11.7. Simulating Media Loads
The bandwidth requirements of a VoIP call can be calculated by adding overhead factors such as RTP and IP headers and Ethernet framing to the size of the actual audio frame. That gives you the packet size , which you can multiply by packet rate in order to figure out the bandwidth required for the call (this was expanded upon significantly in Chapter 6). But that's all academic.
In a real-world scenario, you may need to fill, or nearly fill, a call path with traffic in order to test its behavior under heavy loads. By tweaking the amount of load you place on the call path , you can figure out the failure thresholds for it. Use IPerf to simulate a fixed load, and continually increase it, recording the call quality and completion rate each time you do.
Once you hit the performance ceiling, you should be able to state that a particular link will carry no more than x G.711 calls or that a certain call path through the network will carry no more than y G.729A calls. You would almost certainly have to do this kind of simulation during off-hours.
But how can you account for the bandwidth consumption of non-VoIP applications on the network? One way is to measure it during regular business hours, and then add it into your IPerf simulations.
Use your managed switches or MRTG (multi-router traffic grapher ) to record traffic levels during regular hours. Then, later, use an IPerf client/server pair to simulate that traffic during off-hours. Use a second IPerf client/server pair to simulate VoIP traffic, and continually raise the traffic level, until your test call breaks down or fails.