Burning the playbooks: fireside chat with DHH
summary
Claims that moving off hyperscalers saves real money: company spends 3.2 million dollars a year on cloud and finds buying servers with 192 vCPUs per box is far cheaper and yields better uptime for SLA commitments. Describes a self-built deployment tool Kamal that uses KVM and Docker, provides default Rails Docker configs and lets teams deploy identically to cloud or on-prem, and reports Rails 7 + importmaps work with these defaults. Uses ChatGPT but avoids strong AI predictions.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:55Evaluating new tech
- 12:39Cloud vs self-hosting
- 24:49Shape Up at 37signals
- 32:32Work-life balance
- 35:20Rails & Kamal contributions
- 40:36Open source impact
- 42:39Closing remarks
related talks
26mFireside Chat: Going Cloud-Native: Supercharging the Journey to Digitalization
21mAvoiding That $1M Dollar AWS Bill - Mark Michael, Glenn Oppegard
26mCloud Agnostic Webscale
21mHybrid Cloud - nothing to worry about! 🌥️
25mFrom Duct Tape To Declarative: Playtika's Platform Overhaul at Unicorn Scale
46mDeploying and maintaining software for today's infrastructure
01h 03mCloud Devops
47mIgnites