DNS Privacy Considerations
Elkins
26241
United States of America
WV
tjw.ietf@gmail.com
Internet Area
dprive
DNS
This document describes the privacy issues associated with the use of the DNS
by Internet users. It provides general observations about typical current
privacy practices. It is intended to be an analysis of the present situation
and does not prescribe solutions. This document obsoletes RFC 7626.
Introduction
This document is an analysis of the DNS privacy issues, in the spirit
of .
The Domain Name System (DNS) is specified in , , and
many later RFCs, which have never been consolidated. It is one of the most
important infrastructure components of the Internet and is often ignored or
misunderstood by Internet users (and even by many professionals). Almost
every activity on the Internet starts with a DNS query (and often several).
Its use has many privacy implications, and this document is an attempt at a
comprehensive and accurate list.
Let us begin with a simplified reminder of how the DNS works (see also
). A client, the stub resolver, issues a
DNS query to a server called the recursive resolver (also called caching
resolver, full resolver, or recursive name server). Let's use the query
"What are the AAAA records for www.example.com?" as an example. AAAA is the
QTYPE (Query Type), and www.example.com is the QNAME (Query Name). (The
description that follows assumes a cold cache, for instance, because the
server just started.) The recursive resolver will first query the root name
servers. In most cases, the root name servers will send a referral. In this
example, the referral will be to the .com name servers. The resolver repeats
the query to one of the .com name servers. The .com name servers, in turn,
will refer to the example.com name servers. The example.com name servers will
then return the answers. The root name servers, the name servers of .com, and
the name servers of example.com are called authoritative name servers. It is
important, when analyzing the privacy issues, to remember that the question
asked to all these name servers is always the original question, not a
derived question. The question sent to the root name servers is "What are
the AAAA records for www.example.com?", not "What are the name servers of
.com?". By repeating the full question, instead of just the relevant part of
the question to the next in line, the DNS provides more information than
necessary to the name server. In this simplified description, recursive
resolvers do not implement QNAME minimization as described in ,
which will only send the relevant part of the question to the upstream name
server.
DNS relies heavily on caching, so the algorithm described
above is actually a bit more complicated, and not all questions are
sent to the authoritative name servers. If the
stub resolver asks the recursive resolver a few seconds later, "What are the SRV records
of _xmpp-server._tcp.example.com?", the recursive resolver will
remember that it knows the name servers of example.com and will just
query them, bypassing the root and .com. Because there is typically
no caching in the stub resolver, the recursive resolver, unlike the
authoritative servers, sees all the DNS traffic. (Applications, like
web browsers, may have some form of caching that does not follow DNS
rules, for instance, because it may ignore the TTL. So, the
recursive resolver does not see all the name resolution activity.)
It should be noted that DNS recursive resolvers sometimes forward
requests to other recursive resolvers, typically bigger machines,
with a larger and more shared cache (and the query hierarchy can be
even deeper, with more than two levels of recursive resolvers). From
the point of view of privacy, these forwarders are like resolvers
except that they do not see all of the requests being made (due to
caching in the first resolver).
At the time of writing, almost all this DNS traffic is currently
sent unencrypted. However, there is increasing deployment
of DNS over TLS (DoT) and DNS over HTTPS (DoH)
, particularly in mobile devices, browsers, and by
providers of anycast recursive DNS resolution services. There are a
few cases where there is some alternative channel encryption, for
instance, in an IPsec VPN tunnel, at least between the stub resolver and
the resolver. Some recent analysis on the service quality of encrypted DNS
traffic can be found in .
Today, almost all DNS queries are sent over UDP . This has
practical consequences when considering encryption of the traffic as a
possible privacy technique. Some encryption solutions are only designed for
TCP, not UDP, although new solutions are still emerging
.
Another important point to keep in mind when analyzing the privacy
issues of DNS is the fact that DNS requests received by a server are
triggered for different reasons. Let's assume an eavesdropper wants
to know which web page is viewed by a user. For a typical web page,
there are three sorts of DNS requests being issued:
- Primary request:
- This is the domain name in the URL that the user
typed, selected from a bookmark, or chose by clicking on a
hyperlink. Presumably, this is what is of interest for the
eavesdropper.
