On Fri, Mar 22, 2013 at 9:33 PM, Oleg Muravskiy <o...@ripe.net> wrote: > > Randy Bush wrote: > > I have, a model that says: > "If you want to publish a ROA, you need to have a CA and you need to > run a publication point" > land this a roa and a CA). > > Wherever did you get that? what is the ratio of hosted LIRs to delegated > today? > > > Although in a hosted model it's possible to run just one CA and one > publication point for all hosted clients, in practice no one from RIRs is > doing that. > In our hosted system we have 1370 CAs, and each CA has it's own publication > point, although they all appear to be in the same rsync tree.
so, this leads to some confusion, but does make the repository one monolithic item to kill/sync-from. It's convenient, right up until it's not :( I'm not sure when that point will arrive, but ... > If the number of CAs is used to estimate the size of a global RPKI > repository (number of objects), then the distinction between hosted and > delegated model doesn't matter. It matters if you want to estimate the > number of different repositories to query. But I don't know what to do with > that number afterwards. I hope that the next step after that estimate is: "real world testing shows pulling a repository/update-set takes X seconds, as we get larger in the number of repositories, we need to keep that in mind" -chris _______________________________________________ sidr mailing list sidr@ietf.org https://www.ietf.org/mailman/listinfo/sidr
