What is Little's Law?

Little's Law states that in any stable system, the average number of items in the system (L) equals the average arrival rate (λ) multiplied by the average time each item spends in the system (W). Written as L = λ × W. For APIs: L is concurrent requests, λ is throughput (requests per second), and W is average latency in seconds.

When does this formula break down?

Little's Law assumes a stable, ergodic system where arrival rate equals departure rate (no queue build-up). It breaks down when the system is overloaded (queue grows without bound), during bursts, or when latency distribution is heavily skewed by outliers. In those cases the law still holds on long-run averages, but snapshot readings can be misleading.

What units should I use?

The formula requires consistent units: if throughput is in requests/second, latency must be in seconds, and concurrency will be in requests. This tool converts ms to seconds automatically when you select milliseconds for latency. Always check that your monitoring tools report the same unit before plugging in real data.

How do I use this for thread pool sizing?

Set the target throughput (req/s) and your expected p99 latency (in seconds), then solve for L — that is the minimum thread pool size needed to sustain the target without queuing. Add a headroom multiplier (typically 1.2×–1.5×) for burst traffic and GC pauses.

What is the difference between throughput and RPS?

They are the same concept: throughput (λ in queuing theory) is the rate at which work items arrive or complete. Requests per second (RPS) is simply the common engineering label for throughput in HTTP contexts. You can also use requests per minute (RPM) — this tool converts it for you.

Throughput Latency Helper

Name: Throughput Latency Helper
Availability: InStock
Author: Nham Vu

Enter any two of throughput, latency, or concurrency and instantly derive the third using Little's Law — the foundational queuing-theory formula for API and system performance.

Inputs

Select the variable to solve for, then fill in the other two.

Solve for

Throughput (λ)

Latency (W)

Concurrency (L)

requests

Quick Scenarios

Fill in two values to compute the third.

Little's Law — L = λ × W

Concurrency (L)

—

concurrent requests

Variable	Symbol	Value
Concurrency	L	—
Throughput	λ	—
Latency	W	—

Copied!

Summary

Enter any two of throughput, latency, or concurrency and instantly derive the third using Little's Law — the foundational queuing-theory formula for API and system performance.

How it works

Select which variable you want to calculate: Concurrency (L), Throughput (λ), or Latency (W).
Enter values for the other two variables using the input fields.
Choose the correct units for each value (ms/s for latency, req/s or req/min for throughput).
The result is computed instantly and displayed with a plain-language explanation.
Use the scenario buttons to pre-fill common API benchmarking examples.
Copy the result or the full formula breakdown to share with your team.

Use cases

Estimating the number of concurrent connections a web server must handle at a given request rate and latency.
Determining the maximum throughput achievable before a service becomes overloaded.
Capacity planning: how much latency budget remains before concurrency limits are breached.
Debugging API slowdowns by checking whether observed concurrency matches expected values.
Load-test planning: deriving ramp targets from SLA latency and target RPS.
Teaching queuing theory fundamentals to engineers new to distributed systems.
Validating k6, Gatling, or Locust load-test parameters before a test run.
Sizing thread pools and connection pools for backend services.

Frequently Asked Questions

Related tools

Network Speed Calculator Http Status Code Lookup Regex Tester

Last updated: 2026-05-23 · Reviewed by Nham Vu