Orchestrating Quasi-Real Time Data Processing in the Computing Farm of the ATLAS Experiment at CERN

Explore a conference talk that delves into how CERN's ATLAS experiment leverages Kubernetes to orchestrate quasi-real-time data processing in its computing farm. Learn about the challenges and solutions of managing one million physics events per second at a data rate of 5 TB/s using a combination of custom electronics and a massive computing farm of up to 5000 commodity servers. Discover the optimization strategies implemented to launch over 25000 PODs across 2500+ worker nodes in approximately 50 seconds, the impact of Kubernetes Scheduler and Controller Manager QPS values on POD operations, and how custom scheduler profiles achieve POD scheduling rates of around 500 Hz. Gain insights into the intersection of high-energy physics and cloud-native technologies through this practical implementation at one of the world's most advanced scientific facilities.