DNS Query Name Minimisation to Improve Privacy
draft-ietf-dnsop-rfc7816bis-00

Abstract

This document describes a technique to improve DNS privacy, a technique called "QNAME minimisation", where the DNS resolver no longer sends the full original QNAME to the upstream name server. It obsoletes RFC 7816.

This document is part of the IETF DNSOP (DNS Operations) Working Group. The source of the document, as well as a list of open issues, is at <https://framagit.org/bortzmeyer/rfc7816-bis>

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 March 18, 2019.

Copyright Notice

Copyright (c) 2018 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 and Background
• 2. QNAME Minimisation
• 3. Possible Issues
• 4. Protocol and Compatibility Discussion
• 5. Operational Considerations
• 6. Performance Considerations
• 7. Results of the Experimentation
• 8. Security Considerations
• 9. Implementation status - RFC EDITOR: PLEASE REMOVE BEFORE PUBLICATION
• 10. References
• 10.1. Normative References
• 10.2. Informative References
• Appendix A. An Algorithm to Perform QNAME Minimisation
• Appendix B. Alternatives
• Acknowledgments
• Changes from RFC 7816
• Authors' Addresses

1. Introduction and Background

The problem statement for this document and its predecessor [RFC7816] is described in [I-D.bortzmeyer-dprive-rfc7626-bis]. The terminology ("QNAME", "resolver", etc.) is defined in [I-D.ietf-dnsop-terminology-bis]. This specific solution is not intended to fully solve the DNS privacy problem; instead, it should be viewed as one tool amongst many.

QNAME minimisation follows the principle explained in Section 6.1 of [RFC6973]: the less data you send out, the fewer privacy problems you have.

Before QNAME minimisation, when a resolver received the query "What is the AAAA record for www.example.com?", it sent to the root (assuming a cold resolver, whose cache is empty) the very same question. Sending the full QNAME to the authoritative name server was a tradition, not a protocol requirement. In a conversation with the author in January 2015, Paul Mockapetris explained that this tradition comes from a desire to optimise the number of requests, when the same name server is authoritative for many zones in a given name (something that was more common in the old days, where the same name servers served .com and the root) or when the same name server is both recursive and authoritative (something that is strongly discouraged now). Whatever the merits of this choice at this time, the DNS is quite different now.

2. QNAME Minimisation

The idea is to minimise the amount of data sent from the DNS resolver to the authoritative name server. In the example in the previous section, sending "What are the NS records for .com?" would have been sufficient (since it will be the answer from the root anyway). The rest of this section describes the recommended way to do QNAME minimisation -- the way that maximises privacy benefits (other alternatives are discussed in the appendices).

Instead of sending the full QNAME and the original QTYPE upstream, a resolver that implements QNAME minimisation and does not already have the answer in its cache sends a request to the name server authoritative for the closest known ancestor of the original QNAME. The request is done with:

• the QTYPE NS
• the QNAME that is the original QNAME, stripped to just one label more than the zone for which the server is authoritative

For example, a resolver receives a request to resolve foo.bar.baz.example. Let's assume that it already knows that ns1.nic.example is authoritative for .example and the resolver does not know a more specific authoritative name server. It will send the query QTYPE=NS,QNAME=baz.example to ns1.nic.example.

The minimising resolver works perfectly when it knows the zone cut (zone cuts are described in Section 6 of [RFC2181]). But zone cuts do not necessarily exist at every label boundary. If we take the name www.foo.bar.example, it is possible that there is a zone cut between "foo" and "bar" but not between "bar" and "example". So, assuming that the resolver already knows the name servers of .example, when it receives the query "What is the AAAA record of www.foo.bar.example?", it does not always know where the zone cut will be. To find the zone cut, it will query the .example name servers for the NS records for bar.example. It will get a NODATA response, indicating that there is no zone cut at that point, so it has to query the .example name servers again with one more label, and so on. (Appendix A describes this algorithm in deeper detail.)

www.isc.org, cold cache, aggressive algorithm:

QTYPE   QNAME           TARGET                 NOTE
NS      org             root nameserver
NS      isc.org         Afilias nameserver
NS      www.isc.org     ISC nameserver         "www" may be delegated
A       www.isc.org     ISC nameserver

www.isc.org, cold cache, lazy algorithm (for a cold cache, it is the
same algorithm as now):

