DS Algorithms for Securing NS and Glue

Status of This Memo

This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.¶

Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.¶

Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."¶

This Internet-Draft will expire on 23 March 2022.¶

Copyright Notice

Copyright (c) 2021 IETF Trust and the persons identified as the document authors. All rights reserved.¶

This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.¶

Table of Contents

1. Introduction

Currently, any query for delegation NS records over an unprotected transport path returns results which do not have protection from tampering by an active on-path attacker, or against successful cache poisoning attackes. This is because the parent NS records are being authoritative, and thus do not have RRSIGs. The child NS records with the same owner name are authoritave, but the parent NS records are what get used for delegations.¶

There is new privacy work that relies on the name server names in the delgation RDATA. Unsigned records are vulnerable to modification by on-path attackers and to cache poisoning by off-path attackers. That privacy work uses the name for TLS validation, and the only source of the name server name is the NS record in the delgation.¶

This document is about protecting the RDATA of NS record, not the privacy issues per se.¶

Note that the use of an encrypted trasport (such as DoT [RFC7858] to the parent would be an alternative approach, but in the absence of encrypted transport, the current approach is recommended.¶

If an attacker alters the NS records returned, or poisons the resolver's cache for the unsigned delegation NS, the recursive resolver could be directed to a server operated by an attacker.¶

2. Conventions and Definitions

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here.¶

3. Background

The methods developed for adding security to the Domain Name System, collectively refered to as DNSSEC, had as a primary requirement that they be backward compatible. The original specifications for DNS used the same Resourc Record Type (RRTYPE) on both the parent and child side of a zone cut (the NS record). The main goal of DNSSEC was to ensure data integrity by using cryptographic signatures. However, owing to this overlap in the NS record type where the records above and below the zone cut have the same owner name created an inherent conflict, as only the child zone is authoritative for these records.¶

The result is that the parent side of the zone cut has records needed for DNS resolution which are not signed and not validatable.¶

This has no security (data validation) impact on DNS zones which are fully DNSSEC signed (anchored at the IANA DNS Trust Anchor), but does impact unsigned zones regardless of where the transition from secure to insecure occurs.¶

4. New DNSKEY Algorithm

This new DNSKEY algorithm conforms to the structure requirements from [RFC4034], but is not itself used as actual DNSKEY algorithm. It is assigned a value from the DNSKEY algorithm table. No DNSKEY records are published in the child zone using this algorithm.¶

This DNSKEY is used only as the input to the corresponding DS hashs published in the parent zone.¶

Note that this method is orthogonal to the specific choice of DS hashes. Examples here refer to the what is published currently in the IANA tables for recommended DNSSEC parameters, including recommended choices. Any valid supported hash for DS records MAY be used.¶

example.com NS ns1.Example.Net
example.com NS ns2.Example.Net

example.com NS ns1.Example.Net
example.com NS ns2.Example.Net
; example.com DS KeyTag=FOO Algorithm={TBD1}
;   DigestType=2 Digest=sha2-256(wireformat("ns1.example.net"))
example.com DS KeyTag=FOO Algorithm={TBD1} DigestType=2 Digest=...
; example.com DS KeyTag=FOO Algorithm={TBD1}
;   DigestType=2 Digest=sha2-256(wireformat("ns2.example.net"))
example.com DS KeyTag=FOO Algorithm={TBD1} DigestType=2 Digest=...

5. Validation Using These DS Records

These new DS records are used to validate corresponding delegation records and glue. Each NS record must have a matching DS record. The expected DS record RDATA is constructed, and a matching DS record with identical RDATA MUST be present. Any NS record without matching valid DS record MUST be ignored.¶

• NS records are validated using {TBD1}. The RDATA consists of only the RDATA from the NS record.¶

6. Protection of glue records

For the issue of glue records (parent side A/AAAA records which are not signed), please see the proposal [I-D.dickson-dnsop-glueless].¶

7. Security Considerations

As outlined earlier in FIXME, there could be security issues in various use cases.¶

The target domain containing each name server name is presumed (and required) to be DNSSEC signed.¶

8. IANA Considerations

This document has no IANA actions. (FIXME - update this doc to specify the required IANA actions - add TBD1 to the DNSKEY algorithm table)¶

Appendix A. Acknowledgments

Thanks to everyone who helped create the tools that let everyone use Markdown to create Internet Drafts, and the RFC Editor for xml2rfc.¶

Thanks to Dan York for his Tutorial on using Markdown (specifically mmark) for writing IETF drafts.¶

Thanks to YOUR NAME HERE for contributions, reviews, etc.¶

Author's Address