Re: draft-varlashkin-router-conv-bench-00
iLya <ilya <at> nobulus.com>
2011-12-13 16:04:57 GMT
Hi Jan,
--------------------------------------------------
> Hi,
>
>>This are two main categories who will execute tests. So it looks like
> it's >reasonable to assume that testers will have means to execute the
> tests >according to relevant scope. Would you agree?
>
> Yes, it was just the wording in the text with just the physical
> devices, otherwise yes, it gives the freedom to do what everybody
> needs to do.
>
If you or someone else have an alternative wording, I'll be happy to
consider it. Original text was not meant to confuse anyone but it is
probably result of me not being native english speaker.
>> What do you (and others in BMWG) think about described approach, would
> that address your concerns?
>
> 5.4 Explicitly requires a BGP exchange - the non-DUT has to initiate
> a BGP exchange and that may depend on the non-DUT - maybe easiest this
> could be circumvented by requiring the other BGP peer is from the same
> vendor/same implementation or tested always with the same one
> when comparing two DUT implementation ?? Although maybe it is covered
> already by the requirement the update is immediate - so as long as this
> is considered it is ok.
>
(Continue reading)