Using the Python Fuel SDK, we'd like to maximize our throughput with parallel processing (multithreading or multiprocessing). Are there any hard limits on the number of simultaneous connections imposed by the Fuel SDK or Marketing Cloud?
As far as I know there are no documented or known limitations, however, our team says the threshold is about 4 threads. Any more than that and things start getting unstable. Of course it depends on the operation.
Perhaps the best course of action would be to slowly increase the number of threads until they start timing out. I'd certainly be careful to not go crazy with it. Bad things can happen.