On 20/12/2012 15:35, Eric Osterweil wrote:
>
> On Dec 20, 2012, at 11:59 AM, Arturo Servin wrote:
>
>> "> I would really like to echo Chris's last paragraph here. What do you
>>> think is a reasonable time to propagate from an operator editing the
>>> RPKI (A) -> 99.9% of Bs?
>>>
>>> I understand that half a day is way too long. Instantaneous is
>>> theoretically impossible when BGP and RPKI are separate. But is there
>>> really no reasonable pragmatic indication of what would be 'good
>>> enough' for the real world? E.g. if we can come up with a structure
>>> that enables repositories to support 100k RP tools ('B', assuming 2
>>> gatherers per ASN) getting their *updates* (full dump separate thread
>>> please) every 10 minutes, is that good enough?
>>>
>>
>> So, instead of continuing a futile discussion, why not better start
>> working on a workable requirement?
>
> I think a number of us in this ``futile'' argument have been pushing pretty
> hard to start (or some may claim revive) the requirements process. The very
> fact that we are even talking about freshness and liveliness of data is a
> result of a few of us having to request/re-request/write drafts/do formal
> analysis on existing measurement/swrite tech notes/re-request again/etc.
> imho, the only reason anyone is _now_ saying, ``half a day is way too long''
> is because it has taken this long to get traction from the group.
>
That's is not true. We have seen some challenges in the current
architecture since long ago and some are trying to address them:
https://datatracker.ietf.org/doc/draft-rogaglia-sidr-multiple-publication-points/
https://datatracker.ietf.org/doc/draft-tbruijnzeels-sidr-delta-protocol/
https://datatracker.ietf.org/doc/draft-tbruijnzeels-sidr-validation-local-cache/
> My 0.02 about the above is that the first part of Tim's para talks about
> finding a requirement, but the 2nd part presumes the existence of a design
> (repos and structures)...
Yes. That is what we have now. If you have a better way I would like to
hear a proposal.
Let's talk about requirements first. If we need use cases and threat
models before that, then let's go _there_ first, and then on to
requirements. I feel like your comments above take for granted the fact
that several of us have had to work very hard to get to this stage.
>
Let's talk about requirements, what do you think is a reasonable time to
propagate?
> Eric
>
Regards,
as
_______________________________________________
sidr mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/sidr
