DNSIND
97.08.11
Munich
DynUpd Testing
o Insecure Dynamic Update being coordinated by map
o Secure Dynamic Update being coordinated by Olafur
o soliciting implementations, send mail to
[email protected] or
[email protected]
o plan to run over the open internet
o prefer authors
o private mailing list
o methodology being developed. based on old mail to list
o three insecure servers known: BIND, American Internet, Keio Univ (thinks
it is complete)
o some more clients
o TIS is working on a secure server
Notify Testing
o
[email protected] is coordinating testing
o in BIND 8.1.1 4.9.6, IBM OS/2 has notify
o Andreas has a primary server
o will write up methodology
IXFR Testing
o Ohta-san coordinating testing, but too busy
o implementations: IBM OS/2, MetaInfo on NT, possibly Inria too
o no methodolgy yet, will write up
draft-ietf-dnsind-tsig-01.txt
o making specification more strict
o TCP connection support
o included timer values included in digest
o error generation model is now specified
o will come out soon after dnssec release
o Andreas has tsig-01, and will update to 02 version
o secret exchange should be a separate draft
o we should have a WG last call
draft-ietf-dnsind-ncache-04.txt
o no one here to speak for it
o it needs work, spec getting closer but language problems
draft-eastlake-kitchen-sink-02.txt
o recent criticism regarding ASN.1 examples will be cleaned up
o if too successful, there may be a problem of many parts of the type then
returning too many data, might resolve by bucket hashing
draft-ietf-dnsind-test-tlds-01.txt
o reserved names for testing and examples would be useful
o some opinion that one is sufficient and english based ones might be ill
advised
o suggestion that we look at the ISO reserved list
o suggest WG last call
draft-watson-dns-error-00.txt
o ran out of RCODEs while doing tsig work
o hence reserve one for indicating more error codes
o need for more detail by some protocols, e.g. dynamic update also needs
error detail
o need a time indicator for when generator does not know what time it is
Domain name compression in new RR types
o description of the problem
o recommends that ranges of RR numbers for compressable and uncompressable
o dnssec's NXT RR bitmap would make this difficult
o alternative: new style compression which is local to the RR, not for the
entire message
draft-eastlake-local-names-00.txt
o see the draft
o has been no significant objection
-30-