Scalable is Awesome, Literally!
summary
Argues that “scalable” is hand‑wavy and proposes measurable engineering: use Erlang, measure latency and throughput, prefer suitability over vague scale claims, and design evolvable systems. Describes building CloudBees RUN PaaS (used by tens of thousands of customers and serving tens of millions of end‑users) and a real outage caused by a validating XML parser fetching an external XSD that made every deployment fail. Prescribes concrete patterns: build naive, avoid dependencies, observe, measure, and iterate.
AI-written summary · about
chapters (29)
- 0:00Introduction
- 1:10Step 1 Build
- 2:05Scalable
- 3:50The Premise
- 4:33The Myth of Scalability
- 5:36Infinite Scalability
- 6:12Outages
- 7:27No magical property of scalability
- 8:02The term scalable
- 8:30The word awesome
- 9:21The cheetah
- 11:48Joe Armstrong
- 14:53Antipatterns
- 16:09Vetting Problems
- 17:51Build a Naive
- 19:39Overestimating
- 21:03dependencies bad
- 22:08XML
- 24:02Dependencies
- 25:03System
- 25:58Dependency
- 28:07Cloud
- 29:13Laytons
- 30:49The Ecosystem
- 32:03Evolvable
- 38:02Exercise
- 38:54Fitness Test
- 40:08Evolution
- 42:50Monolithic Architecture
related talks
56mScaling Up!
27mHow can you Scale if you Don’t Trust It?
19mScaling Fast – Engineering Lessons From ~15 Years of Tech Startups
48mScale, Microservices and Flow
37mScale, Flow and Microservices
42mScale, Flow and Microservices
46mAgile at Scale Generative Principles, an Alternative to Scaled Frameworks
26mScaling engineering organizations