Understanding LPS - More Than Just a Load Testing Tool

See this scenario and think how you could do it using existing load testing tools:

✅ Send a large number of requests in one burst
⏱ Wait a bit like for 5 seconds
🔁 Repeat the same burst and wait again
⏱ Repeat that operation for 5 mins
❌ Or stop the test based on a specific condition

How straight forward to achieve this using the tools you know? Not that straight, right?

✨ So I thought: what if I could build a tool that gives you full control with a simple structure?

That's how the LPS structure came to life:
Clients + Rounds + Iterations + Iteration Modes + Termination Rules

Each one of these components plays a role:

👥 Clients

These are your virtual users. Each client runs a group of iterations. You can define how many clients to launch and when — like "start a new client every 3 seconds" — to simulate gradual user entry into the system.

🔁 Rounds

Each Round represents a phase of your test. You might have a Round with 100 users, another with 1000 — and each round can contain multiple iterations, which can be executed in parallel or in flow (sequence) depending on the behavior you want.

🔄 Iterations

This is the heart of a round. Each iteration defines the target endpoint and a specific execution behavior, like:

Send requests with a fixed number
Send requests for a fixed duration
Send requests in bursts

🧠 Iteration Modes

The mode is what makes each iteration smart:

D: Run for a specified duration
R: Run for a fixed number of requests
DCB: Burst + Cooldown + Repeat
CB: Continuous bursts until manually stopped or through a termination rule
...and more!

Each mode gives you a different load shape — with zero scripting.

⏹️ Termination Rules

This will complete the picture. You'll be able to say things like:

"Stop the test if error rate reaches 5% for 3 seconds in a row - This feature enables you to define what status codes are considered error status codes"

That means your tests will become smart, dynamic, and outcome-aware — not just time-based.

🎯 Why does this matter?

When you combine all these components:

Clients = simulate real users
Rounds = simulate stages of your test and load ramp-up
Iterations = simulate specific user behaviors or request patterns and defines the target endpoint
Iteration Modes = simulate realistic timing and burst patterns
Termination Rules = allow advanced, intelligent test control

🔥 It unlocks the ability to build almost any load testing scenario you can imagine...

From the simplest burst load test to highly sensitive API stress scenarios — all of it can be modeled in YAML, or even without YAML, using the CLI directly.