Loader.io supports three test types: Clients per test, Clients per second, and Maintain client load tests.
Clients per test #
With the "clients per test" type, you specify the total number of clients to connect to your app over the duration of the test. If you specify 20,000 clients for a 20-second test, 1,000 clients will connect each second during that test.
Here is what that looks like for a 20-second, 100-client test. Each arrow represents a client making requests against the server, and the length of the arrow represents the the time it took for the client to get a response from all URLs in the test:
Clients per second #
The "Clients per Second" test type is similar to "Clients per Test", but instead of specifying the total, you specify the number of clients to start each second. A 20-second test with 1,000 clients per second is the same as a 20-second test at 20,000 clients per test.
The graphical representation of this looks similar, too:
Even though you might specify 5 clients per second, there may be more than 5 clients active at any point during the test, especially if server response times are long.
Maintain client load #
This test allows you to specify a
tovalue for clients. If you specify 0 and 10,000 here for example, the test will start with 0 clients and increase up to 10,000 simultaneous clients by the end of the test.
The key difference here is that each client keeps making requests after finishing, instead of stopping. This type of test generally makes a lot more requests. Here is what it looks like in arrow graphs: