My Server for Aiur: How Starcraft Taught Me To Scale
summary
Measure everything in production, then optimize where the measurements point: use logs and New Relic/Scout to find N+1s and slow queries, add eager loading and indexes, set config.active_record.auto_explain_threshold to catch scans, wrap expensive results with rails.cache.fetch backed by memcache (dolly), move static assets to a CDN and enable assets.digest true with long expires headers, push background jobs to Resque/ActiveJob and scale web dynos with heroku ps:scale.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 3:24Starcraft scaling strategies
- 7:15Backend speed optimization
- 15:36Caching expensive ops
- 20:49Avoid premature optimization
- 22:13Throughput with capacity
- 29:56Data storage patterns
- 33:08Client-side performance
- 40:58Summary and Q&A