QTYPE   QNAME           TARGET                 NOTE
A       www.isc.org     root nameserver
A       www.isc.org     Afilias nameserver
A       www.isc.org     ISC nameserver

www.isc.org, warm cache (all NS RRsets are known), both algorithms:

QTYPE   QNAME           TARGET                 NOTE
A       www.isc.org     ISC nameserver

www.example.org, warm cache (but for isc.org only, example.org's
NS RRset is not known), aggressive algorithm

QTYPE   QNAME               TARGET             NOTE
NS      example.org      Afilias nameserver
NS      www.example.org  Example nameserver
A       www.example.org  Example nameserver

Here are more detailed examples of queries with QNAME minimisation:

Since the information about the zone cuts will be stored in the resolver's cache, the performance cost is probably reasonable. Section 6 discusses this performance discrepancy further.

Note that DNSSEC-validating resolvers already have access to this information, since they have to know the zone cut (the DNSKEY record set is just below; the DS record set is just above).

3. Possible Issues

TODO may be remove the whole section now that it is no longer experimental?

QNAME minimisation is legal, since the original DNS RFCs do not mandate sending the full QNAME. So, in theory, it should work without any problems. However, in practice, some problems may occur (see [Huque-QNAME-Min] for an analysis and [Huque-QNAME-Discuss] for an interesting discussion on this topic).

Some broken name servers do not react properly to QTYPE=NS requests. For instance, some authoritative name servers embedded in load balancers reply properly to A queries but send REFUSED to NS queries. This behaviour is a protocol violation, and there is no need to stop improving the DNS because of such behaviour. However, QNAME minimisation may still work with such domains, since they are only leaf domains (no need to send them NS requests). Such a setup breaks more than just QNAME minimisation. It breaks negative answers, since the servers don't return the correct SOA, and it also breaks anything dependent upon NS and SOA records existing at the top of the zone.

Another way to deal with such incorrect name servers would be to try with QTYPE=A requests (A being chosen because it is the most common and hence a QTYPE that will always be accepted, while a QTYPE NS may ruffle the feathers of some middleboxes). Instead of querying name servers with a query "NS example.com", we could use "A _.example.com" and see if we get a referral. TODO this is what Unbound does

A problem can also appear when a name server does not react properly to ENTs (Empty Non-Terminals). If ent.example.com has no resource records but foobar.ent.example.com does, then ent.example.com is an ENT. Whatever the QTYPE, a query for ent.example.com must return NODATA (NOERROR / ANSWER: 0). However, some name servers incorrectly return NXDOMAIN for ENTs. If a resolver queries only foobar.ent.example.com, everything will be OK, but if it implements QNAME minimisation, it may query ent.example.com and get an NXDOMAIN. See also Section 3 of [DNS-Res-Improve] for the other bad consequences of this bad behaviour.

A possible solution, currently implemented in Knot or Unbound, is to retry with the full query when you receive an NXDOMAIN. It works, but it is not ideal for privacy.

*.example.          60  IN  A   192.0.2.6

Other practices that do not conform to the DNS protocol standards may pose a problem: there is a common DNS trick used by some web hosters that also do DNS hosting that exploits the fact that the DNS protocol (pre-DNSSEC) allows certain serious misconfigurations, such as parent and child zones disagreeing on the location of a zone cut. Basically, they have a single zone with wildcards for each TLD, like:

This lets them have many web-hosting customers without having to configure thousands of individual zones on their name servers. They just tell the prospective customer to point their NS records at the hoster's name servers, and the web hoster doesn't have to provision anything in order to make the customer's domain resolve. NS queries to the hoster will therefore not give the right result, which may endanger QNAME minimisation (it will be a problem for DNSSEC, too).

TODO report by Akamai about why they return erroneous responses https://mailarchive.ietf.org/arch/msg/dnsop/XIX16DCe2ln3ZnZai723v32ZIjE

4. Protocol and Compatibility Discussion

QNAME minimisation is compatible with the current DNS system and therefore can easily be deployed; since it is a unilateral change to the resolver, it does not change the protocol. (Because it is a unilateral change, resolver implementers may do QNAME minimisation in slightly different ways; see the appendices for examples.)

