Trivially bug free code for a greener planet
summary
Claim of reduced cloud costs by ~80% backed by concrete rewrites: replace heavy Spring Boot Kafka stream pipelines that serialize 40 MB JSON blobs (peaking at 460 GB/s of data), with a Postgres listen/notify connector and compact services in Zig and Go using zero‑copy AVRO. Measured wins: latency drops from ~33 days backlog to 4–15 ms, memory per service falls from multiple GB to ~10–500 MB, fleet shrinks from ~780 CPUs and $400k/year. Libraries: go‑listen and a tiny Kafka pooper.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 0:30About me
- 2:10Prelude
- 4:13Problem description
- 9:51refactoring
- 11:44Persistence
- 15:51Kafka
- 18:49Kafka Pooper
- 21:34Latency
- 22:30Cloud costs
related talks
46mMaking Software & Data Architectures Sustainable
46mAdopting Rust by gracefully oxidising web applications
45mThe Joy of Building Large Scale Systems
09m1000+ Compile-time Errors Later and Running Smoothly in Prod
49mHow to Save the Planet Without Leaving your Desk!
03h 04mHow We Evolve Clever Cloud – Postmortem on a 10‑Year Project
39mApplying refactoring techniques ...
50mModern Java microservices in the Cloud