Re: DS queries on parents vs. "correct behaviour" in answering

2010-12-04 Thread Mark Andrews

Mark Andrews writes:
> 
> In message <02d001cb93f5$513ca2b0$f3b5e8...@janssen@eurid.eu>, "Peter Janssen
> " 
> writes:
> > When a validating resolver queries the parent of a zone for the DS
> > record(s),
> > and the (child) zone is NOT signed,  the response contains no answer
> > but it does contain NSEC (NSEC3) record(s) in the authority section
> > together with corresponding RRSIG records (parent zone is signed).
> > Would it be considered ok, harmfull, not allowed, (any other word)
> > to include in that answer the NS RRSET for the child zone
> > (obviously without any RRSIG)?
> > 
> > Against RFC? Not specified?
> > Would it break resolvers?  Any or all implementations?
> > 
> > What do you think?
> 
> The server is broken.  The DS records are part of the parent zone
> and the authority section should reflect that.  DNSSEC unaware parent
> servers return referrals to the child zone.  A resolver see such a
> referral is likely to just drop the response and move on to the next
> server.
> 
> I suspect you are asking this because of x.dns.be's answers.  Note
> the anwer is also missing the SOA record required for negative caching
> (RFC 2308).
> 
> Mark

It helps if I have the right type in the question.

; <<>> DiG 9.6.0-APPLE-P2 <<>> foo.be +dnssec @x.dns.be +norec ds
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 37780
;; flags: qr; QUERY: 1, ANSWER: 0, AUTHORITY: 6, ADDITIONAL: 1

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags: do; udp: 4096
;; QUESTION SECTION:
;foo.be.IN  DS

;; AUTHORITY SECTION:
foo.be. 86400   IN  NS  ns6.gandi.net.
foo.be. 86400   IN  NS  ka.quuxlabs.com.
ba141snrnoe1rc9mddgrest23g657rir.be. 600 IN NSEC3 1 1 5 1A4E9B6C 
BB7ONI6L9S8J5E3K6HUQ7C41J1AN85CR NS SOA RRSIG DNSKEY NSEC3PARAM TYPE65534
ba141snrnoe1rc9mddgrest23g657rir.be. 600 IN RRSIG NSEC3 8 2 600 20101207140244 
20101130135115 61344 be. 
ZzvHV36wtbQ9woSfpc6nltz+tPc9GStoiEj4Fux+w70xkroPgjCtXhoY 
jC1uErBEAIKVoMKijb4TbFkssppxTZPvsqqYO3nE6ANWm85pHpP/q9VI 
eMk8RKcopptowjT9opikpvOJnOxlq3zTWBBoUjpyc6ZhJAPun3RPbQg5 Lfw=
040gpts32ds6q6unjgf8eh7bpal1m1ik.be. 600 IN NSEC3 1 1 5 1A4E9B6C 
06JFHM0ATMQQJ2C08HOFHCO313VOSEEG NS DS RRSIG
040gpts32ds6q6unjgf8eh7bpal1m1ik.be. 600 IN RRSIG NSEC3 8 2 600 20101207152009 
20101130151117 61344 be. 
Rk1cwdoDfSo99pNPyBzducYv3CRa4qh3fpQmJifDWCxnR3WIAElwaqrV 
dh9czL06jPBBGFTJLzSYs+jbxmrt/iK3EE7E/0Z+AJiZTMBhO+LOY2YM 
U2sU9SX7/cZvtKvIN73/HI1VegcNrDFCqrJvU9zsaUmDwynLGqolzWBV tGI=

;; Query time: 483 msec
;; SERVER: 2001:678:4::a#53(2001:678:4::a)
;; WHEN: Sun Dec  5 09:06:10 2010
;; MSG SIZE  rcvd: 620