One should note that the behaviour suggested here (minimising the amount of data sent in QNAMEs from the resolver) is NOT forbidden by Section 5.3.3 of [RFC1034] or Section 7.2 of [RFC1035]. As stated in Section 1, the current method, sending the full QNAME, is not mandated by the DNS protocol.

One may notice that many documents that explain the DNS and that are intended for a wide audience incorrectly describe the resolution process as using QNAME minimisation (e.g., by showing a request going to the root, with just the TLD in the query). As a result, these documents may confuse readers that use them for privacy analysis.

5. Operational Considerations

TODO what to do if the resolver forwards? Unbound disables QNAME minimisation in that case, since the forwarder will see everything, anyway. What should a minimising resolver do when forwading the request to a forwarder, not to an authoritative name server? Send the full qname? Minimises? (But how since we do not know the zone cut?)

The administrators of the forwarders, and of the authoritative name servers, will get less data, which will reduce the utility of the statistics they can produce (such as the percentage of the various QTYPEs).

DNS administrators are reminded that the data on DNS requests that they store may have legal consequences, depending on your jurisdiction (check with your local lawyer).

6. Performance Considerations

The main goal of QNAME minimisation is to improve privacy by sending less data. However, it may have other advantages. For instance, if a root name server receives a query from some resolver for A.example followed by B.example followed by C.example, the result will be three NXDOMAINs, since .example does not exist in the root zone. Under query name minimisation, the root name servers would hear only one question (for .example itself) to which they could answer NXDOMAIN, thus opening up a negative caching opportunity in which the full resolver could know a priori that neither B.example nor C.example could exist. Thus, in this common case the total number of upstream queries under QNAME minimisation would be counterintuitively less than the number of queries under the traditional iteration (as described in the DNS standard). TODO mention [RFC8020]? And [RFC8198], the latter depending on DNSSEC?

QNAME minimisation may also improve lookup performance for TLD operators. For a typical TLD, delegation-only, and with delegations just under the TLD, a two‑label QNAME query is optimal for finding the delegation owner name.

QNAME minimisation can decrease performance in some cases -- for instance, for a deep domain name (like www.host.group.department.example.com, where host.group.department.example.com is hosted on example.com's name servers). Let's assume a resolver that knows only the name servers of example.com. Without QNAME minimisation, it would send these example.com name servers a query for www.host.group.department.example.com and immediately get a specific referral or an answer, without the need for more queries to probe for the zone cut. For such a name, a cold resolver with QNAME minimisation will, depending on how QNAME minimisation is implemented, send more queries, one per label. Once the cache is warm, there will be no difference with a traditional resolver. Actual testing is described in [Huque-QNAME-Min]. Such deep domains are especially common under ip6.arpa.

7. Results of the Experimentation

TODO various experiences from actual deployments, problems heard. TODO the Knot bug #339 https://gitlab.labs.nic.cz/knot/knot-resolver/issues/339? TODO Problems with AWS https://forums.aws.amazon.com/thread.jspa?threadID=269116?

8. Security Considerations

QNAME minimisation's benefits are clear in the case where you want to decrease exposure to the authoritative name server. But minimising the amount of data sent also, in part, addresses the case of a wire sniffer as well as the case of privacy invasion by the servers. (Encryption is of course a better defense against wire sniffers, but, unlike QNAME minimisation, it changes the protocol and cannot be deployed unilaterally. Also, the effect of QNAME minimisation on wire sniffers depends on whether the sniffer is on the DNS path.)

QNAME minimisation offers zero protection against the recursive resolver, which still sees the full request coming from the stub resolver.

All the alternatives mentioned in Appendix B decrease privacy in the hope of improving performance. They must not be used if you want maximum privacy.

9. Implementation status - RFC EDITOR: PLEASE REMOVE BEFORE PUBLICATION

This section records the status of known implementations of the protocol defined by this specification at the time of posting of this Internet-Draft, and is based on a proposal described in [RFC7942]. The description of implementations in this section is intended to assist the IETF in its decision processes in progressing drafts to RFCs. Please note that the listing of any individual implementation here does not imply endorsement by the IETF. Furthermore, no effort has been spent to verify the information presented here that was supplied by IETF contributors. This is not intended as, and must not be construed to be, a catalog of available implementations or their features. Readers are advised to note that other implementations may exist.

