Network Working Group R. Arends Request for Comments: 4035 Telematica Instituut Obsoletes: 2535, 3008, 3090, 3445, 3655, 3658, R. Austein 3755, 3757, 3845 ISC Updates: 1034, 1035, 2136, 2181, 2308, 3225, M. Larson 3007, 3597, 3226 VeriSign Category: Standards Track D. Massey Colorado State University S. Rose NIST March 2005
Protocol Modifications for the DNS Security Extensions
Status of This Memo
This document specifies an Internet standards track protocol for the Internet community, and requests discussion and suggestions for improvements. Please refer to the current edition of the "Internet Official Protocol Standards" (STD 1) for the standardization state and status of this protocol. Distribution of this memo is unlimited.
Copyright Notice
Copyright (C) The Internet Society (2005).
Abstract
This document is part of a family of documents that describe the DNS Security Extensions (DNSSEC). The DNS Security Extensions are a collection of new resource records and protocol modifications that add data origin authentication and data integrity to the DNS. This document describes the DNSSEC protocol modifications. This document defines the concept of a signed zone, along with the requirements for serving and resolving by using DNSSEC. These techniques allow a security-aware resolver to authenticate both DNS resource records and authoritative DNS error indications.
This document obsoletes RFC 2535 and incorporates changes from all updates to RFC 2535.
Arends, et al. Standards Track [Page 1]
RFC 4035 DNSSEC Protocol Modifications March 2005
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.1. Background and Related Documents . . . . . . . . . . . . 4 1.2. Reserved Words . . . . . . . . . . . . . . . . . . . . . 4 2. Zone Signing . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.1. Including DNSKEY RRs in a Zone . . . . . . . . . . . . . 5 2.2. Including RRSIG RRs in a Zone . . . . . . . . . . . . . 5 2.3. Including NSEC RRs in a Zone . . . . . . . . . . . . . . 6 2.4. Including DS RRs in a Zone . . . . . . . . . . . . . . . 7 2.5. Changes to the CNAME Resource Record. . . . . . . . . . 7 2.6. DNSSEC RR Types Appearing at Zone Cuts. . . . . . . . . 8 2.7. Example of a Secure Zone . . . . . . . . . . . . . . . . 8 3. Serving . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3.1. Authoritative Name Servers . . . . . . . . . . . . . . . 9 3.1.1. Including RRSIG RRs in a Response . . . . . . . 10 3.1.2. Including DNSKEY RRs in a Response . . . . . . . 11 3.1.3. Including NSEC RRs in a Response . . . . . . . . 11 3.1.4. Including DS RRs in a Response . . . . . . . . . 14 3.1.5. Responding to Queries for Type AXFR or IXFR . . 15 3.1.6. The AD and CD Bits in an Authoritative Response. 16 3.2. Recursive Name Servers . . . . . . . . . . . . . . . . . 17 3.2.1. The DO Bit . . . . . . . . . . . . . . . . . . . 17 3.2.2. The CD Bit . . . . . . . . . . . . . . . . . . . 17 3.2.3. The AD Bit . . . . . . . . . . . . . . . . . . . 18 3.3. Example DNSSEC Responses . . . . . . . . . . . . . . . . 19 4. Resolving . . . . . . . . . . . . . . . . . . . . . . . . . . 19 4.1. EDNS Support . . . . . . . . . . . . . . . . . . . . . . 19 4.2. Signature Verification Support . . . . . . . . . . . . . 19 4.3. Determining Security Status of Data . . . . . . . . . . 20 4.4. Configured Trust Anchors . . . . . . . . . . . . . . . . 21 4.5. Response Caching . . . . . . . . . . . . . . . . . . . . 21 4.6. Handling of the CD and AD Bits . . . . . . . . . . . . . 22 4.7. Caching BAD Data . . . . . . . . . . . . . . . . . . . . 22 4.8. Synthesized CNAMEs . . . . . . . . . . . . . . . . . . . 23 4.9. Stub Resolvers . . . . . . . . . . . . . . . . . . . . . 23 4.9.1. Handling of the DO Bit . . . . . . . . . . . . . 24 4.9.2. Handling of the CD Bit . . . . . . . . . . . . . 24 4.9.3. Handling of the AD Bit . . . . . . . . . . . . . 24 5. Authenticating DNS Responses . . . . . . . . . . . . . . . . . 25 5.1. Special Considerations for Islands of Security . . . . . 26 5.2. Authenticating Referrals . . . . . . . . . . . . . . . . 26 5.3. Authenticating an RRset with an RRSIG RR . . . . . . . . 28 5.3.1. Checking the RRSIG RR Validity . . . . . . . . . 28 5.3.2. Reconstructing the Signed Data . . . . . . . . . 29 5.3.3. Checking the Signature . . . . . . . . . . . . . 31 5.3.4. Authenticating a Wildcard Expanded RRset Positive Response. . . . . . . . . . . . . . . . 32
The DNS Security Extensions (DNSSEC) are a collection of new resource records and protocol modifications that add data origin authentication and data integrity to the DNS. This document defines the DNSSEC protocol modifications. Section 2 of this document defines the concept of a signed zone and lists the requirements for zone signing. Section 3 describes the modifications to authoritative name server behavior necessary for handling signed zones. Section 4 describes the behavior of entities that include security-aware resolver functions. Finally, Section 5 defines how to use DNSSEC RRs to authenticate a response.
This document is part of a family of documents defining DNSSEC that should be read together as a set.
[RFC4033] contains an introduction to DNSSEC and definitions of common terms; the reader is assumed to be familiar with this document. [RFC4033] also contains a list of other documents updated by and obsoleted by this document set.
The reader is also assumed to be familiar with the basic DNS concepts described in [RFC1034], [RFC1035], and the subsequent documents that update them; particularly, [RFC2181] and [RFC2308].
This document defines the DNSSEC protocol operations.
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in [RFC2119].
DNSSEC introduces the concept of signed zones. A signed zone includes DNS Public Key (DNSKEY), Resource Record Signature (RRSIG), Next Secure (NSEC), and (optionally) Delegation Signer (DS) records according to the rules specified in Sections 2.1, 2.2, 2.3, and 2.4, respectively. A zone that does not include these records according to the rules in this section is an unsigned zone.
DNSSEC requires a change to the definition of the CNAME resource record ([RFC1035]). Section 2.5 changes the CNAME RR to allow RRSIG and NSEC RRs to appear at the same owner name as does a CNAME RR.
DNSSEC specifies the placement of two new RR types, NSEC and DS, which can be placed at the parental side of a zone cut (that is, at a delegation point). This is an exception to the general prohibition against putting data in the parent zone at a zone cut. Section 2.6 describes this change.
To sign a zone, the zone's administrator generates one or more public/private key pairs and uses the private key(s) to sign authoritative RRsets in the zone. For each private key used to create RRSIG RRs in a zone, the zone SHOULD include a zone DNSKEY RR containing the corresponding public key. A zone key DNSKEY RR MUST have the Zone Key bit of the flags RDATA field set (see Section 2.1.1 of [RFC4034]). Public keys associated with other DNS operations MAY be stored in DNSKEY RRs that are not marked as zone keys but MUST NOT be used to verify RRSIGs.
If the zone administrator intends a signed zone to be usable other than as an island of security, the zone apex MUST contain at least one DNSKEY RR to act as a secure entry point into the zone. This secure entry point could then be used as the target of a secure delegation via a corresponding DS RR in the parent zone (see [RFC4034]).
For each authoritative RRset in a signed zone, there MUST be at least one RRSIG record that meets the following requirements:
o The RRSIG owner name is equal to the RRset owner name.
o The RRSIG class is equal to the RRset class.
o The RRSIG Type Covered field is equal to the RRset type.
o The RRSIG Original TTL field is equal to the TTL of the RRset.
o The RRSIG RR's TTL is equal to the TTL of the RRset.
o The RRSIG Labels field is equal to the number of labels in the RRset owner name, not counting the null root label and not counting the leftmost label if it is a wildcard.
o The RRSIG Signer's Name field is equal to the name of the zone containing the RRset.
o The RRSIG Algorithm, Signer's Name, and Key Tag fields identify a zone key DNSKEY record at the zone apex.
The process for constructing the RRSIG RR for a given RRset is described in [RFC4034]. An RRset MAY have multiple RRSIG RRs associated with it. Note that as RRSIG RRs are closely tied to the RRsets whose signatures they contain, RRSIG RRs, unlike all other DNS
Arends, et al. Standards Track [Page 5]
RFC 4035 DNSSEC Protocol Modifications March 2005
RR types, do not form RRsets. In particular, the TTL values among RRSIG RRs with a common owner name do not follow the RRset rules described in [RFC2181].
An RRSIG RR itself MUST NOT be signed, as signing an RRSIG RR would add no value and would create an infinite loop in the signing process.
