whirlwind-example

An example repo of how to use xolvio/whirlwind.

To use on AWS Lambda you need to deploy it first:

1. clone this repo

2. run npm install

3. run npm run deploy

4. Customise runTest.js to define your test scenario and load parameters.

And then you can run tests by eg. npm run soak

We created four test types: stress, load, soak and spite.

Stress test

Stress test is a test which starts with some predefined startLoad load and increases the load during the test to a predefined endLoad. Defaults are:

startLoad: 1, // number of requests per second at the beginning of the test
endLoad: 3, // number of requests per second at the end of the test
rampUpTime: 5 // time duration

Load test

Load test is similar to stress test, but it has stationary phases, where the load is kept at the same requests per second. It allows to check if the server is able to hold up with a current level of load. Defaults are:

startLoad: 1, // number of requests per second at the beginning of the test
endLoad: 3, // number of requests per second at the end of the test
rampUpTimePerStep: 15, // time in which load will be increasing during each step/phase of the test
flatDurationPerStep: 40, // time in each step/phase with a constant load
numberOfSteps: 2, // number of step/phases in test, in each step there will be a higher load
rampUpType: 'linear' // defines if each step has linear, or exponential growth

Soak test

Soak test is running on a constant load. It allows to check if the server is stable with a consistent load. It's best to run it for a longer time periods, eg. 8 hours. Defaults are:

load: 1, // number of requests per second during the whole test
duration: 1 // duration of the test in seconds

Spike test

Spike test checks if the server can handle spike loads. It defines some consistentLoad and spikeLoad, which should be considerably higher. After a few spikes of load you can check if any requests were dropped and how long it takes for the server to get back to low response latency on a consisntent load phase.

consistentLoad: 1, // number of requests per second during the consistent load / base load
spikeLoad: 5, // number of requests per second during the spike load
loadTestDuration: 1 * 60, // total test duration in seconds
spikeDuration: 15, // duration of each spike load phase
numberOfSpikes: 1 // number of spike load phases