Collection of streams for freeloader. They're roughly divided into 3 categories:
-
Emitters
-
Stop conditions
-
Reporters
All these modules are Node.js Transform streams, so you can also easily create your own.
Emits count requests for each incoming request.
Example:
emit(r)
.pipe(times(5))
.send()Emits count requests per second for each incoming request.
This emitter only stops when you press Ctrl-C or when a downstream module requests a shutdown.
Example:
emit(r)
.pipe(perSecond(10))
.send()Note: the emitter can push out thousands of requests per second, but you will most likely be limited by the local network bottlneck.
Maintains count requests in flight for each incoming request.
This emitter only stops when you press Ctrl-C or when a downstream module requests a shutdown.
This is the equivalent of threads in JMeter.
Example:
emit(r)
.pipe(concurrent(50))
.send()Applies the fn function to every incoming request. For example, the function can add headers or modify the payload.
Example:
function randomId(req) {
req.body.myId = Math.floor(Math.random() * 1000);
}
emit(r)
.pipe(times(1000))
.pipe(transform(randomId))
.send()Stop sending any more requests after duration. This module needs to be downstream of any emitting module, since the pause event bubbles up.
duration is a human readable string like 5s, 20s, 3m, 1h.
Example:
emit(r)
.pipe(perSecond(5))
.pipe(stopTimer('10s'))
.send()Note: this does not terminate the pipeline immediately. It simply asks upstream modules to stop sending requests. The shutdown can take a few seconds if modules are still waiting for responses to arrive (ex: consoleSummary).
Shuts down the pipeline after count requests have gone through. This module needs to be downstream of any emitting module, since the pause event bubbles up.
duration is a human readable string like 5s, 20s, 3m, 1h.
Example:
emit(r)
.pipe(perSecond(5))
.pipe(stopCount(30))
.send()Note: this does not terminate the pipeline immediately. It simply asks upstream modules to stop sending requests. The shutdown can take a few seconds if modules are still waiting for responses to arrive (ex: consoleSummary).
Prints every request and response as they arrive. This is useful for debugging, but usually too verbose for actual load tests.
Example:
emit(r)
.pipe(print())
.send()Prints a dot for every request going through, as a way to track progress.
Example:
emit(r)
.pipe(requestDots())
.send()For every response that comes back, prints a green 'o' (success) or a red 'x' (failure).
Example:
emit(r)
.pipe(responseDots())
.send()Prints the state of the test to the console every millis milliseconds. This includes total requests count, response count, and number of requests in flight.
Example:
emit(r)
.pipe(periodicSnapshot(1000))
.send()Prints useful statistics to the console once all the responses have arrived, including the average response times.
Example:
emit(r)
.pipe(consoleSummary())
.send()Similar to consoleSummary, but prints the statistics to a file. This is useful to integrate into a CI pipeline.
Example:
emit(r)
.pipe(jsonSummary('test-report.json'))
.send()Prints bar charts to the console once the test has finished (response time distribution, ...).
Example:
emit(r)
.pipe(consoleCharts())
.send()Calls the fn function once the test is finished. The function will be called with an Error if any of the requests failed.
Example:
function done(err) {
console.log(err ? ('Test failed: ' + err) : 'Success');
}
emit(r)
.pipe(callback(done))
.send()Promise-like API for when the test has finished. Will call fnSuccess if all the requests were successful, and fnSuccess if there were any errors.
Example:
function success() {
console.log('Success!');
}
function failure() {
console.log('Failure:', err);
}
emit(r)
.pipe(then(success, failure))
.send()