Work back to the consumer count that satisfies the stability condition
How many consumers a message queue needs follows from the arrival rate and the service rate of one consumer. Enter both and this page computes the minimum consumer count that keeps utilization ρ below 1, plus the spare capacity at that point. Add a target utilization or the consumer count you already run and both are compared on the same screen.
The formula is the queueing-theory utilization ρ = λ / (c × μ), where λ is the arrival rate, μ the service rate of one consumer and c the number of consumers. What matters is that the stability condition is ρ strictly below 1. A ρ of exactly 1.000 looks balanced, but because arrivals and service are uneven the queue length keeps growing. Verdicts use the 3-decimal value shown on screen. Current as of October 2026.
Keeping ρ below 1 is only the minimum condition for a queue not to blow up; it does not mean a latency target is met. The closer ρ gets to 1, the more steeply average waiting time rises. There is no official figure for how much headroom to keep, so it is taken as your own input. Sizing queue storage and draining an existing backlog belong to other tools and are not covered here.
Frequently Asked Questions
Sometimes that value lands on a utilization of exactly 1. With 500 arrivals and 50 served per second, 10 consumers give 1.000 and the queue keeps growing. This page returns the smallest count that stays strictly below 1.
The stability condition holds, but average waiting time rises steeply. It comes from uneven arrivals and service, so adding a little headroom helps latency even when the queue is technically stable.
No. This page only works back to the consumer count; storage sizing and backlog drain time are not supported and are covered by separate calculators.