- Secondary requests:
- These are the additional requests performed by
the user agent (here, the web browser) without any direct
involvement or knowledge of the user. For the Web, they are
triggered by embedded content, Cascading Style Sheets (CSS),
JavaScript code, embedded images, etc. In some cases, there can
be dozens of domain names in different contexts on a single web
page.
- Tertiary requests:
- These are the additional requests performed by
the DNS service itself. For instance, if the answer to a query is
a referral to a set of name servers and the glue records are not
returned, the resolver will have to send additional requests to turn
the name servers' names into IP addresses. Similarly, even if
glue records are returned, a careful recursive server will send
tertiary requests to verify the IP addresses of those records.
It can also be noted that, in the case of a typical web browser, more
DNS requests than strictly necessary are sent, for instance, to
prefetch resources that the user may query later or when
autocompleting the URL in the address bar. Both are a significant privacy
concern since they may leak information even about non-explicit
actions. For instance, just reading a local HTML page, even without
selecting the hyperlinks, may trigger DNS requests.
Privacy-related terminology is from
. This document obsoletes .
Scope
This document focuses mostly on the study of privacy risks for the
end user (the one performing DNS requests). The risks of
pervasive surveillance are considered as well as risks coming from a more
focused surveillance. In this document, the term "end user" is used
as defined in .
This document does not attempt a comparison of specific privacy protections
provided by individual networks or organizations; it makes only general
observations about typical current practices.
Privacy risks for the holder of a zone (the risk that someone gets the data)
are discussed in and .
Privacy risks for recursive operators (including access providers and
operators in enterprise networks) such as leakage of private namespaces or
blocklists are out of scope for this document.
Non-privacy risks (e.g., security-related considerations such as cache poisoning) are
also out of scope.
The privacy risks associated with the use of other protocols that make use of
DNS information are not considered here.
Risks
The following four sections outline the privacy considerations associated with
different aspects of the DNS for the end user. When reading these sections, it
needs to be kept in mind that many of the considerations (for example, recursive
resolver and transport protocol) can be specific to the network context that a
device is using at a given point in time. A user may have many devices, and each
device might utilize many different networks (e.g., home, work, public, or
cellular) over a period of time or even concurrently. An exhaustive analysis of
the privacy considerations for an individual user would need to take into
account the set of devices used and the multiple dynamic contexts of each
device. This document does not attempt such a complex analysis; instead, it
presents an overview of the various considerations that could form the basis of
such an analysis.
Risks in the DNS Data
The Public Nature of DNS Data
It has been stated that "the data in the DNS is public". This sentence
makes sense for an Internet-wide lookup system, and there
are multiple facets to the data and metadata involved that deserve a
more detailed look. First, access control lists (ACLs) and private
namespaces notwithstanding, the DNS operates under the assumption
that public-facing authoritative name servers will respond to "usual"
DNS queries for any zone they are authoritative for, without further
authentication or authorization of the client (resolver). Due to the
lack of search capabilities, only a given QNAME will reveal the
resource records associated with that name (or that name's nonexistence). In other words: one needs to know what to ask for in
order to receive a response. There are many ways in which supposedly "private"
resources currently leak. A few examples are DNSSEC NSEC zone walking ,
passive DNS services , etc. The zone transfer QTYPE is
often blocked or restricted to authenticated/authorized access to
enforce this difference (and maybe for other reasons).
Another difference between the DNS data and a particular DNS
transaction (i.e., a DNS name lookup): DNS data and the results of a
DNS query are public, within the boundaries described above, and may
not have any confidentiality requirements. However, the same is not
true of a single transaction or a sequence of transactions; those
transactions are not / should not be public. A single transaction
reveals both the originator of the query and the query contents; this
potentially leaks sensitive information about a specific user. A
typical example from outside the DNS world is that the website of Alcoholics Anonymous is public but the fact that you visit it should not be. Furthermore,
the ability to link queries reveals information about individual use
patterns.