The NS RRset that appears at the zone apex name MUST be signed, but the NS RRsets that appear at delegation points (that is, the NS RRsets in the parent zone that delegate the name to the child zone's name servers) MUST NOT be signed. Glue address RRsets associated with delegations MUST NOT be signed.
There MUST be an RRSIG for each RRset using at least one DNSKEY of each algorithm in the zone apex DNSKEY RRset. The apex DNSKEY RRset itself MUST be signed by each algorithm appearing in the DS RRset located at the delegating parent (if any).
Each owner name in the zone that has authoritative data or a delegation point NS RRset MUST have an NSEC resource record. The format of NSEC RRs and the process for constructing the NSEC RR for a given name is described in [RFC4034].
The TTL value for any NSEC RR SHOULD be the same as the minimum TTL value field in the zone SOA RR.
An NSEC record (and its associated RRSIG RRset) MUST NOT be the only RRset at any particular owner name. That is, the signing process MUST NOT create NSEC or RRSIG RRs for owner name nodes that were not the owner name of any RRset before the zone was signed. The main reasons for this are a desire for namespace consistency between signed and unsigned versions of the same zone and a desire to reduce the risk of response inconsistency in security oblivious recursive name servers.
The type bitmap of every NSEC resource record in a signed zone MUST indicate the presence of both the NSEC record itself and its corresponding RRSIG record.
The difference between the set of owner names that require RRSIG records and the set of owner names that require NSEC records is subtle and worth highlighting. RRSIG records are present at the owner names of all authoritative RRsets. NSEC records are present at the owner names of all names for which the signed zone is authoritative and also at the owner names of delegations from the
Arends, et al. Standards Track [Page 6]
RFC 4035 DNSSEC Protocol Modifications March 2005
signed zone to its children. Neither NSEC nor RRSIG records are present (in the parent zone) at the owner names of glue address RRsets. Note, however, that this distinction is for the most part visible only during the zone signing process, as NSEC RRsets are authoritative data and are therefore signed. Thus, any owner name that has an NSEC RRset will have RRSIG RRs as well in the signed zone.
The bitmap for the NSEC RR at a delegation point requires special attention. Bits corresponding to the delegation NS RRset and any RRsets for which the parent zone has authoritative data MUST be set; bits corresponding to any non-NS RRset for which the parent is not authoritative MUST be clear.
The DS resource record establishes authentication chains between DNS zones. A DS RRset SHOULD be present at a delegation point when the child zone is signed. The DS RRset MAY contain multiple records, each referencing a public key in the child zone used to verify the RRSIGs in that zone. All DS RRsets in a zone MUST be signed, and DS RRsets MUST NOT appear at a zone's apex.
A DS RR SHOULD point to a DNSKEY RR that is present in the child's apex DNSKEY RRset, and the child's apex DNSKEY RRset SHOULD be signed by the corresponding private key. DS RRs that fail to meet these conditions are not useful for validation, but because the DS RR and its corresponding DNSKEY RR are in different zones, and because the DNS is only loosely consistent, temporary mismatches can occur.
The TTL of a DS RRset SHOULD match the TTL of the delegating NS RRset (that is, the NS RRset from the same zone containing the DS RRset).
Construction of a DS RR requires knowledge of the corresponding DNSKEY RR in the child zone, which implies communication between the child and parent zones. This communication is an operational matter not covered by this document.
If a CNAME RRset is present at a name in a signed zone, appropriate RRSIG and NSEC RRsets are REQUIRED at that name. A KEY RRset at that name for secure dynamic update purposes is also allowed ([RFC3007]). Other types MUST NOT be present at that name.
This is a modification to the original CNAME definition given in [RFC1034]. The original definition of the CNAME RR did not allow any other types to coexist with a CNAME record, but a signed zone
Arends, et al. Standards Track [Page 7]
RFC 4035 DNSSEC Protocol Modifications March 2005
requires NSEC and RRSIG RRs for every authoritative name. To resolve this conflict, this specification modifies the definition of the CNAME resource record to allow it to coexist with NSEC and RRSIG RRs.
DNSSEC introduced two new RR types that are unusual in that they can appear at the parental side of a zone cut. At the parental side of a zone cut (that is, at a delegation point), NSEC RRs are REQUIRED at the owner name. A DS RR could also be present if the zone being delegated is signed and seeks to have a chain of authentication to the parent zone. This is an exception to the original DNS specification ([RFC1034]), which states that only NS RRsets could appear at the parental side of a zone cut.
This specification updates the original DNS specification to allow NSEC and DS RR types at the parent side of a zone cut. These RRsets are authoritative for the parent when they appear at the parent side of a zone cut.
This section describes the behavior of entities that include security-aware name server functions. In many cases such functions will be part of a security-aware recursive name server, but a security-aware authoritative name server has some of the same requirements. Functions specific to security-aware recursive name servers are described in Section 3.2; functions specific to authoritative servers are described in Section 3.1.
In the following discussion, the terms "SNAME", "SCLASS", and "STYPE" are as used in [RFC1034].
A security-aware name server MUST support the EDNS0 ([RFC2671]) message size extension, MUST support a message size of at least 1220 octets, and SHOULD support a message size of 4000 octets. As IPv6 packets can only be fragmented by the source host, a security aware name server SHOULD take steps to ensure that UDP datagrams it transmits over IPv6 are fragmented, if necessary, at the minimum IPv6 MTU, unless the path MTU is known. Please see [RFC1122], [RFC2460], and [RFC3226] for further discussion of packet size and fragmentation issues.
Arends, et al. Standards Track [Page 8]
RFC 4035 DNSSEC Protocol Modifications March 2005
A security-aware name server that receives a DNS query that does not include the EDNS OPT pseudo-RR or that has the DO bit clear MUST treat the RRSIG, DNSKEY, and NSEC RRs as it would any other RRset and MUST NOT perform any of the additional processing described below. Because the DS RR type has the peculiar property of only existing in the parent zone at delegation points, DS RRs always require some special processing, as described in Section 3.1.4.1.
Security aware name servers that receive explicit queries for security RR types that match the content of more than one zone that it serves (for example, NSEC and RRSIG RRs above and below a delegation point where the server is authoritative for both zones) should behave self-consistently. As long as the response is always consistent for each query to the name server, the name server MAY return one of the following:
o The above-delegation RRsets. o The below-delegation RRsets. o Both above and below-delegation RRsets. o Empty answer section (no records). o Some other response. o An error.
DNSSEC allocates two new bits in the DNS message header: the CD (Checking Disabled) bit and the AD (Authentic Data) bit. The CD bit is controlled by resolvers; a security-aware name server MUST copy the CD bit from a query into the corresponding response. The AD bit is controlled by name servers; a security-aware name server MUST ignore the setting of the AD bit in queries. See Sections 3.1.6, 3.2.2, 3.2.3, 4, and 4.9 for details on the behavior of these bits.
A security aware name server that synthesizes CNAME RRs from DNAME RRs as described in [RFC2672] SHOULD NOT generate signatures for the synthesized CNAME RRs.
Upon receiving a relevant query that has the EDNS ([RFC2671]) OPT pseudo-RR DO bit ([RFC3225]) set, a security-aware authoritative name server for a signed zone MUST include additional RRSIG, NSEC, and DS RRs, according to the following rules:
o RRSIG RRs that can be used to authenticate a response MUST be included in the response according to the rules in Section 3.1.1.
Arends, et al. Standards Track [Page 9]
RFC 4035 DNSSEC Protocol Modifications March 2005
o NSEC RRs that can be used to provide authenticated denial of existence MUST be included in the response automatically according to the rules in Section 3.1.3.
o Either a DS RRset or an NSEC RR proving that no DS RRs exist MUST be included in referrals automatically according to the rules in Section 3.1.4.
These rules only apply to responses where the semantics convey information about the presence or absence of resource records. That is, these rules are not intended to rule out responses such as RCODE 4 ("Not Implemented") or RCODE 5 ("Refused").
DNSSEC does not change the DNS zone transfer protocol. Section 3.1.5 discusses zone transfer requirements.
When responding to a query that has the DO bit set, a security-aware authoritative name server SHOULD attempt to send RRSIG RRs that a security-aware resolver can use to authenticate the RRsets in the response. A name server SHOULD make every attempt to keep the RRset and its associated RRSIG(s) together in a response. Inclusion of RRSIG RRs in a response is subject to the following rules:
o When placing a signed RRset in the Answer section, the name server MUST also place its RRSIG RRs in the Answer section. The RRSIG RRs have a higher priority for inclusion than any other RRsets that may have to be included. If space does not permit inclusion of these RRSIG RRs, the name server MUST set the TC bit.
o When placing a signed RRset in the Authority section, the name server MUST also place its RRSIG RRs in the Authority section. The RRSIG RRs have a higher priority for inclusion than any other RRsets that may have to be included. If space does not permit inclusion of these RRSIG RRs, the name server MUST set the TC bit.
o When placing a signed RRset in the Additional section, the name server MUST also place its RRSIG RRs in the Additional section. If space does not permit inclusion of both the RRset and its associated RRSIG RRs, the name server MAY retain the RRset while dropping the RRSIG RRs. If this happens, the name server MUST NOT set the TC bit solely because these RRSIG RRs didn't fit.
