Harmonizing how applications specify DANE-like usage
draft-ogud-dane-vocabulary-01

Abstract

There is no standard terminology as how to talk about use of DNS in various application contexts, this document goal is to facilitate creation of such a vocabulary/taxonomy.

This document started out as proposal for specific word usage for specifications of adding DANE like technology by different protocols/services. DANE is a method for specifying in DNS records acceptable keys/certificates for application servers.

The terms defined in this document should be applicable to all uses of service specification that uses DNS records.

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 http://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 April 24, 2014.

Copyright Notice

Copyright (c) 2013 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 (http://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
• 1.1. Requirements notation
• 2. Proposed Terms
• 2.1. DNS Navigation Records
• 2.2. DNS Integrity
• 2.3. Service Specification Records
• 2.4. Service Address Records
• 2.5. Application Authentication Records
• 2.6. Offered Names
• 3. DANE Example specifcation
• 4. IANA considerations
• 5. Security considerations
• 6. Internationalizaiton Considerations
• 7. References
• 7.1. Normative References
• 7.2. Informative References
• Appendix A. Document history
• Author's Address

1. Introduction

DNS [RFC1034] is being used by many protocols to express where services are located on the internet, today there is no good way to express exactly what people have in mind when specifying a new service/protocol exactly and in concise manner how the service is looked up in the DNS.

DANE [RFC6698] is a powerful new way to provide/amend how authentication/authorization/confidencialty of a connection to a server can be protected by leveraging DNSSEC [RFC4033] [RFC4034] [RFC4035] for the establishment of TLS connection [RFC5246] [RFC6347] which in many cases uses PKIX [RFC5280]. All of these technologies are complicated. People familiar with one or two are not necessarily familiar with all the parts that needed to apply DANE like mechanism to other protocols.

The goal of this document is three fold:

• To provide common vocabulary for usage of DNS records in service specification.
• To provide an overview of the non protocol specific parts needed to specify an DANE like addition.
• To provide a common framework for such specifications making it easy to review/compare the specifications. An important goal is to allow the new specifications to avoid repeating explanations and/or definitions.

When below the notation "foo/bar" is used that is because the editor is not sure if both apply or which one is more appropriate, please advise.

1.1. Requirements notation

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].

2. Proposed Terms

The terms below are being proposed to reduce confusion when reading protocol specifications, related to DNS and DANE, for various application protocols.

At this point all the terms below are proposals and better terms are welcome.

2.1. DNS Navigation Records

DNS Navigation refers to any records used to traverse the DNS tree to find the records requested. This includes

NS records:

that provide a referral to DNS servers for more specific part of the name being looked up. Example: name server for "example." will hand out a referral to server for "bar.example." when asked about "foo.bar.example."

CNAME records:

records that change the location of an record, this for all practical purposes a pointer that only applies to that specific name.

DNAME records:

specify a rewrite rule for a name to a new name. Example: "bar.example." DNAME "foo.example." means that "www.bar.example." is to be looked up as "www.foo.example". DNAME applies to names that are longer than the name it, i.e. "bar.example." is not rewritten but "www.bar.example." is

DANE specification explicitly requires all of these records to be validated by DNSSEC.

See section Section 2.2

While traversing the DNS tree other records like A and AAAA are used but these records do not change the "navigation", these records do not explicitly need to be protected as the data retrieved from the addresses is expected to be protected.

2.2. DNS Integrity

DNSSEC defines a records and procedures to provide integrity and authentication to data stored in DNS [RFC4033]. The records used to provide the keying information and chain of trust are DNSKEY, DS records. NSEC/NSEC3 provide information about existence/non- existence of the requested information. RRSIG provides a digital signature for a RRset.

DNSSEC provides both Integrity and Authenticity i.e. it says the records came from the right source and have not been changed.

Any DNS record(s) that is DNS Integrity protected, will pass DNSSEC validation for all DNS Navigation records leading to the name and the record(s) also pass DNSSEC validation.

In the case of CNAME and DNAME that go "sideways" i.e. to a different branch of the DNS tree, both branches MUST be validated, this principle needs to be applied to all such redirection.

