DNS Resolver Information

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 27 May 2024.¶

Copyright Notice

Copyright (c) 2023 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 Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶

Table of Contents

1. Introduction

Historically, DNS stub resolvers communicated with upstream resolvers without needing to know anything about the features supported by these recursive resolvers. As more and more recursive resolvers expose different features that may impact delivered DNS services, means to help stub resolvers to identify the capabilities of resolvers are valuable. Typically, stub resolvers can discover and authenticate encrypted DNS resolvers provided by a local network, for example, using the Discovery of Network-designated Resolvers (DNR) [RFC9463] and the Discovery of Designated Resolvers (DDR) [RFC9462]. However, these stub resolvers need a mechanism to retrieve information from the discovered recursive resolvers about their capabilities.¶

This document fills that void by specifying a method for stub resolvers to retrieve such information. To that aim, a new resource record (RR) type is defined for stub resolvers to query the recursive resolvers. The information that a resolver might want to expose is defined in Section 5.¶

Retrieved information can be used to feed the server selection procedure. However, that selection procedure is out of the scope of this document.¶

2. Terminology

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.¶

This document makes use of the terms defined in [RFC8499]. The following additional terms are used:¶

Encrypted DNS:

Refers to a DNS scheme where DNS exchanges are transported over an encrypted channel between a DNS client and server (e.g., DNS over HTTPS (DoH) [RFC8484], DNS over TLS (DoT) [RFC7858], or DNS over QUIC (DoQ) [RFC9250]).¶

Encrypted DNS resolver:

Refers to a DNS resolver that supports any encrypted DNS scheme.¶

3. Retrieving Resolver Information

A stub resolver that wants to retrieve the resolver information may use the RR type "RESINFO" defined in this document.¶

The content of the RDATA in a response to a RESINFO RR type query is defined in Section 5. If the resolver understands the RESINFO RR type, the RRSet in the Answer section MUST have exactly one record.¶

A DNS client can retrieve the resolver information using the RESINFO RR type and the QNAME of the domain name that is used to authenticate the DNS resolver (referred to as the Authentication Domain Name (ADN) in [RFC9463]).¶

When clients use the Special-Use Domain Name "resolver.arpa" with DDR to discover a designated encrypted resolver based on an IP address (Section 4 of [RFC9462]), they need to handle RESINFO responses specially. By using the DNS server's domain name from the DDR SVCB response to issue the RESINFO query, a client accepts the risk that a resolver supports DDR but does not support RESINFO. In this scenario, the resolver might pass the query upstream, and then the client can receive a positive RESINFO response either from a legitimate upstream DNS resolver or an attacker.¶

While DNSSEC can be considered as a candidate mechanism to protect against the attack, it is important to note that the name was received over unencrypted DNS and that the RESINFO response can be both validly DNSSEC-signed and not signed by the name that the original DDR resolution intended. To reduce the scope of such an attack, clients wishing to retrieve resolver information from resolvers discovered when performing DDR discovery using resolver IP address (Section 4 of [RFC9462]) MUST ensure during the TLS handshake that the TLS certificate presented by the resolver contains in its SubjectAltName (SAN) the domain name in the TargetName of the DDR SVCB response. If that succeeds, clients MAY choose to retrieve the resolver information using the RESINFO RR type and the QNAME set to the TargetName in the DDR SVCB response.¶

4. Format of the Resolver Information

The resolver information uses the same format as DNS TXT records. The motivation for using the same format as TXT records is to convey a small amount of useful information about a DNS resolver. As a reminder, the format rules for TXT records are defined in the base DNS specification (Section 3.3.14 of [RFC1035]) and further elaborated in the DNS-based Service Discovery (DNS-SD) specification (Section 6.1 of [RFC6763]). The recommendations to limit the TXT record size are discussed in Section 6.1 of [RFC6763].¶

Similar to DNS-SD, the RESINFO RR type uses "key/value" pairs to convey the resolver information. Each "key/value" pair is encoded using the format rules defined in Section 6.3 of [RFC6763]. Using standardized "key/value" syntax within the RESINFO RR type makes it easier for future keys to be defined. If a DNS client sees unknown keys in a RESINFO RR type, it MUST silently ignore them. The same rules for the keys as those defined in Section 6.4 of [RFC6763] MUST be followed for RESINFO.¶