Data in the DNS Request
The DNS request includes many fields, but two of them seem particularly
relevant for the privacy issues: the QNAME and the source IP address.
"Source IP address" is used in a loose sense of "source IP address + maybe
source
port number", because the port number is also in the request and can be used to
differentiate between several users sharing an IP address (behind a
Carrier-Grade NAT (CGN), for instance ).
The QNAME is the full name sent by the user. It gives information
about what the user does ("What are the MX records of example.net?"
means they probably want to send email to someone at example.net,
which may be a domain used by only a few persons and is therefore
very revealing about communication relationships). Some QNAMEs are
more sensitive than others. For instance, querying the A record of a
well-known web statistics domain reveals very little (everybody
visits websites that use this analytics service), but querying the A
record of www.verybad.example where verybad.example is the domain of
an organization that some people find offensive or objectionable may
create more problems for the user. Also, sometimes, the QNAME embeds
the software one uses, which could be a privacy issue (for instance,
_ldap._tcp.Default-First-Site-Name._sites.gc._msdcs.example.org.
There are also some BitTorrent clients that query an SRV record for
_bittorrent-tracker._tcp.domain.example.
Another important thing about the privacy of the QNAME is future
usages. Today, the lack of privacy is an obstacle to putting
potentially sensitive or personally identifiable data in the DNS. At
the moment, your DNS traffic might reveal that you are exchanging emails but not with whom. If your Mail User Agent (MUA) starts looking up
Pretty Good Privacy (PGP) keys in the DNS , then
privacy becomes a lot more important. And email is just an example;
there would be other really interesting uses for a more privacy-friendly DNS.
For the communication between the stub resolver and the recursive resolver,
the source IP address is the address of the user's machine. Therefore, all
the issues and warnings about collection of IP addresses apply here. For the communication between the recursive resolver and the authoritative name
servers, the source IP address has a different meaning; it does not have the
same status as the source address in an HTTP connection. It is typically the
IP address of the recursive resolver that, in a way, "hides" the real user.
However, hiding does not always work. The edns-client-subnet (ECS) EDNS0 option is sometimes used (see one privacy analysis in ).
Sometimes the end user has a personal recursive resolver on their machine.
In both cases, the IP address originating queries to the authoritative server
is as sensitive as it is for HTTP .
A note about IP addresses: there is currently no IETF document that describes
in detail all the privacy issues around IP addressing in general, although
does discuss privacy considerations for IPv6 address generation
mechanisms. In the meantime, the discussion here is intended to include both
IPv4 and IPv6 source addresses. For a number of reasons, their assignment and
utilization characteristics are different, which may have implications for
details of information leakage associated with the collection of source
addresses. (For example, a specific IPv6 source address seen on the public
Internet is less likely than an IPv4 address to originate behind an address-sharing scheme.) However, for both IPv4 and IPv6 addresses, it is
important to note that source addresses are propagated with queries
via the ECS option and comprise metadata about the host, user,
or application that originated them.
Data in the DNS Payload
At the time of writing, there are no standardized client identifiers contained in
the DNS payload itself (ECS, as described in , is widely used; however, is only an Informational RFC).
DNS Cookies are a lightweight DNS transaction security mechanism that
provides limited protection against a variety of increasingly common
denial-of-service and amplification/forgery or cache poisoning attacks by
off-path attackers. It is noted, however, that they are designed to just verify
IP addresses (and should change once a client's IP address changes), but they are
not designed to actively track users (like HTTP cookies).
There are anecdotal accounts of Media Access Control (MAC) addresses
and even user names being inserted in nonstandard EDNS(0) options
for stub-to-resolver communications to support proprietary functionality
implemented at the resolver (e.g., parental filtering).
Cache Snooping
The content of recursive resolvers' caches can reveal data about the
clients using it (the privacy risks depend on the number of clients).
