Re: [Sidrops] I-D Action: draft-ietf-sidrops-rov-no-rr-01.txt

[Randy Bush <[e-mail]>]

>>   If new RPKI data arrive which invalidate the best path, and the
>>   router did not keep all alternatives, then it MUST issue a route
>>   refresh so those alternatives may be evaluated for best path.
> 
> Just to make sure this is the intended behaviour: When new RPKI data
> arrives that transitions a 'better' route that was invalid to unknown
> or valid, the router will stick to the old preferred route.

i don't think the above was talking about that.  the paragraph above
that in the draft says

   When RPKI data cause one or more paths to be dropped due to ROV,
   those paths MUST NOT be evaluated for best path, but MUST be saved
   (either separately or marked) so they may be reevaluated with respect
   to new RPKI data.

> It only switches when the best path is invalidated.

no.  it'll switch if a ROA comes in which validates a previously
invalidated path, and that path competes successfully with the current
best path.

> Can we simulate or estimate the impact of the specification in section
> 4?

could be a fun measurement project.  but the IMC 2022 dreadline is now,
and we're kinda focused on that.  gladly collaborate with someone on
this new measurement starting in a week+.

randy