Are there any analytic or simulation models of DNS traffic and caches?
Say I have a DNS cache that handles two different kinds of traffic,
DNSBL lookups that are almost never reused, and web page lookups that
are frequently reused. Is there a model that will predict whether
partitioning the cache would be a good idea, or capping TTL on the
DNSBLs, or other sorts of tricks?
Pointers are fine. TIA.
R's,
John