This information can sometimes be examined by sending DNS queries
with RD=0 to inspect cache content, particularly looking at the DNS
TTLs . Since this also is a reconnaissance
technique for subsequent cache poisoning attacks, some countermeasures have already been developed and deployed .
Risks on the Wire
Unencrypted Transports
For unencrypted transports, DNS traffic can be seen by an eavesdropper like
any other traffic. (DNSSEC, specified in , explicitly excludes
confidentiality from its goals.) So, if an initiator starts an HTTPS
communication with a recipient, the HTTP traffic will be encrypted, but the
DNS exchange prior to it will not be. When other protocols become more
and more privacy aware and secured against surveillance (e.g., ,
), the use of unencrypted transports for DNS may
become "the weakest link" in privacy. It is noted that, at the time of writing,
there is ongoing work attempting to encrypt the Server Name Identification (SNI) in the TLS handshake
, which is one of the
last remaining non-DNS cleartext identifiers of a connection target.
An important specificity of the DNS traffic is that it may take a
different path than the communication between the initiator and the
recipient. For instance, an eavesdropper may be unable to tap the
wire between the initiator and the recipient but may have access to
the wire going to the recursive resolver or to the authoritative
name servers.
The best place to tap, from an eavesdropper's point of view, is
clearly between the stub resolvers and the recursive resolvers,
because traffic is not limited by DNS caching.
The attack surface between the stub resolver and the rest of the
world can vary widely depending upon how the end user's device is
configured. By order of increasing attack surface:
- The recursive resolver can be on the end user's device. In (currently) a small number of cases, individuals may choose to
operate their own DNS resolver on their local machine. In this
case, the attack surface for the connection between the stub
resolver and the caching resolver is limited to that single
machine. The recursive resolver will expose data to authoritative
resolvers as discussed in .
- The recursive resolver may be at the local network edge. For
many/most enterprise networks and for some residential networks, the
caching resolver may exist on a server at the edge of the local
network. In this case, the attack surface is the local network.
Note that in large enterprise networks, the DNS resolver may not
be located at the edge of the local network but rather at the edge
of the overall enterprise network. In this case, the enterprise
network could be thought of as similar to the Internet Access
Provider (IAP) network referenced below.
- The recursive resolver can be in the IAP network. For most residential
networks and potentially other networks, the typical case is for the
user's device to be configured (typically automatically through DHCP or
relay agent options) with the addresses of the DNS proxy in the Customer
Premises Equipment (CPE), which in turn
points to the DNS recursive resolvers at the IAP. The attack surface for
on-the-wire attacks is therefore from the end user system across the
local network and across the IAP network to the IAP's recursive resolvers.
- The recursive resolver can be a public DNS service (or a privately run DNS
resolver hosted on the public Internet). Some machines
may be configured to use public DNS resolvers such as those
operated by Google Public DNS or OpenDNS. The user may
have configured their machine to use these DNS recursive resolvers
themselves -- or their IAP may have chosen to use the public DNS
resolvers rather than operating their own resolvers. In this
case, the attack surface is the entire public Internet between the
user's connection and the public DNS service. It can be noted that if the
user selects a single resolver with a small client population (even when using
an encrypted transport), it can actually serve to aid tracking of that user as
they move across network environments.
It is also noted that, typically, a device connected only to a modern cellular
network is
- directly configured with only the recursive resolvers of the IAP and
-
afforded some level of protection against some types of eavesdropping
for all traffic (including DNS traffic) due to the cellular network
link-layer encryption.
The attack surface for this specific scenario is not considered here.
Encrypted Transports
The use of encrypted transports directly mitigates passive surveillance of the
DNS payload; however, some privacy attacks are still possible. This section
enumerates the residual privacy risks to an end user when an attacker can
passively monitor encrypted DNS traffic flows on the wire.
These are cases where user identification, fingerprinting, or correlations may be
possible due to the use of certain transport layers or cleartext/observable
features. These issues are not specific to DNS, but DNS traffic is susceptible
to these attacks when using specific transports.
