On 11/11/2009, at 3:29 PM, Duane Wessels wrote: > > > On Wed, 4 Nov 2009, Nicholas Weaver wrote: > >> Also, has someone done a study what the major recursive resolvers do on >> response failures from a root? Do they go to another first or do they try a >> smaller EDNS MTU? > > I gave a presentation on this at the DNS-OARC meeting last week: > > https://www.dns-oarc.net/files/workshop-200911/Duane_Wessels.pdf > > I was only able to test BIND (9.4.3) and Unbound (1.3.3) before the > workshop. > > I've since learned that since my graphs only show 7 seconds after > the initial query, it misses Unbound's fallback to TCP, which > takes longer than that.
Great presentation. I encourage everyone to read this document. Duane shows an incredibly high packet amplification in some circumstances, of 30 or more subsequent (sys/glue) queries each having to hunt to the optimal MTU. And the total traffic count rises as well of course. yes, its a corner case. Alas, its a corner case we can predict will exist for disadvantaged people behind broken middleware and old software. -George _______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop