On 06/06/2012 11:56, Gilles Massen wrote: > There are 2 groups of additional sections: > A, C, E, I, J: 2 names, with their v4+v6 addresses > B, D, E, G, H, K, L, M: 4 names, only v4 addresses > > Now I don't want to pick on any implementation, rather get an idea of > the reasoning behind the behaviour... (I do agree that the query is a > bit extreme). And if operators think that this might lead to problems.
Hi Gilles, We operate K-root with NSD, which prefers A glue over AAAA glue. If the client is not ENDS0-capable, then NSD will first fill the additional section with as much A glue as it can. This can sometimes result in the client not getting any AAAA glue. I believe the reasoning for this decision is that if a client isn't ENDS0-capable, then it's a reasonable assumption that it is also not IPv6-capable, so it's better to give it as much A glue as possible. Regards, Anand Buddhdev RIPE NCC _______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop
