Clojure at 4,000 msg/s-What We Learned, Loved, and Loathed
summary
Swarmify builds a video resilience layer that sustains 4,000 msg/s from browsers to cloud, sending EDN over WebSockets and running unified Clojure/Clojurescript code both client and server. Architecture uses core.async and STM, Aleph for networking, Google Dataflow for big‑data pipelines, CircleCI and a REPL-on-production for debugging. It works in production since 2013 but requires careful JVM memory management, cautious dependency upgrades, and tooling fixes for advanced-compile, race-condition memory leaks, and tricky stack traces.
AI-written summary · about
chapters (10)
- 0:00Intro
- 0:56Experiment
- 2:37Why Clojure?
- 5:37System
- 8:12Dev - Loved
- 13:07Dev - Loathed
- 19:14Ops - Love
- 21:59Ops - Loathed
- 26:58Team - Loved
- 29:31Summary
