airhacks.fm podcast with adam bien 1 Billion Jobs Daily with Zero Dependencies Java
Sep 28, 2025
Ronald Dehuysser, creator of JobRunner, discusses revolutionizing job scheduling with a focus on efficiency and sustainability. He shares insights on processing 1 billion jobs daily while utilizing carbon-aware scheduling linked to European energy data. Ronald dives into JobRunner's evolution from open-source to a professional model with advanced features, and emphasizes Java's performance over Python. The conversation also touches on the implications of serverless architectures, operational risks in critical sectors, and future opportunities in edge computing and WebAssembly.
AI Snips
Chapters
Books
Transcript
Episode notes
From Hobby Project To Billion-Jobs Platform
- Ronald built JobRunner pro after maintaining open source and a first customer enabled him to work full-time for months.
- Usage grew to ~1 billion jobs/day and inspired carbon-aware scheduling features.
Schedule Jobs Around Renewable Peaks
- JobRunner uses European grid forecasts (ENTSO-E) to schedule jobs when renewables peak and carbon intensity is low.
- Energy prices correlate with CO2 so scheduling by price usually also reduces emissions and cost.
Use Price As A Carbon Proxy
- Use energy prices as a practical proxy for carbon intensity because they match carbon forecasts ~99.9% of the time.
- This lets organizations reduce emissions and lower compute costs simultaneously.