When responding to a query that has the DO bit set and that requests the SOA or NS RRs at the apex of a signed zone, a security-aware authoritative name server for that zone MAY return the zone apex DNSKEY RRset in the Additional section. In this situation, the DNSKEY RRset and associated RRSIG RRs have lower priority than does any other information that would be placed in the additional section. The name server SHOULD NOT include the DNSKEY RRset unless there is enough space in the response message for both the DNSKEY RRset and its associated RRSIG RR(s). If there is not enough space to include these DNSKEY and RRSIG RRs, the name server MUST omit them and MUST NOT set the TC bit solely because these RRs didn't fit (see Section 3.1.1).
When responding to a query that has the DO bit set, a security-aware authoritative name server for a signed zone MUST include NSEC RRs in each of the following cases:
No Data: The zone contains RRsets that exactly match <SNAME, SCLASS> but does not contain any RRsets that exactly match <SNAME, SCLASS, STYPE>.
Name Error: The zone does not contain any RRsets that match <SNAME, SCLASS> either exactly or via wildcard name expansion.
Wildcard Answer: The zone does not contain any RRsets that exactly match <SNAME, SCLASS> but does contain an RRset that matches <SNAME, SCLASS, STYPE> via wildcard name expansion.
Wildcard No Data: The zone does not contain any RRsets that exactly match <SNAME, SCLASS> and does contain one or more RRsets that match <SNAME, SCLASS> via wildcard name expansion, but does not contain any RRsets that match <SNAME, SCLASS, STYPE> via wildcard name expansion.
In each of these cases, the name server includes NSEC RRs in the response to prove that an exact match for <SNAME, SCLASS, STYPE> was not present in the zone and that the response that the name server is returning is correct given the data in the zone.
If the zone contains RRsets matching <SNAME, SCLASS> but contains no RRset matching <SNAME, SCLASS, STYPE>, then the name server MUST include the NSEC RR for <SNAME, SCLASS> along with its associated RRSIG RR(s) in the Authority section of the response (see Section 3.1.1). If space does not permit inclusion of the NSEC RR or its associated RRSIG RR(s), the name server MUST set the TC bit (see Section 3.1.1).
Since the search name exists, wildcard name expansion does not apply to this query, and a single signed NSEC RR suffices to prove that the requested RR type does not exist.
If the zone does not contain any RRsets matching <SNAME, SCLASS> either exactly or via wildcard name expansion, then the name server MUST include the following NSEC RRs in the Authority section, along with their associated RRSIG RRs:
o An NSEC RR proving that there is no exact match for <SNAME, SCLASS>.
o An NSEC RR proving that the zone contains no RRsets that would match <SNAME, SCLASS> via wildcard name expansion.
In some cases, a single NSEC RR may prove both of these points. If it does, the name server SHOULD only include the NSEC RR and its RRSIG RR(s) once in the Authority section.
If space does not permit inclusion of these NSEC and RRSIG RRs, the name server MUST set the TC bit (see Section 3.1.1).
The owner names of these NSEC and RRSIG RRs are not subject to wildcard name expansion when these RRs are included in the Authority section of the response.
Note that this form of response includes cases in which SNAME corresponds to an empty non-terminal name within the zone (a name that is not the owner name for any RRset but that is the parent name of one or more RRsets).
3.1.3.3. Including NSEC RRs: Wildcard Answer Response
If the zone does not contain any RRsets that exactly match <SNAME, SCLASS> but does contain an RRset that matches <SNAME, SCLASS, STYPE> via wildcard name expansion, the name server MUST include the
Arends, et al. Standards Track [Page 12]
RFC 4035 DNSSEC Protocol Modifications March 2005
wildcard-expanded answer and the corresponding wildcard-expanded RRSIG RRs in the Answer section and MUST include in the Authority section an NSEC RR and associated RRSIG RR(s) proving that the zone does not contain a closer match for <SNAME, SCLASS>. If space does not permit inclusion of the answer, NSEC and RRSIG RRs, the name server MUST set the TC bit (see Section 3.1.1).
3.1.3.4. Including NSEC RRs: Wildcard No Data Response
This case is a combination of the previous cases. The zone does not contain an exact match for <SNAME, SCLASS>, and although the zone does contain RRsets that match <SNAME, SCLASS> via wildcard expansion, none of those RRsets matches STYPE. The name server MUST include the following NSEC RRs in the Authority section, along with their associated RRSIG RRs:
o An NSEC RR proving that there are no RRsets matching STYPE at the wildcard owner name that matched <SNAME, SCLASS> via wildcard expansion.
o An NSEC RR proving that there are no RRsets in the zone that would have been a closer match for <SNAME, SCLASS>.
In some cases, a single NSEC RR may prove both of these points. If it does, the name server SHOULD only include the NSEC RR and its RRSIG RR(s) once in the Authority section.
The owner names of these NSEC and RRSIG RRs are not subject to wildcard name expansion when these RRs are included in the Authority section of the response.
If space does not permit inclusion of these NSEC and RRSIG RRs, the name server MUST set the TC bit (see Section 3.1.1).
As explained above, there are several situations in which a security-aware authoritative name server has to locate an NSEC RR that proves that no RRsets matching a particular SNAME exist. Locating such an NSEC RR within an authoritative zone is relatively simple, at least in concept. The following discussion assumes that the name server is authoritative for the zone that would have held the non-existent RRsets matching SNAME. The algorithm below is written for clarity, not for efficiency.
To find the NSEC that proves that no RRsets matching name N exist in the zone Z that would have held them, construct a sequence, S, consisting of the owner names of every RRset in Z, sorted into
Arends, et al. Standards Track [Page 13]
RFC 4035 DNSSEC Protocol Modifications March 2005
canonical order ([RFC4034]), with no duplicate names. Find the name M that would have immediately preceded N in S if any RRsets with owner name N had existed. M is the owner name of the NSEC RR that proves that no RRsets exist with owner name N.
The algorithm for finding the NSEC RR that proves that a given name is not covered by any applicable wildcard is similar but requires an extra step. More precisely, the algorithm for finding the NSEC proving that no RRsets exist with the applicable wildcard name is precisely the same as the algorithm for finding the NSEC RR that proves that RRsets with any other owner name do not exist. The part that's missing is a method of determining the name of the non- existent applicable wildcard. In practice, this is easy, because the authoritative name server has already checked for the presence of precisely this wildcard name as part of step (1)(c) of the normal lookup algorithm described in Section 4.3.2 of [RFC1034].
When responding to a query that has the DO bit set, a security-aware authoritative name server returning a referral includes DNSSEC data along with the NS RRset.
If a DS RRset is present at the delegation point, the name server MUST return both the DS RRset and its associated RRSIG RR(s) in the Authority section along with the NS RRset.
If no DS RRset is present at the delegation point, the name server MUST return both the NSEC RR that proves that the DS RRset is not present and the NSEC RR's associated RRSIG RR(s) along with the NS RRset. The name server MUST place the NS RRset before the NSEC RRset and its associated RRSIG RR(s).
Including these DS, NSEC, and RRSIG RRs increases the size of referral messages and may cause some or all glue RRs to be omitted. If space does not permit inclusion of the DS or NSEC RRset and associated RRSIG RRs, the name server MUST set the TC bit (see Section 3.1.1).
The DS resource record type is unusual in that it appears only on the parent zone's side of a zone cut. For example, the DS RRset for the delegation of "foo.example" is stored in the "example" zone rather than in the "foo.example" zone. This requires special processing rules for both name servers and resolvers, as the name server for the child zone is authoritative for the name at the zone cut by the normal DNS rules but the child zone does not contain the DS RRset.
Arends, et al. Standards Track [Page 14]
RFC 4035 DNSSEC Protocol Modifications March 2005
A security-aware resolver sends queries to the parent zone when looking for a needed DS RR at a delegation point (see Section 4.2). However, special rules are necessary to avoid confusing security-oblivious resolvers which might become involved in processing such a query (for example, in a network configuration that forces a security-aware resolver to channel its queries through a security-oblivious recursive name server). The rest of this section describes how a security-aware name server processes DS queries in order to avoid this problem.
The need for special processing by a security-aware name server only arises when all the following conditions are met:
o The name server has received a query for the DS RRset at a zone cut.
o The name server is authoritative for the child zone.
o The name server is not authoritative for the parent zone.
o The name server does not offer recursion.
