Re: [tsvwg] Switch testing at 25G with ECN --SCE Draft

[Wesley Eddy <[e-mail]>]

On 8/9/2019 8:02 PM, Rodney W. Grimes wrote:
>> Right.  Per the SCE method, the switch would mark ECT(1) using a ramp, and then when the ramp would exceed 100% marking, it would change to using CE.
>>
>> The implementation discussed here marks CE using a ramp, and when the ramp would exceed 100% marking, it changes to packet drop.  This, as you said, is simply RFC3168 ECN marking, not SCE ECN marking.
> Purely due to a choice to an expedient experiment.  This is not "simply RFC3168 ECN marking" which would occur at the point the switch was about to start dropping, it is "linear ramp response proportionaly to queue depth SCE marking" using the CE bit pattern.  I can not get into the details as to why this was done this way.

The advice in 3168 is to do marking based on average queue length, but 
*not* just before it needs to start dropping.  Is it more correct 
instead to understand what you're saying as the experiment being based 
on using the instantaneous queue length rather than average?


> RFC3168 ECN marking is already in the switch, the switch was modified to behave in a different manner, using the
> RFC3168 CE bits.  This different manner was turned OFF to do the dctcp tests and turned ON to do the dctcp-sce tests.

Since the marking probability function and marking decision isn't part 
of 3168/ECN, the description here sounds like the switch is always 
conforming to 3168 behavior, but just using some custom decision logic 
for marking.

Just trying to be clear on what we're really talking about ...