RE: A6 Query Header and Response.
-Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Thursday, June 26, 2003 10:46 AM To: Sarang Karandikar Cc: [EMAIL PROTECTED] Subject: Re: A6 Query Header and Response. I am trying to write DNS resolver for A6 queries. Does anyone know the exact format in which the A6 query header etc. will be sent/received ? before you proceed, you may want to know that A6 is now experimental (so in real life you would use ). RFC3363 Itojun Yes, I do know that A6 is experimental. I am already supporting . I feel that A6 is not being deployed because not much is known about it. I still feel it will be deployed eventually. When will be next action be taken on deciding this ? But I am not finding the format of A6 replies, for both forward and backward queries. In particular, will the Header, Authority, Additional Record section be modified ? In what way will the new Answer section will le different ? What is the value of DNAME (as CNAME is 5) ? What will be the length of Answer section, and its internal values ? Regards, Sarang. - The IPv6 Users Mailing List Unsubscribe by sending unsubscribe users to [EMAIL PROTECTED]
Re: A6 Query Header and Response.
On Thu, Jun 26, 2003 at 10:56:16AM +0530, Sarang Karandikar wrote: Yes, I do know that A6 is experimental. I am already supporting . I feel that A6 is not being deployed because not much is known about it. I still feel it will be deployed eventually. When will be next action be taken on deciding this ? It also suffers from serious performance and security problems (one DNS request may cause multiple requests to unfold a prefix, and compromising a DNS node may also compromise other nodes because prefixes will be propagated), not to mention other minor problems (deploying might not be easier considering the configuration) See also http://www.faqs.org/rfcs/rfc3364.html Resolving a chain of A6 RRs involves resolving a series of what are almost independent queries, but not quite. Each of these sub-queries takes some non-zero amount of time, unless the answer happens to be in the resolver's local cache already. How many times did we have problems with slow/unreachable DNS servers ? Having multiple DNS involved will dramatically increase this risk. A6 are looking pretty nice, but are they the perfect solution considering the need for reliability/performances/security? - The IPv6 Users Mailing List Unsubscribe by sending unsubscribe users to [EMAIL PROTECTED]
Re: A6 Query Header and Response.
On Thu, 2003-06-26 at 08:38, Xavier Roche wrote: On Thu, Jun 26, 2003 at 10:56:16AM +0530, Sarang Karandikar wrote: Yes, I do know that A6 is experimental. I am already supporting . I feel that A6 is not being deployed because not much is known about it. I still feel it will be deployed eventually. When will be next action be taken on deciding this ? It also suffers from serious performance and security problems ... Yes, yes, yes, the horse is dead for the moment. Stop flogging it. The writer only asked for pointers on the structure of the query to experiment with it. That is, afterall, what *experimental* is for, isn't it? Does anyone have any constructive pointers, or is all we have just untested criticism for lack of experimenting? -- Aaron J. Angel [EMAIL PROTECTED] - The IPv6 Users Mailing List Unsubscribe by sending unsubscribe users to [EMAIL PROTECTED]