In all other cases, the name server either has some way of obtaining the DS RRset or could not have been expected to have the DS RRset even by the pre-DNSSEC processing rules, so the name server can return either the DS RRset or an error response according to the normal processing rules.
If all the above conditions are met, however, the name server is authoritative for SNAME but cannot supply the requested RRset. In this case, the name server MUST return an authoritative "no data" response showing that the DS RRset does not exist in the child zone's apex. See Appendix B.8 for an example of such a response.
3.1.5. Responding to Queries for Type AXFR or IXFR
DNSSEC does not change the DNS zone transfer process. A signed zone will contain RRSIG, DNSKEY, NSEC, and DS resource records, but these records have no special meaning with respect to a zone transfer operation.
An authoritative name server is not required to verify that a zone is properly signed before sending or accepting a zone transfer. However, an authoritative name server MAY choose to reject the entire zone transfer if the zone fails to meet any of the signing requirements described in Section 2. The primary objective of a zone transfer is to ensure that all authoritative name servers have identical copies of the zone. An authoritative name server that
Arends, et al. Standards Track [Page 15]
RFC 4035 DNSSEC Protocol Modifications March 2005
chooses to perform its own zone validation MUST NOT selectively reject some RRs and accept others.
DS RRsets appear only on the parental side of a zone cut and are authoritative data in the parent zone. As with any other authoritative RRset, the DS RRset MUST be included in zone transfers of the zone in which the RRset is authoritative data. In the case of the DS RRset, this is the parent zone.
NSEC RRs appear in both the parent and child zones at a zone cut and are authoritative data in both the parent and child zones. The parental and child NSEC RRs at a zone cut are never identical to each other, as the NSEC RR in the child zone's apex will always indicate the presence of the child zone's SOA RR whereas the parental NSEC RR at the zone cut will never indicate the presence of an SOA RR. As with any other authoritative RRs, NSEC RRs MUST be included in zone transfers of the zone in which they are authoritative data. The parental NSEC RR at a zone cut MUST be included in zone transfers of the parent zone, and the NSEC at the zone apex of the child zone MUST be included in zone transfers of the child zone.
RRSIG RRs appear in both the parent and child zones at a zone cut and are authoritative in whichever zone contains the authoritative RRset for which the RRSIG RR provides the signature. That is, the RRSIG RR for a DS RRset or a parental NSEC RR at a zone cut will be authoritative in the parent zone, and the RRSIG for any RRset in the child zone's apex will be authoritative in the child zone. Parental and child RRSIG RRs at a zone cut will never be identical to each other, as the Signer's Name field of an RRSIG RR in the child zone's apex will indicate a DNSKEY RR in the child zone's apex whereas the same field of a parental RRSIG RR at the zone cut will indicate a DNSKEY RR in the parent zone's apex. As with any other authoritative RRs, RRSIG RRs MUST be included in zone transfers of the zone in which they are authoritative data.
3.1.6. The AD and CD Bits in an Authoritative Response
The CD and AD bits are designed for use in communication between security-aware resolvers and security-aware recursive name servers. These bits are for the most part not relevant to query processing by security-aware authoritative name servers.
A security-aware name server does not perform signature validation for authoritative data during query processing, even when the CD bit is clear. A security-aware name server SHOULD clear the CD bit when composing an authoritative response.
Arends, et al. Standards Track [Page 16]
RFC 4035 DNSSEC Protocol Modifications March 2005
A security-aware name server MUST NOT set the AD bit in a response unless the name server considers all RRsets in the Answer and Authority sections of the response to be authentic. A security-aware name server's local policy MAY consider data from an authoritative zone to be authentic without further validation. However, the name server MUST NOT do so unless the name server obtained the authoritative zone via secure means (such as a secure zone transfer mechanism) and MUST NOT do so unless this behavior has been configured explicitly.
A security-aware name server that supports recursion MUST follow the rules for the CD and AD bits given in Section 3.2 when generating a response that involves data obtained via recursion.
As explained in [RFC4033], a security-aware recursive name server is an entity that acts in both the security-aware name server and security-aware resolver roles. This section uses the terms "name server side" and "resolver side" to refer to the code within a security-aware recursive name server that implements the security-aware name server role and the code that implements the security-aware resolver role, respectively.
The resolver side follows the usual rules for caching and negative caching that would apply to any security-aware resolver.
The resolver side of a security-aware recursive name server MUST set the DO bit when sending requests, regardless of the state of the DO bit in the initiating request received by the name server side. If the DO bit in an initiating query is not set, the name server side MUST strip any authenticating DNSSEC RRs from the response but MUST NOT strip any DNSSEC RR types that the initiating query explicitly requested.
The CD bit exists in order to allow a security-aware resolver to disable signature validation in a security-aware name server's processing of a particular query.
The name server side MUST copy the setting of the CD bit from a query to the corresponding response.
The name server side of a security-aware recursive name server MUST pass the state of the CD bit to the resolver side along with the rest
Arends, et al. Standards Track [Page 17]
RFC 4035 DNSSEC Protocol Modifications March 2005
of an initiating query, so that the resolver side will know whether it is required to verify the response data it returns to the name server side. If the CD bit is set, it indicates that the originating resolver is willing to perform whatever authentication its local policy requires. Thus, the resolver side of the recursive name server need not perform authentication on the RRsets in the response. When the CD bit is set, the recursive name server SHOULD, if possible, return the requested data to the originating resolver, even if the recursive name server's local authentication policy would reject the records in question. That is, by setting the CD bit, the originating resolver has indicated that it takes responsibility for performing its own authentication, and the recursive name server should not interfere.
If the resolver side implements a BAD cache (see Section 4.7) and the name server side receives a query that matches an entry in the resolver side's BAD cache, the name server side's response depends on the state of the CD bit in the original query. If the CD bit is set, the name server side SHOULD return the data from the BAD cache; if the CD bit is not set, the name server side MUST return RCODE 2 (server failure).
The intent of the above rule is to provide the raw data to clients that are capable of performing their own signature verification checks while protecting clients that depend on the resolver side of a security-aware recursive name server to perform such checks. Several of the possible reasons why signature validation might fail involve conditions that may not apply equally to the recursive name server and the client that invoked it. For example, the recursive name server's clock may be set incorrectly, or the client may have knowledge of a relevant island of security that the recursive name server does not share. In such cases, "protecting" a client that is capable of performing its own signature validation from ever seeing the "bad" data does not help the client.
The name server side of a security-aware recursive name server MUST NOT set the AD bit in a response unless the name server considers all RRsets in the Answer and Authority sections of the response to be authentic. The name server side SHOULD set the AD bit if and only if the resolver side considers all RRsets in the Answer section and any relevant negative response RRs in the Authority section to be authentic. The resolver side MUST follow the procedure described in Section 5 to determine whether the RRs in question are authentic. However, for backward compatibility, a recursive name server MAY set the AD bit when a response includes unsigned CNAME RRs if those CNAME
Arends, et al. Standards Track [Page 18]
RFC 4035 DNSSEC Protocol Modifications March 2005
RRs demonstrably could have been synthesized from an authentic DNAME RR that is also included in the response according to the synthesis rules described in [RFC2672].
This section describes the behavior of entities that include security-aware resolver functions. In many cases such functions will be part of a security-aware recursive name server, but a stand-alone security-aware resolver has many of the same requirements. Functions specific to security-aware recursive name servers are described in Section 3.2.
A security-aware resolver MUST include an EDNS ([RFC2671]) OPT pseudo-RR with the DO ([RFC3225]) bit set when sending queries.
A security-aware resolver MUST support a message size of at least 1220 octets, SHOULD support a message size of 4000 octets, and MUST use the "sender's UDP payload size" field in the EDNS OPT pseudo-RR to advertise the message size that it is willing to accept. A security-aware resolver's IP layer MUST handle fragmented UDP packets correctly regardless of whether any such fragmented packets were received via IPv4 or IPv6. Please see [RFC1122], [RFC2460], and [RFC3226] for discussion of these requirements.
A security-aware resolver MUST support the signature verification mechanisms described in Section 5 and SHOULD apply them to every received response, except when:
o the security-aware resolver is part of a security-aware recursive name server, and the response is the result of recursion on behalf of a query received with the CD bit set;
o the response is the result of a query generated directly via some form of application interface that instructed the security-aware resolver not to perform validation for this query; or
o validation for this query has been disabled by local policy.
Arends, et al. Standards Track [Page 19]
RFC 4035 DNSSEC Protocol Modifications March 2005
A security-aware resolver's support for signature verification MUST include support for verification of wildcard owner names.
Security-aware resolvers MAY query for missing security RRs in an attempt to perform validation; implementations that choose to do so must be aware that the answers received may not be sufficient to validate the original response. For example, a zone update may have changed (or deleted) the desired information between the original and follow-up queries.
