Queuing Theories
summary
Queueing theory applied to Kafka/Kinesis gives concrete, actionable metrics: measure lambda (arrival rate) by counting produced events, record producer and consumer timestamps to get waiting time W, service time S, and residence time R = W + S, and compute utilization = lambda / mu (mu = service rate). The claim is that once utilization approaches or exceeds 1 the system hits a hockey‑stick where residence time explodes, so plot utilization vs residence to predict breakpoints and scale before incidents.
AI-written summary · about
chapters (4)
related talks
40mSystem Performance and Queuing Theory - Concepts and Application
48mKafka as a Message Queue: can you do it, and should you do it?
40mResilient Design 101: Queue Theory
38mKafka Message Broker Under High Load
53mDemystifying Kafka
40mEverything Will Flow
16mTo queue or not to queue, that is the question.
42mA Kafka Client’s Request: There and Back Again