Some general examples exist; for example, certain studies highlight
that the OS fingerprint values of IPv4 TTL, IPv6 Hop Limit, or TCP Window size can be used to fingerprint client OSes or that various techniques can be
used to de-NAT DNS queries .
Note that even when using encrypted transports, the use of cleartext transport
options to decrease latency can provide correlation of a user's connections,
e.g., using TCP Fast Open .
Implementations that support encrypted transports also commonly reuse
connections for multiple DNS queries to optimize performance (e.g., via DNS
pipelining or HTTPS multiplexing). Default configuration options for encrypted
transports could, in principle, fingerprint a specific client application.
For
example:
- TLS version or cipher suite selection
- session resumption
- the maximum number of messages to send and
- a maximum connection time before closing a connections and reopening.
If libraries or applications offer user configuration of such options (e.g.,
), then they could, in principle, help to identify a specific user. Users
may want to use only the defaults to avoid this issue.
While there are known attacks on older versions of TLS, the most recent
recommendations and the development of TLS 1.3 largely
mitigate those.
Traffic analysis of unpadded encrypted traffic is also possible
because the sizes and timing of encrypted DNS
requests and responses can be correlated to unencrypted DNS requests upstream
of a recursive resolver.
Risks in the Servers
Using the terminology of , the DNS servers (recursive
resolvers and authoritative servers) are enablers: "they facilitate
communication between an initiator and a recipient without being
directly in the communications path". As a result, they are often
forgotten in risk analysis. But, to quote again, "Although
[...] enablers may not generally be considered as attackers, they may
all pose privacy threats (depending on the context) because they are
able to observe, collect, process, and transfer privacy-relevant
data". In parlance, enablers become observers when they
start collecting data.
Many programs exist to collect and analyze DNS data at the servers -- from
the "query log" of some programs like BIND to tcpdump and more sophisticated
programs like PacketQ and DNSmezzo . The
organization managing the DNS server can use this data itself, or it can be
part of a surveillance program like PRISM and pass data to an
outside observer.
Sometimes this data is kept for a long time and/or distributed to
third parties for research purposes , security
analysis, or surveillance tasks. These uses are sometimes under some
sort of contract, with various limitations, for instance, on
redistribution, given the sensitive nature of the data. Also, there
are observation points in the network that gather DNS data and then
make it accessible to third parties for research or security purposes
("passive DNS" ).
In the Recursive Resolvers
Recursive resolvers see all the traffic since there is typically no
caching before them. To summarize: your recursive resolver knows a
lot about you. The resolver of a large IAP, or a large public
resolver, can collect data from many users.
Resolver Selection
Given all the above considerations, the choice of recursive resolver has
direct privacy considerations for end users. Historically, end user devices
have used the DHCP-provided local network recursive resolver. The choice by a
user to join a particular network (e.g., by physically plugging in a cable or
selecting a network in an OS dialogue) typically updates a number of system
resources -- these can include IP addresses, the availability of IPv4/IPv6, DHCP
server, and DNS resolver. These individual changes, including the change in
DNS resolver, are not normally communicated directly to the user by the OS
when the network is joined. The choice of network has historically determined
the default system DNS resolver selection; the two are directly coupled in
this model.
The vast majority of users do not change their default system DNS settings
and so implicitly accept the network settings for the DNS. The network resolvers
have therefore historically been the sole destination for all of the DNS
queries from a device. These resolvers may have varied
privacy policies depending on the network. Privacy policies for these servers
may or may not be available, and users need to be aware that privacy
guarantees will vary with the network.
All major OSes expose the system DNS settings and allow users to manually
override them if desired.
More recently, some networks and users have actively chosen
to use a large public resolver, e.g., Google Public
DNS,
Cloudflare,
or Quad9. There can be many reasons: cost
considerations for network operators, better reliability, or anti-censorship
considerations are just a few. Such services typically do provide a privacy
policy, and the user can get an idea of the data collected by such
operators by reading one, e.g., Google Public DNS - Your
Privacy.
In general, as with many other protocols, issues around centralization also
arise with DNS.