When attempting to retrieve missing NSEC RRs that reside on the parental side at a zone cut, a security-aware iterative-mode resolver MUST query the name servers for the parent zone, not the child zone.
When attempting to retrieve a missing DS, a security-aware iterative-mode resolver MUST query the name servers for the parent zone, not the child zone. As explained in Section 3.1.4.1, security-aware name servers need to apply special processing rules to handle the DS RR, and in some situations the resolver may also need to apply special rules to locate the name servers for the parent zone if the resolver does not already have the parent's NS RRset. To locate the parent NS RRset, the resolver can start with the delegation name, strip off the leftmost label, and query for an NS RRset by that name. If no NS RRset is present at that name, the resolver then strips off the leftmost remaining label and retries the query for that name, repeating this process of walking up the tree until it either finds the NS RRset or runs out of labels.
A security-aware resolver MUST be able to determine whether it should expect a particular RRset to be signed. More precisely, a security-aware resolver must be able to distinguish between four cases:
Secure: An RRset for which the resolver is able to build a chain of signed DNSKEY and DS RRs from a trusted security anchor to the RRset. In this case, the RRset should be signed and is subject to signature validation, as described above.
Insecure: An RRset for which the resolver knows that it has no chain of signed DNSKEY and DS RRs from any trusted starting point to the RRset. This can occur when the target RRset lies in an unsigned zone or in a descendent of an unsigned zone. In this case, the RRset may or may not be signed, but the resolver will not be able to verify the signature.
Arends, et al. Standards Track [Page 20]
RFC 4035 DNSSEC Protocol Modifications March 2005
Bogus: An RRset for which the resolver believes that it ought to be able to establish a chain of trust but for which it is unable to do so, either due to signatures that for some reason fail to validate or due to missing data that the relevant DNSSEC RRs indicate should be present. This case may indicate an attack but may also indicate a configuration error or some form of data corruption.
Indeterminate: An RRset for which the resolver is not able to determine whether the RRset should be signed, as the resolver is not able to obtain the necessary DNSSEC RRs. This can occur when the security-aware resolver is not able to contact security-aware name servers for the relevant zones.
A security-aware resolver MUST be capable of being configured with at least one trusted public key or DS RR and SHOULD be capable of being configured with multiple trusted public keys or DS RRs. Since a security-aware resolver will not be able to validate signatures without such a configured trust anchor, the resolver SHOULD have some reasonably robust mechanism for obtaining such keys when it boots; examples of such a mechanism would be some form of non-volatile storage (such as a disk drive) or some form of trusted local network configuration mechanism.
Note that trust anchors also cover key material that is updated in a secure manner. This secure manner could be through physical media, a key exchange protocol, or some other out-of-band means.
A security-aware resolver SHOULD cache each response as a single atomic entry containing the entire answer, including the named RRset and any associated DNSSEC RRs. The resolver SHOULD discard the entire atomic entry when any of the RRs contained in it expire. In most cases the appropriate cache index for the atomic entry will be the triple <QNAME, QTYPE, QCLASS>, but in cases such as the response form described in Section 3.1.3.2 the appropriate cache index will be the double <QNAME,QCLASS>.
The reason for these recommendations is that, between the initial query and the expiration of the data from the cache, the authoritative data might have been changed (for example, via dynamic update).
Arends, et al. Standards Track [Page 21]
RFC 4035 DNSSEC Protocol Modifications March 2005
There are two situations for which this is relevant:
1. By using the RRSIG record, it is possible to deduce that an answer was synthesized from a wildcard. A security-aware recursive name server could store this wildcard data and use it to generate positive responses to queries other than the name for which the original answer was first received.
2. NSEC RRs received to prove the non-existence of a name could be reused by a security-aware resolver to prove the non-existence of any name in the name range it spans.
In theory, a resolver could use wildcards or NSEC RRs to generate positive and negative responses (respectively) until the TTL or signatures on the records in question expire. However, it seems prudent for resolvers to avoid blocking new authoritative data or synthesizing new data on their own. Resolvers that follow this recommendation will have a more consistent view of the namespace.
A security-aware resolver MAY set a query's CD bit in order to indicate that the resolver takes responsibility for performing whatever authentication its local policy requires on the RRsets in the response. See Section 3.2 for the effect this bit has on the behavior of security-aware recursive name servers.
A security-aware resolver MUST clear the AD bit when composing query messages to protect against buggy name servers that blindly copy header bits that they do not understand from the query message to the response message.
A resolver MUST disregard the meaning of the CD and AD bits in a response unless the response was obtained by using a secure channel or the resolver was specifically configured to regard the message header bits without using a secure channel.
While many validation errors will be transient, some are likely to be more persistent, such as those caused by administrative error (failure to re-sign a zone, clock skew, and so forth). Since requerying will not help in these cases, validating resolvers might generate a significant amount of unnecessary DNS traffic as a result of repeated queries for RRsets with persistent validation failures.
To prevent such unnecessary DNS traffic, security-aware resolvers MAY cache data with invalid signatures, with some restrictions.
Arends, et al. Standards Track [Page 22]
RFC 4035 DNSSEC Protocol Modifications March 2005
Conceptually, caching such data is similar to negative caching ([RFC2308]), except that instead of caching a valid negative response, the resolver is caching the fact that a particular answer failed to validate. This document refers to a cache of data with invalid signatures as a "BAD cache".
Resolvers that implement a BAD cache MUST take steps to prevent the cache from being useful as a denial-of-service attack amplifier, particularly the following:
o Since RRsets that fail to validate do not have trustworthy TTLs, the implementation MUST assign a TTL. This TTL SHOULD be small, in order to mitigate the effect of caching the results of an attack.
o In order to prevent caching of a transient validation failure (which might be the result of an attack), resolvers SHOULD track queries that result in validation failures and SHOULD only answer from the BAD cache after the number of times that responses to queries for that particular <QNAME, QTYPE, QCLASS> have failed to validate exceeds a threshold value.
Resolvers MUST NOT return RRsets from the BAD cache unless the resolver is not required to validate the signatures of the RRsets in question under the rules given in Section 4.2 of this document. See Section 3.2.2 for discussion of how the responses returned by a security-aware recursive name server interact with a BAD cache.
A validating security-aware resolver MUST treat the signature of a valid signed DNAME RR as also covering unsigned CNAME RRs that could have been synthesized from the DNAME RR, as described in [RFC2672], at least to the extent of not rejecting a response message solely because it contains such CNAME RRs. The resolver MAY retain such CNAME RRs in its cache or in the answers it hands back, but is not required to do so.
A security-aware stub resolver MUST support the DNSSEC RR types, at least to the extent of not mishandling responses just because they contain DNSSEC RRs.
A non-validating security-aware stub resolver MAY include the DNSSEC RRs returned by a security-aware recursive name server as part of the data that the stub resolver hands back to the application that invoked it, but is not required to do so. A non-validating stub resolver that seeks to do this will need to set the DO bit in order to receive DNSSEC RRs from the recursive name server.
A validating security-aware stub resolver MUST set the DO bit, because otherwise it will not receive the DNSSEC RRs it needs to perform signature validation.
A non-validating security-aware stub resolver SHOULD NOT set the CD bit when sending queries unless it is requested by the application layer, as by definition, a non-validating stub resolver depends on the security-aware recursive name server to perform validation on its behalf.
A validating security-aware stub resolver SHOULD set the CD bit, because otherwise the security-aware recursive name server will answer the query using the name server's local policy, which may prevent the stub resolver from receiving data that would be acceptable to the stub resolver's local policy.
A non-validating security-aware stub resolver MAY chose to examine the setting of the AD bit in response messages that it receives in order to determine whether the security-aware recursive name server that sent the response claims to have cryptographically verified the data in the Answer and Authority sections of the response message. Note, however, that the responses received by a security-aware stub resolver are heavily dependent on the local policy of the security-aware recursive name server. Therefore, there may be little practical value in checking the status of the AD bit, except perhaps as a debugging aid. In any case, a security-aware stub resolver MUST NOT place any reliance on signature validation allegedly performed on its behalf, except when the security-aware stub resolver obtained the data in question from a trusted security-aware recursive name server via a secure channel.
A validating security-aware stub resolver SHOULD NOT examine the setting of the AD bit in response messages, as, by definition, the stub resolver performs its own signature validation regardless of the setting of the AD bit.
To use DNSSEC RRs for authentication, a security-aware resolver requires configured knowledge of at least one authenticated DNSKEY or DS RR. The process for obtaining and authenticating this initial trust anchor is achieved via some external mechanism. For example, a resolver could use some off-line authenticated exchange to obtain a zone's DNSKEY RR or to obtain a DS RR that identifies and authenticates a zone's DNSKEY RR. The remainder of this section assumes that the resolver has somehow obtained an initial set of trust anchors.