> 
> ; <<>> DiG 9.6.0-APPLE-P2 <<>> foo.be +dnssec @x.dns.be +norec
> ;; global options: +cmd
> ;; Got answer:
> ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 40730
> ;; flags: qr; QUERY: 1, ANSWER: 0, AUTHORITY: 6, ADDITIONAL: 1
> 
> ;; OPT PSEUDOSECTION:
> ; EDNS: version: 0, flags: do; udp: 4096
> ;; QUESTION SECTION:
> ;foo.be.  IN  A
> 
> ;; AUTHORITY SECTION:
> foo.be.   86400   IN  NS  ns6.gandi.net.
> foo.be.   86400   IN  NS  ka.quuxlabs.com.
> ba141snrnoe1rc9mddgrest23g657rir.be. 600 IN NSEC3 1 1 5 1A4E9B6C BB7ONI6L9S8J
> 5E3K6HUQ7C41J1AN85CR NS SOA RRSIG DNSKEY NSEC3PARAM TYPE65534
> ba141snrnoe1rc9mddgrest23g657rir.be. 600 IN RRSIG NSEC3 8 2 600 2010120714024
> 4 20101130135115 61344 be. ZzvHV36wtbQ9woSfpc6nltz+tPc9GStoiEj4Fux+w70xkroPgj
> CtXhoY jC1uErBEAIKVoMKijb4TbFkssppxTZPvsqqYO3nE6ANWm85pHpP/q9VI eMk8RKcopptow
> jT9opikpvOJnOxlq3zTWBBoUjpyc6ZhJAPun3RPbQg5 Lfw=
> 040gpts32ds6q6unjgf8eh7bpal1m1ik.be. 600 IN NSEC3 1 1 5 1A4E9B6C 06JFHM0ATMQQ
> J2C08HOFHCO313VOSEEG NS DS RRSIG
> 040gpts32ds6q6unjgf8eh7bpal1m1ik.be. 600 IN RRSIG NSEC3 8 2 600 2010120715200
> 9 20101130151117 61344 be. Rk1cwdoDfSo99pNPyBzducYv3CRa4qh3fpQmJifDWCxnR3WIAE
> lwaqrV dh9czL06jPBBGFTJLzSYs+jbxmrt/iK3EE7E/0Z+AJiZTMBhO+LOY2YM U2sU9SX7/cZvt
> KvIN73/HI1VegcNrDFCqrJvU9zsaUmDwynLGqolzWBV tGI=
> 
> ;; Query time: 483 msec
> ;; SERVER: 2001:678:4::a#53(2001:678:4::a)
> ;; WHEN: Sun Dec  5 09:00:21 2010
> ;; MSG SIZE  rcvd: 620
> 
> > Thanks.
> > 
> > --Pj.
> > =A0=A0=A0 =
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > Register your .eu domain name and win an iPod touch this X-Mas
> > http://www.winwith.eu
> > ___
> > bind-users mailing list
> > bind-users@lists.isc.org
> > https://lists.isc.org/mailman/listinfo/bind-users
> -- 
> Mark Andrews, ISC
> 1 Seymour St., Dundas Valley, NSW 2117, Australia
> PHONE: +61 2 9871 4742 INTERNET: ma...@isc.org
-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742 INTERNET: ma...@isc.org
___
bind-users mailing list
bind-users@lists.isc.org
https://lists.isc.org/mailman/listinfo/bind-users


Re: DS queries on parents vs. "correct behaviour" in answering

2010-12-04 Thread Mark Andrews

In message <02d001cb93f5$513ca2b0$f3b5e8...@janssen@eurid.eu>, "Peter Janssen" 
writes:
> When a validating resolver queries the parent of a zone for the DS
> record(s),
> and the (child) zone is NOT signed,  the response contains no answer
> but it does contain NSEC (NSEC3) record(s) in the authority section
> together with corresponding RRSIG records (parent zone is signed).
> Would it be considered ok, harmfull, not allowed, (any other word)
> to include in that answer the NS RRSET for the child zone
> (obviously without any RRSIG)?
> 
> Against RFC? Not specified?
> Would it break resolvers?  Any or all implementations?
> 
> What do you think?

The server is broken.  The DS records are part of the parent zone
and the authority section should reflect that.  DNSSEC unaware parent
servers return referrals to the child zone.  A resolver see such a
referral is likely to just drop the response and move on to the next
server.

