Users

In Cicada, a User is the smallest unit of testing. You can think of a user as the behavior of a single client accessing your site. Let's take a look at our example function:

@scenario(engine)
def my_first_test(context):
    assert 2 + 2 == 4

    return "Passed!"

During a test, Cicada will create a container for each user which will run the function my_first_test inside of a User Loop. In this case, a user will check that 2 + 2 = 4.

User Loop

The user loop is responsible for calling the test function in a user container and sending back results.

By default, the User Loop will only run the function if it has work. Work is used to divide iterations across the pool of users running for a particular scenario. If a user does not have work, it will wait until work is available or until is is shut down by its parent scenario. This is effectively how the default loop works:

def while_has_work(user_commands: UserCommands, context: dict):
    while True:
        if user_commands.has_work():
            my_first_test(context)
            # report output, exception, logs, time taken

In some cases, it may be useful to override the user loop. This pseudocode illustrates running the user so it keeps making requests until it is shut down without waiting for work:

from cicadad.core.decorators import scenario
...

def while_has_work(user_commands: UserCommands, context: dict):
    while True:
        my_first_test(context)
        # report output, exception, logs, time taken

@scenario(engine)
@user_loop(while_has_work)
def my_first_test(context):
    assert 2 + 2 == 4

    return "Passed!"

...

tip

This type of user loop is built-in at cicadad.core.scenario.while_alive. Don't write your own version of this!

UserCommands

You may have noticed that the User Loop is provided with an argument called UserCommands. This is a class available to the user to interact with the test. It has the following primary functions:

• Checking for work
• Running the test function
• Sending results to the parent scenario

See UserCommands for more information.

Context

Also available to a User is the Context. This is a dictionary containing all the results of previous run scenarios. This is the basic structure of the context object:

{
    "scenario_a": {
        "id": ...,
        "output": ...,
        "exception": ...,
        "logs": ...,
        "time_taken": ...,
        "timestamp": ...
    },
    "scenario_b": ...
}

Organizing scenarios to run before others can be done by marking them with the @dependency decorator.