A Procedure for Cautious Delegation of a DNS Name
draft-kolkman-cautious-delegation-02

Abstract

NOTE: The authors recognize that the statistical models that would inform the process are not well understood and that the possibilities to game the system might be unmountable. Unless we reach more insights on how to deal with this details this work is abandoned.

Sometimes, a DNS name is known to be in use in the wild even though it was never properly delegated. This situation appears particularly, but not only, true in certain domains near the root of the tree: people have independently used those non-existent top-level domains as private namespaces. If those names are to be delegated in the public DNS, prudence dictates that collisions between the private uses and the public use be minimized. We outline a procedure to evaluate the harm of delegation.

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."

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. Terminology
• 2. Background and Introduction
• 2.1. Search-path interaction.
• 3. Predelegation determination of use of a name
• 3.1. Predelegation testing is needed
• 3.2. Determining the names of concern
• 3.2.1. Mode 1: Prior to any delegation
• 3.2.2. Mode 2: After delegation
• 4. Parameters for operation of this procedure
• 4.1. Median or Mean
• 4.2. Discussion of Alternatives
• 5. Security considerations
• 6. IANA Considerations
• 7. Informative References
• Appendix A. Document Editing Details
• A.1. version 00
• A.2. version 01
• Authors' Addresses

1. Terminology

NXDOMAIN:

an alternate expression for the "Name Error" RCODE as described in [[RFC1035] Section 4.1.1]. The two terms are used interchangeably in this document. (definition from [RFC2308])

In this document we will be using the terms zone, domain and sub-domain. When envisioning the domain namespace as a tree, with nodes at the places where the dots separate the labels in a domain name, then:

a 'domain'

is an entire branch. e.g. The .org domain is the branch of the domain name tree for which all names end in .org.

a 'sub-domain'

is a subordinate namespace of a given domain. e.g. all names ending in example.org are in the domain example.org which is a sub-domain of .org

a 'zone'

is a piece of the domain space that is under administrative control of one party. e.g. the .org zone has delegated the example.org domain to the example.org maintainers.

2. Background and Introduction

DNS names have always co-existed with other namespaces that are virtually indistinguishable from the DNS. The DNS was itself deployed alongside the host ### table. NetBIOS ### names, though only one label long, could always interact with the DNS search path mechanism to generate DNS names. Additionally, mDNS [RFC6762] names look just like DNS names. Because different naming systems are usually linked together in the user interface, from an end user's point of view these name spaces are all one -- even though they function differently on the Internet.

While [RFC6761] reserved certain special names for internal or private use, there is evidence [SAC45] that various sites connected to the Internet have used other names for internal purposes. In fact, [RFC6762] advises not to use .local for private use and observes: "the following top-level domains have been used on private internal networks without the problems caused by trying to reuse ".local." for this purpose:"

• .intranet.
• .internal.
• .private.
• .corp.
• .home.
• .lan.

In the event such names are delegated for use in the public DNS, there will be inevitable consequences for such sites. Some of those consequences have implications for security, with the potential for leakage of credentials and HTTP cookies ([RFC6265]). Responsible administration of the public namespace therefore requires great care in permitting public delegation of any name where there is good reason to suppose it is in widespread use as a private namespace, even though such private namespaces are (from the point of view of the DNS) irregular (although not uncommon).

2.1. Search-path interaction.

In many cases a string appears to be used as an "undelegated TLD" (being used as the rightmost label in an name), but this is simply an artifact of domain search list processing.

As an (hypothetical) example, Example Widgets uses a sub-domain (.corp) of their primary domain (example.com) to name their employee workstations, servers, printers and similar. They have an "intranet" server named intranet.corp.example.com. In order to allow their employees to simply type "intranet.corp" to access this server, the users' workstations are configured (probably using [RFC3379]) with the search-list set to "corp.example.com, example.com".

When a user enters "intranet.corp", their workstation will try and resolve the name. RFC1535 [RFC1535] specifies that "in any event where a "." exists in a specified name it should be assumed to be a fully qualified domain name (FQDN) and SHOULD be tried as a rooted name first." and so the users workstation will first try and resolve "intranet.corp.". As there is (currently) no .corp TLD this will result in an NXDOMAIN response. The workstation will then append entries in the search-list until it is able to resolve the (now fully-qualified) name.

If the .corp label were to be delegated as a TLD and the sub-domain "intranet" created within .corp, the first lookup ("intranet.corp") would no longer generate an NXDOMAIN response. This would stop the search-list processing, and direct the user to the incorrect server.

It is worth noting that a researcher analyzing DNS queries hitting the root servers would see queries before search-list processing expands them. While this may not change whether or not it is safe to delegate these names, having an understanding of the cause is valuable.

3. Predelegation determination of use of a name

It is possible for the operator of a zone authoritative for some domain name to tell whether a particular subordinate name has a widespread use outside the DNS. In order to do this, the operator of the zone monitors queries against the zone to learn the names for which there are queries, ignoring those names that actually exist i.e. those names the zone owner delegated or created resource records for (in the remainder of this document we will not make the distinction between entering data with a name or making a delegation; within the context of this document the same considerations apply). The operator then establishes a baseline "noise" level of queries for non-existent subordinate names. Any name that is queried with significantly greater frequency is to be treated as in widespread private use, and it should not be released for delegation. The rest of this section describes the mechanisms for such determination in detail.

