∿
SYSTEMS ENGINEERING LAB
Deterministic model · no services
Streaming systems · interactive simulator

Stream Pressure Lab

Explore how arrival rate, service capacity and buffer limits shape queue growth, latency and loss under overload.

Processed
4,200
events in 60 s
Dropped
0
rejected or shed events
Peak queue
0
events waiting
Estimated p95 wait
0 ms
queue ÷ service rate
Queue depth over timeSteady state · 60 seconds
queue depthbuffer capacity

02 / Read the result

Capacity headroom: consumers can process 20 more events/s than producers send.

Expected outcome: queue stays near zero and all generated work is processed.

Design note: average rates hide bursts. Test bursty arrivals and recovery time before choosing a buffer.

03 / Simulation trace