On the Edge of Rails Performance
summary
Concrete fixes and runnable tools not handwaving: N+1 alerts via Bullet that point to the exact line, rails_indexes rake tasks that suggest missing indexes, Scrooge that learns columns and switches to select-only discovery, Rack::Bug + MemoryLogic + Oink that expose per-request RAM growth and flag actions that balloon by 157 MB, Rubber that scaffolds a full EC2 cluster with backups and monit, CloudCrowd that parallelizes work (five nodes processed Shakespeare jobs in ~21s producing 5,588 words), and MadMimi that outsources editable email templates and mailing lists.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 7:52Rails scaling overview
- 8:35Active Record plugins
- 13:35Query overhead with Scrooge
- 16:55Memory bloat tools
- 26:22Ruby scaling libraries
- 30:57Background processing
- 37:19Email with Mad Mimi
- 44:22Q&A








