchmarks use an actual Raspberry Pi and a stack of Rails + Nginx + Memcached + Redis to show concrete wins: initial workflow takes 116 seconds on the Pi, a few lines of fragment caching / Russian‑doll caching cuts render time dramatically (claiming roughly half the EC2 capacity gone), adding HTTP caching, gzip, preheating and per‑user page cache yields further large drops; numbers include a 0.8s page target, a 0.26TB estimate for 10M per‑user pages, and an Ember front end reaching ~14× speedup.
AI-written summary · about
chapters (8)
- 0:00Introduction to caching
- 1:22Why performance matters
- 5:16Measuring and testing
- 11:14Fragment caching
- 15:09HTTP and brute force caching
- 19:32Nginx and page caching
- 22:29Caching for scale
- 24:33Modern stack alternatives
