I’ve been asked to execute a scaling test for our use case for a fleet of 1k and 10k devices. Our publish model is well established at this point, so I was thinking I could simply use the python module on github and put together a simple multithreaded publisher to achieve the test results. Spinning up a thread for each simulated unit and publishing mock data on our expected cadence.
Two or three days of that kind of testing would be sufficient, I think. Seems pretty straightforward, but I thought I might ping the team there to see if there was a ‘better’ method.
Losant generally does not allow scalability testing on our public cloud. If you require scale testing to prove out a solution, I recommend reaching out to your account manager to arrange for doing so on one of our test environments.
Bear in mind that scalability testing, by its nature, generates a great deal of platform usage and thus incurred costs, especially if you are are not yet under a service agreement with Losant.
Thanks Dylan, that is an understandable position. I’ll put that task on the backlog until we get a formal agreement in place. Thanks for the information.