My setup CBQ tests- CBQ Class tests
- I want to give two hosts a certain bounded bandwidth and they have to be able to borrow unused bandwidth from each other.
- CBQ bounded accuracy
- You can bound a class so it can not transmit more than the rate you give it. It works and it's accurate. But it is not perfect, but with some trial and error, you can bound the traffic to each possible rate.
- Dividing the traffic in different classes
- You can divide the traffic in different class so the classes can borrow bandwidth from each other. You can find more info and an overview on this page. These are some pages with more details:
- dividing when you provide the weight parameter
- With the weight parameter, you can manipulate how many traffic flows in wich class, but you have to make sure that the weights are proportional to the rates to get good results.
- dividing when you provide no weight parameter
- You can influence how the tarffic is splitted with the weight parameter. But to make life easy, you can provide no weight parameter to get the same results.
I found some extra information. It's about the CBQ implementation for BSD and it goes about the bounded parameter ans why it's sometimes inaccurate. Full story or small, local copy.
- Filter tests
- I tried different filters to put traffic in the CBQ classes and subclasses. It seems that you have to attach the filters to the qdiscs where the classes belongs. You can use a filter directly to the right class or use different filters to get there. The isolated parameter is not working and disturbes the traffic control so much, that you better don't use it.
- Other traffic generators
- I tried CBQ with different traffic generators. For now, I generated as much traffic as possible in 1 stream (with iperf). So the classes are allways used for 100%. But what happens when you create more then 1 stream or don't have enough traffic to full fill a class? Or if you have 1 class with 10 data streams?
TBF tests Interactive traffic