2.3. Service Specification Records

Protocols have different ways to provide information about where servers are located. Web servers are frequently specified by name i.e. the "www" prefix. Email servers have a special RR type (MX), Jabber uses SR records, ENUM uses NAPTR records etc. and there are also protocols that use a combination like S-NAPTR a schema where NAPTR records are used to specify where to look for SRV records. For all practical purposes NAPTR + SRV should combined be treated as the Service Specification.

For a DANE like specification it has to be clear as what the service specification records are and that these records use requires DNS Integrity.

NOTE: when a client supplies a string to the server as a indicator of what service the the client wants, the string supplied MAY depend on redirection in DNS navigation as well as results of NAPTR records, etc. See section Section 2.6

2.4. Service Address Records

This is where the address records used by the servers reside, specifications SHOULD NOT make a difference between what kind of address records are used.

In some cases the Service Specification records reside at the same name or are the same as the Service Address records. Example: original TLS/DANE, thus both kinds of records are covered by the same DNS integrity rules.

2.5. Application Authentication Records

This term refers to the records that provide information about what are acceptable keys/certificates for the servers to offer.

Application Authentication Records MUST be protected by DNS Integrity and each protocol specification MUST explicitly state where/how to look up the Authentication records.

In some cases all the servers for a service will have the same authentication information, in other cases it is on server by server case. In the first case it is "natural" to store the Authentication records "at" the Service Specification records. In the second case it more natural to store them "at" the Address Records. In this context "at" means the authentication records are stored at name that is an extension of the location example: "_443._tcp.www.example.com" for [RFC6698]. It is possible that neither of these locations is the right one and in that case the specification MUST explicitly express a rule to find the Authentication Records.

Note: above that there is no a requirement that the Application Address records be covered by DNS Integrity. This is because when the Application Authentication records reside "at" the address records, DNS Integrity is inherited. On the other hand when when Application Authentication Records are stored "at" the Service Specification Record, DNS Integrity for the address records is optional, as any connection to a bogus/wrong server should fail the Authentication tests performed at connection time.

Note: When a Address record search starts with CNAME or DNAME, where should the Authentication Records reside ? With redirection record or with final address record ?

2.6. Offered Names

When DNAME is used the name queried for is rewritten into a new name. To disambiguate these cases following prefix terms are defined. Similar rules apply NAPTR + SRV combinations. It is important for many applications to be able to express what name is presented to the service at connection time.

Query:

The name the application issues the query for that RRset.

Final:

The name after all the indirection records have been applied.

SRV

The name on the SRV record used.

NAPTR

The name on the first NAPTR record used, prefix with Final if that is the one wanted.

Intermediate

A particular location in the indirection chain. The specification needs to handle this case if it ever occurs. NOTE: not sure this is needed???

3. DANE Example specifcation

This section is an short example for a protocol that is like SSH [RFC4253] we will call this protocol HISS. This is not an actual full specification, just here to give an idea of how to go about extending DANE-like to a random protocol using the terminology from this document.

Location of HISS protocol DNS records:

Service Specification Records:

HISS uses the address records as the service specification record. This record MUST have "DNS Integrity" as explained in RFC-to-be-this-document.

Service Address Records:

see: Service Specification Records.

Application Authentication Records:

The protocol uses the DNS HISSFP that is stored at the same name as the service is specified. The HISSFP record, if present, takes precedence over keys stored in client cache.

The HISS protocol and HISSFP DNS RR do not exist

4. IANA considerations

None

[RFC Editor: Please remove this section before publication ]

5. Security considerations

TBD

6. Internationalizaiton Considerations

Does this document, in a later version needs to address how to deal with special cases resulting from IDN DNAME use?

When selecting terms to use in standards documents it is important to select works that do not confuse international readers. This document goes out of its way in selecting English terms that are dissimilar to avoid confusions.

7. References

7.1. Normative References

7.2. Informative References

Appendix A. Document history

[RFC Editor: Please remove this section before publication ]

00 Initial version

01 Updated version: added the section on Offered names, expanded DNAME/CNAME text, added NAPTR and SRV considerations.

Author's Address