What do McDonald’s drive-thru lines, highway on-ramp metering lights, and stadium beer service all have in common with software? They’re all bound to the same laws of queuing theory.
In this talk, 8th Light’s Head of Delivery Kevin Kotowski introduces the key concepts that apply queueing theory to software, and explains the differences between terms like “system time and throughput” and “concurrency and parallelism”. Throughout this presentation, Kevin leans on these real-world examples to cut through the jargon and make these lessons more digestible for you and your team.
Kevin finishes the presentation by going a step further and outlining a systematic approach for applying these concepts to your software system. By observing telemetry latency, mapping the critical path, and improving performance on nodes along that critical path, your team can achieve impressive performance improvements.
For more from 8th Light University, check out this Friday’s Meetup where Eric Smith will demonstrate how to build software games in Rust!
Join the Dialogue
Here are a few background resources shared during the talk:
- Applying Queuing Theory to Optimizing the Performance of Enterprise Software Applications
- On the execution of parallel programs on multiprocessor systems. A queuing theory approach