An initial DNSKEY RR can be used to authenticate a zone's apex DNSKEY RRset. To authenticate an apex DNSKEY RRset by using an initial key, the resolver MUST:
1. verify that the initial DNSKEY RR appears in the apex DNSKEY RRset, and that the DNSKEY RR has the Zone Key Flag (DNSKEY RDATA bit 7) set; and
2. verify that there is some RRSIG RR that covers the apex DNSKEY RRset, and that the combination of the RRSIG RR and the initial DNSKEY RR authenticates the DNSKEY RRset. The process for using an RRSIG RR to authenticate an RRset is described in Section 5.3.
Once the resolver has authenticated the apex DNSKEY RRset by using an initial DNSKEY RR, delegations from that zone can be authenticated by using DS RRs. This allows a resolver to start from an initial key and use DS RRsets to proceed recursively down the DNS tree, obtaining other apex DNSKEY RRsets. If the resolver were configured with a root DNSKEY RR, and if every delegation had a DS RR associated with it, then the resolver could obtain and validate any apex DNSKEY RRset. The process of using DS RRs to authenticate referrals is described in Section 5.2.
Section 5.3 shows how the resolver can use DNSKEY RRs in the apex DNSKEY RRset and RRSIG RRs from the zone to authenticate any other RRsets in the zone once the resolver has authenticated a zone's apex DNSKEY RRset. Section 5.4 shows how the resolver can use authenticated NSEC RRsets from the zone to prove that an RRset is not present in the zone.
When a resolver indicates support for DNSSEC (by setting the DO bit), a security-aware name server should attempt to provide the necessary DNSKEY, RRSIG, NSEC, and DS RRsets in a response (see Section 3). However, a security-aware resolver may still receive a response that lacks the appropriate DNSSEC RRs, whether due to configuration issues such as an upstream security-oblivious recursive name server that
Arends, et al. Standards Track [Page 25]
RFC 4035 DNSSEC Protocol Modifications March 2005
accidentally interferes with DNSSEC RRs or due to a deliberate attack in which an adversary forges a response, strips DNSSEC RRs from a response, or modifies a query so that DNSSEC RRs appear not to be requested. The absence of DNSSEC data in a response MUST NOT by itself be taken as an indication that no authentication information exists.
A resolver SHOULD expect authentication information from signed zones. A resolver SHOULD believe that a zone is signed if the resolver has been configured with public key information for the zone, or if the zone's parent is signed and the delegation from the parent contains a DS RRset.
5.1. Special Considerations for Islands of Security
Islands of security (see [RFC4033]) are signed zones for which it is not possible to construct an authentication chain to the zone from its parent. Validating signatures within an island of security requires that the validator have some other means of obtaining an initial authenticated zone key for the island. If a validator cannot obtain such a key, it SHOULD switch to operating as if the zones in the island of security are unsigned.
All the normal processes for validating responses apply to islands of security. The only difference between normal validation and validation within an island of security is in how the validator obtains a trust anchor for the authentication chain.
Once the apex DNSKEY RRset for a signed parent zone has been authenticated, DS RRsets can be used to authenticate the delegation to a signed child zone. A DS RR identifies a DNSKEY RR in the child zone's apex DNSKEY RRset and contains a cryptographic digest of the child zone's DNSKEY RR. Use of a strong cryptographic digest algorithm ensures that it is computationally infeasible for an adversary to generate a DNSKEY RR that matches the digest. Thus, authenticating the digest allows a resolver to authenticate the matching DNSKEY RR. The resolver can then use this child DNSKEY RR to authenticate the entire child apex DNSKEY RRset.
Given a DS RR for a delegation, the child zone's apex DNSKEY RRset can be authenticated if all of the following hold:
o The DS RR has been authenticated using some DNSKEY RR in the parent's apex DNSKEY RRset (see Section 5.3).
Arends, et al. Standards Track [Page 26]
RFC 4035 DNSSEC Protocol Modifications March 2005
o The Algorithm and Key Tag in the DS RR match the Algorithm field and the key tag of a DNSKEY RR in the child zone's apex DNSKEY RRset, and, when the DNSKEY RR's owner name and RDATA are hashed using the digest algorithm specified in the DS RR's Digest Type field, the resulting digest value matches the Digest field of the DS RR.
o The matching DNSKEY RR in the child zone has the Zone Flag bit set, the corresponding private key has signed the child zone's apex DNSKEY RRset, and the resulting RRSIG RR authenticates the child zone's apex DNSKEY RRset.
If the referral from the parent zone did not contain a DS RRset, the response should have included a signed NSEC RRset proving that no DS RRset exists for the delegated name (see Section 3.1.4). A security-aware resolver MUST query the name servers for the parent zone for the DS RRset if the referral includes neither a DS RRset nor a NSEC RRset proving that the DS RRset does not exist (see Section 4).
If the validator authenticates an NSEC RRset that proves that no DS RRset is present for this zone, then there is no authentication path leading from the parent to the child. If the resolver has an initial DNSKEY or DS RR that belongs to the child zone or to any delegation below the child zone, this initial DNSKEY or DS RR MAY be used to re-establish an authentication path. If no such initial DNSKEY or DS RR exists, the validator cannot authenticate RRsets in or below the child zone.
If the validator does not support any of the algorithms listed in an authenticated DS RRset, then the resolver has no supported authentication path leading from the parent to the child. The resolver should treat this case as it would the case of an authenticated NSEC RRset proving that no DS RRset exists, as described above.
Note that, for a signed delegation, there are two NSEC RRs associated with the delegated name. One NSEC RR resides in the parent zone and can be used to prove whether a DS RRset exists for the delegated name. The second NSEC RR resides in the child zone and identifies which RRsets are present at the apex of the child zone. The parent NSEC RR and child NSEC RR can always be distinguished because the SOA bit will be set in the child NSEC RR and clear in the parent NSEC RR. A security-aware resolver MUST use the parent NSEC RR when attempting to prove that a DS RRset does not exist.
Arends, et al. Standards Track [Page 27]
RFC 4035 DNSSEC Protocol Modifications March 2005
If the resolver does not support any of the algorithms listed in an authenticated DS RRset, then the resolver will not be able to verify the authentication path to the child zone. In this case, the resolver SHOULD treat the child zone as if it were unsigned.
A validator can use an RRSIG RR and its corresponding DNSKEY RR to attempt to authenticate RRsets. The validator first checks the RRSIG RR to verify that it covers the RRset, has a valid time interval, and identifies a valid DNSKEY RR. The validator then constructs the canonical form of the signed data by appending the RRSIG RDATA (excluding the Signature Field) with the canonical form of the covered RRset. Finally, the validator uses the public key and signature to authenticate the signed data. Sections 5.3.1, 5.3.2, and 5.3.3 describe each step in detail.
A security-aware resolver can use an RRSIG RR to authenticate an RRset if all of the following conditions hold:
o The RRSIG RR and the RRset MUST have the same owner name and the same class.
o The RRSIG RR's Signer's Name field MUST be the name of the zone that contains the RRset.
o The RRSIG RR's Type Covered field MUST equal the RRset's type.
o The number of labels in the RRset owner name MUST be greater than or equal to the value in the RRSIG RR's Labels field.
o The validator's notion of the current time MUST be less than or equal to the time listed in the RRSIG RR's Expiration field.
o The validator's notion of the current time MUST be greater than or equal to the time listed in the RRSIG RR's Inception field.
o The RRSIG RR's Signer's Name, Algorithm, and Key Tag fields MUST match the owner name, algorithm, and key tag for some DNSKEY RR in the zone's apex DNSKEY RRset.
o The matching DNSKEY RR MUST be present in the zone's apex DNSKEY RRset, and MUST have the Zone Flag bit (DNSKEY RDATA Flag bit 7) set.
Arends, et al. Standards Track [Page 28]
RFC 4035 DNSSEC Protocol Modifications March 2005
It is possible for more than one DNSKEY RR to match the conditions above. In this case, the validator cannot predetermine which DNSKEY RR to use to authenticate the signature, and it MUST try each matching DNSKEY RR until either the signature is validated or the validator has run out of matching public keys to try.
Note that this authentication process is only meaningful if the validator authenticates the DNSKEY RR before using it to validate signatures. The matching DNSKEY RR is considered to be authentic if:
o the apex DNSKEY RRset containing the DNSKEY RR is considered authentic; or
o the RRset covered by the RRSIG RR is the apex DNSKEY RRset itself, and the DNSKEY RR either matches an authenticated DS RR from the parent zone or matches a trust anchor.
Once the RRSIG RR has met the validity requirements described in Section 5.3.1, the validator has to reconstruct the original signed data. The original signed data includes RRSIG RDATA (excluding the Signature field) and the canonical form of the RRset. Aside from being ordered, the canonical form of the RRset might also differ from the received RRset due to DNS name compression, decremented TTLs, or wildcard expansion. The validator should use the following to reconstruct the original signed data:
signed_data = RRSIG_RDATA | RR(1) | RR(2)... where
"|" denotes concatenation
RRSIG_RDATA is the wire format of the RRSIG RDATA fields with the Signature field excluded and the Signer's Name in canonical form.
