If our operators add streams in large batches without considering server load, what will happen to the probe/statistics server?

Let’s assume you added more streams or configured more complex analysis parameters than your hardware can handle. In this case, the probe won’t be able to evaluate data accurately, but our cloud service will remain unaffected as it's designed for intense workload. If you have your own data aggregation server, it might start missing input data. Still, this scenario is unlikely: We provide hardware recommendations, and typically, no one adds more tasks than they have licenses.