On Aug 7, 2007, at 2:47 PM, Rui Ferreira wrote: > So why not modify bind source code to directly parse a big list of > IP's and return a different group of 14 or so IP's in each query, > and also start using a lower TTL?
I gather that you didn't realize that BIND returns a different group of IPs with each query already; if you run "dig 0.pool.ntp.org" several times against a (properly functioning) nameserver, you should get a different set of IPs each time. The problem isn't with BIND being incapable of performing DNS round- robin and properly rotating multiple IPs which match a query, the problem is with stupid DNS client implementations found in Windows and in hardware appliances like DSL modems and the like which tend to cache only the first answer they get, and use that without paying attention to the other return values in a round-robin fashion. Using a lower TTL might help a little, but the same stupid client resolvers which have problems with DNS round-robin tend to enforce a fairly high minimum TTL value which prevents a short TTL set by the authoritative servers from doing much good with respect to the clients which cause the problem. > Or maybe setting 'pool.ntp.org' with a round-robin list of CNAME's > with very low TTL instead of A records, pointing to > '0.pool.ntp.org', '1.pool.ntp.org' and so on, and these ones having > a higher TTL list of 14 IP's, this would spare bandwith in dns > traffic as 'N.pool.ntp.org' list of 14 IP's would remain cached > with the high TTL and the authoritative dns servers would mostly > handle and repond to queries returning the CNAME's instead of 14 IP's? Returning CNAMES rather than A records would significantly increase the DNS traffic load with almost zero benefit; either they'd return results in the same zone and include the IP info as "glue", but only be able to fit half as many IPs per query, or they'd return CNAMES pointing to another zone which would require clients to perform extra queries to resolve the CNAMES into IPs. -- -Chuck _______________________________________________ timekeepers mailing list [email protected] https://fortytwo.ch/mailman/cgi-bin/listinfo/timekeepers