RR(i) = name | type | class | OrigTTL | RDATA length | RDATA
name is calculated according to the function below
class is the RRset's class
type is the RRset type and all RRs in the class
OrigTTL is the value from the RRSIG Original TTL field
All names in the RDATA field are in canonical form
Arends, et al. Standards Track [Page 29]
RFC 4035 DNSSEC Protocol Modifications March 2005
The set of all RR(i) is sorted into canonical order.
To calculate the name: let rrsig_labels = the value of the RRSIG Labels field
let fqdn = RRset's fully qualified domain name in canonical form
let fqdn_labels = Label count of the fqdn above.
if rrsig_labels = fqdn_labels, name = fqdn
if rrsig_labels < fqdn_labels, name = "*." | the rightmost rrsig_label labels of the fqdn
if rrsig_labels > fqdn_labels the RRSIG RR did not pass the necessary validation checks and MUST NOT be used to authenticate this RRset.
The canonical forms for names and RRsets are defined in [RFC4034].
NSEC RRsets at a delegation boundary require special processing. There are two distinct NSEC RRsets associated with a signed delegated name. One NSEC RRset resides in the parent zone, and specifies which RRsets are present at the parent zone. The second NSEC RRset resides at the child zone and identifies which RRsets are present at the apex in the child zone. The parent NSEC RRset and child NSEC RRset can always be distinguished as only a child NSEC RR will indicate that an SOA RRset exists at the name. When reconstructing the original NSEC RRset for the delegation from the parent zone, the NSEC RRs MUST NOT be combined with NSEC RRs from the child zone. When reconstructing the original NSEC RRset for the apex of the child zone, the NSEC RRs MUST NOT be combined with NSEC RRs from the parent zone.
Note that each of the two NSEC RRsets at a delegation point has a corresponding RRSIG RR with an owner name matching the delegated name, and each of these RRSIG RRs is authoritative data associated with the same zone that contains the corresponding NSEC RRset. If necessary, a resolver can tell these RRSIG RRs apart by checking the Signer's Name field.
Once the resolver has validated the RRSIG RR as described in Section 5.3.1 and reconstructed the original signed data as described in Section 5.3.2, the validator can attempt to use the cryptographic signature to authenticate the signed data, and thus (finally!) authenticate the RRset.
The Algorithm field in the RRSIG RR identifies the cryptographic algorithm used to generate the signature. The signature itself is contained in the Signature field of the RRSIG RDATA, and the public key used to verify the signature is contained in the Public Key field of the matching DNSKEY RR(s) (found in Section 5.3.1). [RFC4034] provides a list of algorithm types and provides pointers to the documents that define each algorithm's use.
Note that it is possible for more than one DNSKEY RR to match the conditions in Section 5.3.1. In this case, the validator can only determine which DNSKEY RR is correct by trying each matching public key until the validator either succeeds in validating the signature or runs out of keys to try.
If the Labels field of the RRSIG RR is not equal to the number of labels in the RRset's fully qualified owner name, then the RRset is either invalid or the result of wildcard expansion. The resolver MUST verify that wildcard expansion was applied properly before considering the RRset to be authentic. Section 5.3.4 describes how to determine whether a wildcard was applied properly.
If other RRSIG RRs also cover this RRset, the local resolver security policy determines whether the resolver also has to test these RRSIG RRs and how to resolve conflicts if these RRSIG RRs lead to differing results.
If the resolver accepts the RRset as authentic, the validator MUST set the TTL of the RRSIG RR and each RR in the authenticated RRset to a value no greater than the minimum of:
o the RRset's TTL as received in the response;
o the RRSIG RR's TTL as received in the response;
o the value in the RRSIG RR's Original TTL field; and
o the difference of the RRSIG RR's Signature Expiration time and the current time.
Arends, et al. Standards Track [Page 31]
RFC 4035 DNSSEC Protocol Modifications March 2005
5.3.4. Authenticating a Wildcard Expanded RRset Positive Response
If the number of labels in an RRset's owner name is greater than the Labels field of the covering RRSIG RR, then the RRset and its covering RRSIG RR were created as a result of wildcard expansion. Once the validator has verified the signature, as described in Section 5.3, it must take additional steps to verify the non- existence of an exact match or closer wildcard match for the query. Section 5.4 discusses these steps.
Note that the response received by the resolver should include all NSEC RRs needed to authenticate the response (see Section 3.1.3).
A resolver can use authenticated NSEC RRs to prove that an RRset is not present in a signed zone. Security-aware name servers should automatically include any necessary NSEC RRs for signed zones in their responses to security-aware resolvers.
Denial of existence is determined by the following rules:
o If the requested RR name matches the owner name of an authenticated NSEC RR, then the NSEC RR's type bit map field lists all RR types present at that owner name, and a resolver can prove that the requested RR type does not exist by checking for the RR type in the bit map. If the number of labels in an authenticated NSEC RR's owner name equals the Labels field of the covering RRSIG RR, then the existence of the NSEC RR proves that wildcard expansion could not have been used to match the request.
o If the requested RR name would appear after an authenticated NSEC RR's owner name and before the name listed in that NSEC RR's Next Domain Name field according to the canonical DNS name order defined in [RFC4034], then no RRsets with the requested name exist in the zone. However, it is possible that a wildcard could be used to match the requested RR owner name and type, so proving that the requested RRset does not exist also requires proving that no possible wildcard RRset exists that could have been used to generate a positive response.
In addition, security-aware resolvers MUST authenticate the NSEC RRsets that comprise the non-existence proof as described in Section 5.3.
To prove the non-existence of an RRset, the resolver must be able to verify both that the queried RRset does not exist and that no relevant wildcard RRset exists. Proving this may require more than
Arends, et al. Standards Track [Page 32]
RFC 4035 DNSSEC Protocol Modifications March 2005
one NSEC RRset from the zone. If the complete set of necessary NSEC RRsets is not present in a response (perhaps due to message truncation), then a security-aware resolver MUST resend the query in order to attempt to obtain the full collection of NSEC RRs necessary to verify the non-existence of the requested RRset. As with all DNS operations, however, the resolver MUST bound the work it puts into answering any particular query.
Since a validated NSEC RR proves the existence of both itself and its corresponding RRSIG RR, a validator MUST ignore the settings of the NSEC and RRSIG bits in an NSEC RR.
5.5. Resolver Behavior When Signatures Do Not Validate
If for whatever reason none of the RRSIGs can be validated, the response SHOULD be considered BAD. If the validation was being done to service a recursive query, the name server MUST return RCODE 2 to the originating client. However, it MUST return the full response if and only if the original query had the CD bit set. Also see Section 4.7 on caching responses that do not validate.
[RFC4034] contains a review of the IANA considerations introduced by DNSSEC. The following are additional IANA considerations discussed in this document:
[RFC2535] reserved the CD and AD bits in the message header. The meaning of the AD bit was redefined in [RFC3655], and the meaning of both the CD and AD bit are restated in this document. No new bits in the DNS message header are defined in this document.
[RFC2671] introduced EDNS, and [RFC3225] reserved the DNSSEC OK bit and defined its use. The use is restated but not altered in this document.
This document describes how the DNS security extensions use public key cryptography to sign and authenticate DNS resource record sets. Please see [RFC4033] for terminology and general security considerations related to DNSSEC; see [RFC4034] for considerations specific to the DNSSEC resource record types.
Arends, et al. Standards Track [Page 33]
RFC 4035 DNSSEC Protocol Modifications March 2005
An active attacker who can set the CD bit in a DNS query message or the AD bit in a DNS response message can use these bits to defeat the protection that DNSSEC attempts to provide to security-oblivious recursive-mode resolvers. For this reason, use of these control bits by a security-aware recursive-mode resolver requires a secure channel. See Sections 3.2.2 and 4.9 for further discussion.
The protocol described in this document attempts to extend the benefits of DNSSEC to security-oblivious stub resolvers. However, as recovery from validation failures is likely to be specific to particular applications, the facilities that DNSSEC provides for stub resolvers may prove inadequate. Operators of security-aware recursive name servers will have to pay close attention to the behavior of the applications that use their services when choosing a local validation policy; failure to do so could easily result in the recursive name server accidentally denying service to the clients it is intended to support.
This document was created from the input and ideas of the members of the DNS Extensions Working Group and working group mailing list. The editors would like to express their thanks for the comments and suggestions received during the revision of these security extension specifications. Although explicitly listing everyone who has contributed during the decade in which DNSSEC has been under development would be impossible, [RFC4033] includes a list of some of the participants who were kind enough to comment on these documents.
