To make the simulation interesting the generator and frameworks have built-in randomness.
Generator
At a random interval the generator will randomise the % distribution of HTTP methods for each enabled provider.
This means that for X seconds one provider may receive POST requests while another provider may have an equal split between all of its enabled methods.
Similarly, the Operations are randomised however this applies to all enabled providers instead of unique distributions per provider.
Provider selection is also random for each request.
Task IDs may be random at times defaulting to either a random string or a random uuid.
Payloads may be created with long secret strings to simulate large requests beyond what the frameworks are willing to accept.
Frameworks
For each received request a framework has a chance to sleep to simulate increased workload/latency.
Moreover, it may return an internal server error to simulate random failures.
Sleep time should be between 500-1500µs.