Hi,
sorry if this is discussed somewhere but the search results I read through often only perpherally tough this but I couldn't find a clear answer.
So I know some QoS mechanisms depend on the overhead figure more than others but I wonder how it is relevant at all to be honset.
So most routers here are not modems right? So they sit behind some sort of modem or ONT and we fill those bandwidth fields based on speedtest results (the general consensus seems to be to set it somewhere around 90-95% of the slowest measued speedtest result).
And this is what I dont get. Whatever overhead the WAN side protocols may or may not cause: they were present in the speedtest anyway and effectively subtracted in the result, weren't they? And "WAN" for most routers here is also just an ethernet connection so a modem, which is handelind the doscis/DSL/whatever part, right? Now I'm not a network guy but isn't that kind of encapsulation (though not neccessarily in terms of bandwidth) the whole point of the OSI layers?
sorry if this is discussed somewhere but the search results I read through often only perpherally tough this but I couldn't find a clear answer.
So I know some QoS mechanisms depend on the overhead figure more than others but I wonder how it is relevant at all to be honset.
So most routers here are not modems right? So they sit behind some sort of modem or ONT and we fill those bandwidth fields based on speedtest results (the general consensus seems to be to set it somewhere around 90-95% of the slowest measued speedtest result).
And this is what I dont get. Whatever overhead the WAN side protocols may or may not cause: they were present in the speedtest anyway and effectively subtracted in the result, weren't they? And "WAN" for most routers here is also just an ethernet connection so a modem, which is handelind the doscis/DSL/whatever part, right? Now I'm not a network guy but isn't that kind of encapsulation (though not neccessarily in terms of bandwidth) the whole point of the OSI layers?