The apex DNSKEY set includes two DNSKEY RRs, and the DNSKEY RDATA Flags indicate that each of these DNSKEY RRs is a zone key. One of these DNSKEY RRs also has the SEP flag set and has been used to sign the apex DNSKEY RRset; this is the key that should be hashed to generate a DS record to be inserted into the parent zone. The other DNSKEY is used to sign all the other RRsets in the zone.
The zone includes a wildcard entry, "*.w.example". Note that the name "*.w.example" is used in constructing NSEC chains, and that the RRSIG covering the "*.w.example" MX RRset has a label count of 2.
The zone also includes two delegations. The delegation to "b.example" includes an NS RRset, glue address records, and an NSEC RR; note that only the NSEC RRset is signed. The delegation to "a.example" provides a DS RR; note that only the NSEC and DS RRsets are signed.
A successful query that was answered via wildcard expansion. The label count in the answer's RRSIG RR indicates that a wildcard RRset was expanded to produce this response, and the NSEC RR proves that no closer match exists in the zone.
;; Header: QR AA DO RCODE=0 ;; ;; Question a.z.w.example. IN MX
A "no data" response for a name covered by a wildcard. The NSEC RRs prove that the matching wildcard name does not have any RRs of the requested type and that no closer match exists in the zone.
;; Header: QR AA DO RCODE=0 ;; ;; Question a.z.w.example. IN AAAA
The query in Appendix B.1 returned an MX RRset for "x.w.example.com". The corresponding RRSIG indicates that the MX RRset was signed by an "example" DNSKEY with algorithm 5 and key tag 38519. The resolver needs the corresponding DNSKEY RR in order to authenticate this answer. The discussion below describes how a resolver might obtain this DNSKEY RR.
The RRSIG indicates the original TTL of the MX RRset was 3600, and, for the purpose of authentication, the current TTL is replaced by 3600. The RRSIG labels field value of 3 indicates that the answer was not the result of wildcard expansion. The "x.w.example.com" MX RRset is placed in canonical form, and, assuming the current time falls between the signature inception and expiration dates, the signature is authenticated.
This example shows the logical authentication process that starts from the a configured root DNSKEY (or DS RR) and moves down the tree to authenticate the desired "example" DNSKEY RR. Note that the logical order is presented for clarity. An implementation may choose to construct the authentication as referrals are received or to construct the authentication chain only after all RRsets have been obtained, or in any other combination it sees fit. The example here demonstrates only the logical process and does not dictate any implementation rules.
We assume the resolver starts with a configured DNSKEY RR for the root zone (or a configured DS RR for the root zone). The resolver checks whether this configured DNSKEY RR is present in the root DNSKEY RRset (or whether the DS RR matches some DNSKEY in the root DNSKEY RRset), whether this DNSKEY RR has signed the root DNSKEY RRset, and whether the signature lifetime is valid. If all these
Arends, et al. Standards Track [Page 49]
RFC 4035 DNSSEC Protocol Modifications March 2005
conditions are met, all keys in the DNSKEY RRset are considered authenticated. The resolver then uses one (or more) of the root DNSKEY RRs to authenticate the "example" DS RRset. Note that the resolver may have to query the root zone to obtain the root DNSKEY RRset or "example" DS RRset.
Once the DS RRset has been authenticated using the root DNSKEY, the resolver checks the "example" DNSKEY RRset for some "example" DNSKEY RR that matches one of the authenticated "example" DS RRs. If such a matching "example" DNSKEY is found, the resolver checks whether this DNSKEY RR has signed the "example" DNSKEY RRset and the signature lifetime is valid. If these conditions are met, all keys in the "example" DNSKEY RRset are considered authenticated.
Finally, the resolver checks that some DNSKEY RR in the "example" DNSKEY RRset uses algorithm 5 and has a key tag of 38519. This DNSKEY is used to authenticate the RRSIG included in the response. If multiple "example" DNSKEY RRs match this algorithm and key tag, then each DNSKEY RR is tried, and the answer is authenticated if any of the matching DNSKEY RRs validate the signature as described above.
The query in Appendix B.2 returned NSEC RRs that prove that the requested data does not exist and no wildcard applies. The negative reply is authenticated by verifying both NSEC RRs. The NSEC RRs are authenticated in a manner identical to that of the MX RRset discussed above.
The query in Appendix B.3 returned an NSEC RR that proves that the requested name exists, but the requested RR type does not exist. The negative reply is authenticated by verifying the NSEC RR. The NSEC RR is authenticated in a manner identical to that of the MX RRset discussed above.
The query in Appendix B.4 returned a referral to the signed "a.example." zone. The DS RR is authenticated in a manner identical to that of the MX RRset discussed above. This DS RR is used to authenticate the "a.example" DNSKEY RRset.
Once the "a.example" DS RRset has been authenticated using the "example" DNSKEY, the resolver checks the "a.example" DNSKEY RRset for some "a.example" DNSKEY RR that matches the DS RR. If such a matching "a.example" DNSKEY is found, the resolver checks whether
Arends, et al. Standards Track [Page 50]
RFC 4035 DNSSEC Protocol Modifications March 2005
this DNSKEY RR has signed the "a.example" DNSKEY RRset and whether the signature lifetime is valid. If all these conditions are met, all keys in the "a.example" DNSKEY RRset are considered authenticated.
The query in Appendix B.5 returned a referral to an unsigned "b.example." zone. The NSEC proves that no authentication leads from "example" to "b.example", and the NSEC RR is authenticated in a manner identical to that of the MX RRset discussed above.
The query in Appendix B.6 returned an answer that was produced as a result of wildcard expansion. The answer section contains a wildcard RRset expanded as it would be in a traditional DNS response, and the corresponding RRSIG indicates that the expanded wildcard MX RRset was signed by an "example" DNSKEY with algorithm 5 and key tag 38519. The RRSIG indicates that the original TTL of the MX RRset was 3600, and, for the purpose of authentication, the current TTL is replaced by 3600. The RRSIG labels field value of 2 indicates that the answer is the result of wildcard expansion, as the "a.z.w.example" name contains 4 labels. The name "a.z.w.w.example" is replaced by "*.w.example", the MX RRset is placed in canonical form, and, assuming that the current time falls between the signature inception and expiration dates, the signature is authenticated.
The NSEC proves that no closer match (exact or closer wildcard) could have been used to answer this query, and the NSEC RR must also be authenticated before the answer is considered valid.
The query in Appendix B.7 returned NSEC RRs that prove that the requested data does not exist and no wildcard applies. The negative reply is authenticated by verifying both NSEC RRs.
The query in Appendix B.8 returned NSEC RRs that shows the requested was answered by a child server ("example" server). The NSEC RR indicates the presence of an SOA RR, showing that the answer is from the child . Queries for the "example" DS RRset should be sent to the parent servers ("root" servers).
Arends, et al. Standards Track [Page 51]
RFC 4035 DNSSEC Protocol Modifications March 2005
Authors' Addresses
Roy Arends Telematica Instituut Brouwerijstraat 1 7523 XC Enschede NL
EMail: roy.arends@telin.nl
Rob Austein Internet Systems Consortium 950 Charter Street Redwood City, CA 94063 USA
EMail: sra@isc.org
Matt Larson VeriSign, Inc. 21345 Ridgetop Circle Dulles, VA 20166-6503 USA
EMail: mlarson@verisign.com
Dan Massey Colorado State University Department of Computer Science Fort Collins, CO 80523-1873
EMail: massey@cs.colostate.edu
Scott Rose National Institute for Standards and Technology 100 Bureau Drive Gaithersburg, MD 20899-8920 USA
EMail: scott.rose@nist.gov
Arends, et al. Standards Track [Page 52]
RFC 4035 DNSSEC Protocol Modifications March 2005
Full Copyright Statement
Copyright (C) The Internet Society (2005).
This document is subject to the rights, licenses and restrictions contained in BCP 78, and except as set forth therein, the authors retain all their rights.
This document and the information contained herein are provided on an "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY AND THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Intellectual Property
The IETF takes no position regarding the validity or scope of any Intellectual Property Rights or other rights that might be claimed to pertain to the implementation or use of the technology described in this document or the extent to which any license under such rights might or might not be available; nor does it represent that it has made any independent effort to identify any such rights. Information on the procedures with respect to rights in RFC documents can be found in BCP 78 and BCP 79.
Copies of IPR disclosures made to the IETF Secretariat and any assurances of licenses to be made available, or the result of an attempt made to obtain a general license or permission for the use of such proprietary rights by implementers or users of this specification can be obtained from the IETF on-line IPR repository at http://www.ietf.org/ipr.
The IETF invites any interested party to bring to its attention any copyrights, patents or patent applications, or other proprietary rights that may cover technology that may be required to implement this standard. Please address the information to the IETF at ietf- ipr@ietf.org.
Acknowledgement
Funding for the RFC Editor function is currently provided by the Internet Society.