According to [RFC7942], "this will allow reviewers and working groups to assign due consideration to documents that have the benefit of running code, which may serve as evidence of valuable experimentation and feedback that have made the implemented protocols more mature. It is up to the individual working groups to use this information as they see fit".

Unbound has QNAME minimisation for several years, and it is now the default. It has two modes, strict (no workaround for broken authoritative name servers) and "lax" (retries when there is a NXDOMAIN). TODO Ralph Dolmans talk at OARC https://indico.dns-oarc.net/event/22/contributions/332/attachments/310/542/unbound_qnamemin_oarc24.pdf

Knot resolver also has QNAME minimisation since 2016, and it is activated by default.

BIND has QNAME minimisation since BIND 9.13.2, released in july 2018. Like Unbound, it has several modes, with or without workarounds for broken authoritative name servers.

PowerDNS does not have QNAME minimisation. TODO https://github.com/PowerDNS/pdns/issues/2311

The public DNS resolver at Cloudflare ("1.1.1.1") has QNAME minimisation (it uses Knot).

10. References

10.1. Normative References

10.2. Informative References

Appendix A. An Algorithm to Perform QNAME Minimisation

This algorithm performs name resolution with QNAME minimisation in the presence of zone cuts that are not yet known.

Although a validating resolver already has the logic to find the zone cuts, implementers of other resolvers may want to use this algorithm to locate the cuts. This is just a possible aid for implementers; it is not intended to be normative:

(0)

If the query can be answered from the cache, do so; otherwise, iterate as follows:

(1)

Find the closest enclosing NS RRset in your cache. The owner of this NS RRset will be a suffix of the QNAME -- the longest suffix of any NS RRset in the cache. Call this ANCESTOR.

(2)

Initialise CHILD to the same as ANCESTOR.

(3)

If CHILD is the same as the QNAME, resolve the original query using ANCESTOR's name servers, and finish.

(4)

Otherwise, add a label from the QNAME to the start of CHILD.

(5)

If you have a negative cache entry for the NS RRset at CHILD, go back to step 3.

(6)

Query for CHILD IN NS using ANCESTOR's name servers. The response can be:

(6a)

A referral. Cache the NS RRset from the authority section, and go back to step 1.

(6b)

An authoritative answer. Cache the NS RRset from the answer section, and go back to step 1.

(6c)

An NXDOMAIN answer. Return an NXDOMAIN answer in response to the original query, and stop.

(6d)

A NOERROR/NODATA answer. Cache this negative answer, and go back to step 3.

Appendix B. Alternatives

Remember that QNAME minimisation is unilateral, so a resolver is not forced to implement it exactly as described here.

There are several ways to perform QNAME minimisation. See Section 2 for the suggested way. It can be called the aggressive algorithm, since the resolver only sends NS queries as long as it does not know the zone cuts. This is the safest, from a privacy point of view. Another possible algorithm, not fully studied at this time, could be to "piggyback" on the traditional resolution code. At startup, it sends traditional full QNAMEs and learns the zone cuts from the referrals received, then switches to NS queries asking only for the minimum domain name. This leaks more data but could require fewer changes in the existing resolver codebase.

In the above specification, the original QTYPE is replaced by NS (or may be A, if too many servers react incorrectly to NS requests); this is the best approach to preserve privacy. But this erases information about the relative use of the various QTYPEs, which may be interesting for researchers (for instance, if they try to follow IPv6 deployment by counting the percentage of AAAA vs. A queries). A variant of QNAME minimisation would be to keep the original QTYPE.

Another useful optimisation may be, in the spirit of the HAMMER idea [HAMMER], to probe in advance for the introduction of zone cuts where none previously existed (i.e., confirm their continued absence, or discover them).

To address the "number of queries" issue described in Section 6, a possible solution is to always use the traditional algorithm when the cache is cold and then to move to QNAME minimisation (precisely defining what is "hot" or "cold" is left to the implementer). This will decrease the privacy but will guarantee no degradation of performance.

Acknowledgments

TODO (refer to 7816)

Changes from RFC 7816

• Made changes to deal with Errata #4644
• Changed status to be on standards track

Authors' Addresses