It's very common, and does not present any problems in operational
practice. For example, from the root zone:
Number of IPv4 hosts: 1209
Number of IPv4 addresses: 1157
Number of IPv6 hosts: 551
Number of IPv6 addresses: 520
Situations where there are more addresses than hosts are not uncommon
either, and often result from either a transition in progress, or a
misunderstanding of what "multiple NS records" means.
hth,
Doug
On 05/03/2013 03:15 PM, John Kristoff wrote:
Most authoritative servers, and presumably most operators, associate a
single A, AAAA or pair of those two to a single NS RR, but I have seen
cases where this is not true.
For instance, representative of the more common configuration,
example.org has an NS RR of b.iana-servers.net and that name in turn has
one A and one AAAA associated with it.
However, imagine b.iana-servers.net actually maps to multiple A RRs.
I've seen cases where PTRs can get out of whack.
I could imagine server selection or round robin algorithms giving
somewhat unpredictable and potentially suboptimal results.
Perhaps even some issues with an suboptimal set of additional records
being returned.
On the other hand... if the NS names are in different zones, perhaps
this adds some reliability.
I'm curious if anyone is aware of, or can envision, any actual problems
or real benefits with this A/AAAA overloading, for a lack of a better
term since I'm not sure what to call it.
Thanks to my friend Ed Lewis who entertained a version of this question
some time back off-list and began his response in his characteristically
delightful way with "Problems get solved more easily when it's a genius
and an idiot working together".
John
_______________________________________________
dns-operations mailing list
[email protected]
https://lists.dns-oarc.net/mailman/listinfo/dns-operations
dns-jobs mailing list
https://lists.dns-oarc.net/mailman/listinfo/dns-jobs