Keys MUST either be defined in the IANA registry (Section 7.2) or begin with the substring "temp-" for names defined for local use only.¶

5. Resolver Information Keys/Values

The following resolver information keys are defined:¶

qnamemin:

If the DNS resolver supports QNAME minimisation [RFC9156] to improve DNS privacy, the key is present. Note that, as per the rules for the keys defined in Section 6.4 of [RFC6763], if there is no '=' in a key, then it is a boolean attribute, simply identified as being present, with no value.¶

This is an optional attribute.¶

exterr:

If the DNS resolver supports extended DNS errors (EDE) option [RFC8914] to return additional information about the cause of DNS errors, the value of this key lists the possible extended DNS error codes that can be returned by this DNS resolver. When multiple values are present, these values MUST be comma-separated.¶

This is an optional attribute.¶

infourl:

An URL that points to the generic unstructured resolver information (e.g., DoH APIs supported, possible HTTP status codes returned by the DoH server, how to report a problem) for troubleshooting purposes.¶

The server MUST support the content-type 'text/html'. The DNS client MUST reject the URL if the scheme is not "https". The URL SHOULD be treated only as diagnostic information for IT staff. It is not intended for end user consumption as the URL can possibily provide misleading information. A DNS client MAY choose to display the URL to the end user, if and only if the encrypted resolver has sufficient reputation, according to some local policy (e.g., user configuration, administrative configuration, or a built-in list of respectable resolvers).¶

This is an optional attribute. For example, a DoT server may not want to host an HTTPS server.¶

New keys can be defined as per the procedure defined in Section 7.2.¶

Figure 1 shows an example of a published resolver information record:¶

resolver.example.net. 7200 IN RESINFO qnamemin exterr=15,16,17
                      infourl=https://resolver.example.com/guide

6. Security Considerations

DNS clients communicating with DNS resolvers discovered using DNR MUST employ one of the following measures to prevent DNS response forgery attacks:¶

1. Establish an authenticated secure connection to the DNS resolver.¶

2. Implement local DNSSEC validation (Section 10 of [RFC8499]) to verify the authenticity of the resolver information.¶

DNS clients communicating with DNS resolvers discovered using DDR's discovery using resolver IP addresses ( Section 4 of [RFC9462]) MUST perform the validation described in Section 3 to limit the effectiveness of upstream attacks (because then the attacker can only redirect the client to another server with a valid TLS certificate for the original IP address but possibly with a different domain name).¶

An encrypted resolver may return incorrect information in RESINFO. If the client cannot validate the attributes received from the resolver, which will be used for resolver selection or display to the end-user, the client should process those attributes only if the encrypted resolver has sufficient reputation according to local policy (e.g., user configuration, administrative configuration, or a built-in list of respectable resolvers). This approach limits the ability of a malicious encrypted resolver to cause harm.¶

7. IANA Considerations

• Note to the RFC Editor: Please update "RFCXXXX" occurences with the RFC number to be assigned to this document.¶

Type: RESINFO
Value: 261
Meaning: Resolver Information as Key/Value Pairs
Reference: RFCXXXX

Acknowledgments

This specification leverages the work that has been documented in [I-D.pp-add-resinfo].¶

Thanks to Tommy Jensen, Vittorio Bertola, Vinny Parla, Chris Box, Ben Schwartz, Tony Finch, Daniel Kahn Gillmor, Eric Rescorla, Shashank Jain, Florian Obser, Richard Baldry, and Martin Thomson for the discussion and comments.¶

Thanks to Mark Andrews, Joe Abley, Paul Wouters, Tim Wicinski, and Steffen Nurpmeso for the discussion on the RR formatting rules.¶

Special thanks to Tommy Jensen for the careful and thoughtful Shepherd review.¶

Thanks to Johan Stenstam for the dns-dir review and Ray Bellis for the RRTYPE allocation review.¶

Authors' Addresses