3.1. Predelegation testing is needed

In order for this procedure to be useful, it should be undertaken before any subordinate names are delegated. Otherwise, it will be difficult to tell whether a subordinate name is being queried because it is already delegated or because it is in private use.

At the same time, it is possible that the operator of a zone may wish to consider the private use of a descendant name, where some intermediate namespace has been delegated. In that case, it is necessary to ensure that the descendant name is not actually delegated when evaluating queries against that name.

3.2. Determining the names of concern

[ ED NOTE: This methodology needs to tested. First assessment of data indicates that this approach may be far to trivial ]

There are two modes of operation for determining names of concern. The most usual is to examine names for which there is no intermediate delegation. This is useful in case the operator of the zone is deciding whether to permit delegation or addition of a particular name. The second, more unusual mode, is to examine subordinate names inside a sub-domain that has already been delegated. This mode is useful only as part of a regime of contract enforcement with the operator of the (already delegated) sub-domain. [WK Note: Are we sure we even want to address/suggest this second "limited delegation" option? If we are going to discuss it, referring to it as "limited delegation" or similar may help clarify. Personally I think 'tis a silly idea, but... There is talk of doing "test" delegations - basically launch a TLD / domain with a low TTL. If nothing goes "boom" then delegate for longer...]

3.2.1. Mode 1: Prior to any delegation

The procedure starts with the name of a zone, which is called the "starting domain". In order to determine what subordinate names may be problematic, the starting domain zone operator captures all the names it receives in queries. The operator discards as irrelevant any sub-domain it has already delegated in its namespace. Every other queried name will result in a response of Name Error, RCODE=3 ###STD13 ("NXDOMAIN" ###Negative cache). We call the resulting list the "NX names". (See Section 4 for guidance on the sample size.)

The operator then takes the list of NX names, and builds a frequency of queries for each potential delegation point (in practice all immediately subordinate names). The operator proceeds in the fully-qualified domain name ("FQDN") label by label until the next label past the operator's namespace (in practice, these are the names at which delegation will potentially take place). We call these the "target names". The operator counts the number of queries for each target name.

The operator determines the mean and median number of queries over the set of target names. Any name that receives more queries than ###SIGMA -- needs xref to params### greater than the mean, or ###SIGMA2### greater than the median, should be regarded as in widespread private use on the Internet and therefore not a candidate for delegation.

It is possible that only a portion of a namespace subordinate to a target name is actually in private use. It is possible to measure this situation simply by treating the beginning of the namespace in question as the starting domain, and then repeating the procedure above. This could be useful in order to establish baseline restrictions on the operator of a subordinate namespace prior to delegation.

3.2.2. Mode 2: After delegation

This mode is more likely to be useful if the evaluation at the end of the previous section has already been performed. In this case, some sub-domain to the operator's zone is to be evaluated for possible private use, where that sub domain has already been delegated. The zone operator operates the "parent starting zone", and is evaluating use inside a starting domain already operated by someone else. The very same mechanisms as are outlined in Section 3.2.1 are used, but the evaluation must take into consideration the effects of negative TTLs ### for the starting domain. Because of the combining effects of multiple negative TTLs, it is inadvisable to attempt to perform this evaluation beyond the boundary of a single delegation.

4. Parameters for operation of this procedure

This section ought to have some words about sane parameters to use for the procedure.?

4.1. Median or Mean

In this section we would like to describe some likely distributions. Our assumption is that incoming queries will usually follow some dictionary pattern. The 'everybody wants to be Mr. Black' [ResevoirDogs] effect is that queries are much more likely for popular names than for labels filled with random content. Therefore distributions for non-existent names will have relatively little power in the long tail. However, the long tail is significant in the sense that the names in the long tail are most likely not to exist.

The exact type of distribution and the statistical parameters that signify it is subject for a future version of the draft.

4.2. Discussion of Alternatives

The above method is based on looking at names that the querying population perceives to exist. Alternatively one could count queries for a set of random name like "ao42hft3tofj4irsavc4owajhro.example". That type of measurement will set the baseline of _real_ non-existing names and set the noise level (likely zero queries within a reasonable timescale). However, using truly random names introduced the problem that any signal (e.g. a handful of queries used for probing of availability) will make the domain name unavailable.

5. Security considerations

Applying this mechanism as the basis for decisions on whether or not to delegate domains introduces a motivation for gaming the system. The reception of a lot of queries for a particular domain may cause it to not be delegated, while the reception of many random queries (changing the properties of the query distribution) may cause a domain that is in common use to be delegated (by hiding the actual use of names in that domain in the noise). Careful analysis of data (for example, by studying root for queries, and taking into account historical trending) could, in case of suspicion of gaming, help to supplement decisions.

6. IANA Considerations

This document makes no requests of the IANA.

7. Informative References

Appendix A. Document Editing Details

[To Be Removed before publication]

$Id: draft-kolkman-cautious-delegation.xml 3 2013-05-02 14:27:06Z olaf $

A.1. version 00

Documenting the first rough outline based on hallway discussions with the specific purpose to document the idea in the public domain.

$Id: draft-kolkman-cautious-delegation.xml 5 2013-06-11 21:49:28Z warren $

A.2. version 01

• Bunch 'o nits.
• Added section on search-path processing.

Authors' Addresses