I suspect you are asking this because of x.dns.be's answers.  Note
the anwer is also missing the SOA record required for negative caching
(RFC 2308).

Mark

; <<>> DiG 9.6.0-APPLE-P2 <<>> foo.be +dnssec @x.dns.be +norec
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 40730
;; flags: qr; QUERY: 1, ANSWER: 0, AUTHORITY: 6, ADDITIONAL: 1

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags: do; udp: 4096
;; QUESTION SECTION:
;foo.be.IN  A

;; AUTHORITY SECTION:
foo.be. 86400   IN  NS  ns6.gandi.net.
foo.be. 86400   IN  NS  ka.quuxlabs.com.
ba141snrnoe1rc9mddgrest23g657rir.be. 600 IN NSEC3 1 1 5 1A4E9B6C 
BB7ONI6L9S8J5E3K6HUQ7C41J1AN85CR NS SOA RRSIG DNSKEY NSEC3PARAM TYPE65534
ba141snrnoe1rc9mddgrest23g657rir.be. 600 IN RRSIG NSEC3 8 2 600 20101207140244 
20101130135115 61344 be. 
ZzvHV36wtbQ9woSfpc6nltz+tPc9GStoiEj4Fux+w70xkroPgjCtXhoY 
jC1uErBEAIKVoMKijb4TbFkssppxTZPvsqqYO3nE6ANWm85pHpP/q9VI 
eMk8RKcopptowjT9opikpvOJnOxlq3zTWBBoUjpyc6ZhJAPun3RPbQg5 Lfw=
040gpts32ds6q6unjgf8eh7bpal1m1ik.be. 600 IN NSEC3 1 1 5 1A4E9B6C 
06JFHM0ATMQQJ2C08HOFHCO313VOSEEG NS DS RRSIG
040gpts32ds6q6unjgf8eh7bpal1m1ik.be. 600 IN RRSIG NSEC3 8 2 600 20101207152009 
20101130151117 61344 be. 
Rk1cwdoDfSo99pNPyBzducYv3CRa4qh3fpQmJifDWCxnR3WIAElwaqrV 
dh9czL06jPBBGFTJLzSYs+jbxmrt/iK3EE7E/0Z+AJiZTMBhO+LOY2YM 
U2sU9SX7/cZvtKvIN73/HI1VegcNrDFCqrJvU9zsaUmDwynLGqolzWBV tGI=

;; Query time: 483 msec
;; SERVER: 2001:678:4::a#53(2001:678:4::a)
;; WHEN: Sun Dec  5 09:00:21 2010
;; MSG SIZE  rcvd: 620

> Thanks.
> 
> --Pj.
> =A0=A0=A0 =
> 
> 
> 
> 
> 
> 
> 
> Register your .eu domain name and win an iPod touch this X-Mas
> http://www.winwith.eu
> ___
> bind-users mailing list
> bind-users@lists.isc.org
> https://lists.isc.org/mailman/listinfo/bind-users
-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742 INTERNET: ma...@isc.org
___
bind-users mailing list
bind-users@lists.isc.org
https://lists.isc.org/mailman/listinfo/bind-users


DS queries on parents vs. "correct behaviour" in answering

2010-12-04 Thread Peter Janssen
When a validating resolver queries the parent of a zone for the DS
record(s),
and the (child) zone is NOT signed,  the response contains no answer
but it does contain NSEC (NSEC3) record(s) in the authority section
together with corresponding RRSIG records (parent zone is signed).
Would it be considered ok, harmfull, not allowed, (any other word)
to include in that answer the NS RRSET for the child zone
(obviously without any RRSIG)?

Against RFC? Not specified?
Would it break resolvers?  Any or all implementations?

What do you think?

Thanks.

--Pj.
    






Register your .eu domain name and win an iPod touch this X-Mas
http://www.winwith.eu
___
bind-users mailing list
bind-users@lists.isc.org
https://lists.isc.org/mailman/listinfo/bind-users