The picture is fluid with several competing factors
contributing, where these factors can also vary by geographic region. These include:
- ISP outsourcing, including to third-party and public resolvers
- regional market domination by one or only a few ISPs
- applications directing DNS traffic by default to a limited subset of resolvers (see )
An increased proportion of the global DNS resolution traffic being served by
only a few entities means that the privacy considerations for users are
highly dependent on the privacy policies and practices of those
entities. Many of the issues around centralization are discussed in
.
Dynamic Discovery of DoH and Strict DoT
While support for opportunistic DoT can be determined by probing a resolver on
port 853, there is currently no standardized discovery mechanism for DoH and
Strict DoT servers.
This means that clients that might want to dynamically discover such encrypted
services, and where users are willing to trust such services, are not able to do
so. At the time of writing, efforts to provide standardized signaling mechanisms
to discover the services offered by local resolvers are in progress
. Note that an increasing number of ISPs
are deploying encrypted DNS; for example, see the Encrypted DNS Deployment
Initiative .
Application-Specific Resolver Selection
An increasing number of applications are offering application-specific encrypted DNS resolution settings, rather than defaulting to
using only the system resolver. A variety of heuristics and
resolvers are available in different applications, including hard-coded lists of recognized DoH/DoT servers.
Generally, users are not aware of application-specific DNS settings and may
not have control over those settings. To address these limitations, users
will only be aware of and have the ability to control such settings if
applications provide the following functions:
- communicate the change clearly to users when the default application
resolver changes away from the system resolver
- provide configuration options to change the default
application resolver, including a choice to always use the system resolver
- provide mechanisms for users to locally inspect, selectively forward,
and filter queries (either via the application itself or use of the
system resolver)
Application-specific changes to default destinations for users' DNS
queries might increase or decrease user privacy; it is highly
dependent on the network context and the application-specific
default. This is an area of active debate, and the IETF is working on
a number of issues related to application-specific DNS settings.
Active Attacks on Resolver Configuration
The previous section discussed DNS privacy, assuming that all the traffic
was directed to the intended servers (i.e., those that would be used in the
absence of an active attack) and that the potential attacker was purely
passive. But, in reality, there can be active attackers in the network.
The Internet Threat model, as described in , assumes that the attacker
controls the network. Such an attacker can completely control any insecure DNS
resolution, both passively monitoring the queries and responses and substituting
their own responses. Even if encrypted DNS such as DoH or DoT is used, unless
the client has been configured in a secure way with the server identity, an active attacker can impersonate the server. This implies that opportunistic
modes of DoH/DoT as well as modes where the client learns of the DoH/DoT server
via in-network mechanisms such as DHCP are vulnerable to attack. In addition, if
the client is compromised, the attacker can replace the DNS configuration with
one of its own choosing.
Blocking of DNS Resolution Services
User privacy can also be at risk if there is blocking
of access to remote recursive servers
that offer encrypted transports, e.g., when the local resolver does not offer
encryption and/or has very poor privacy policies. For example, active blocking
of port 853 for DoT or blocking of specific IP addresses could restrict the resolvers
available to the user. The extent of the risk to user privacy is highly
dependent on the specific network and user context; a user on a network that
is known to perform surveillance would be compromised if they could not access
such services, whereas a user on a trusted network might have no privacy
motivation to do so.
As a matter of policy, some recursive resolvers use their position in the query
path to selectively block access to certain DNS records. This is a form of
rendezvous-based blocking as described in . Such
blocklists often include servers known to be used for malware, bots, or other
security risks. In order to prevent circumvention of their blocking policies,
some networks also block access to resolvers with incompatible policies.
It is also noted that attacks on remote resolver services, e.g., DDoS, could
force users to switch to other services that do not offer encrypted transports
for DNS.
Encrypted Transports and Recursive Resolvers
DoT and DoH
Use of encrypted transports does not reduce the data available in the recursive
resolver and ironically can actually expose more information about users to
operators. As described in , use of session-based encrypted
transports (TCP/TLS) can expose correlation data about users.
DoH-Specific Considerations
DoH inherits the full privacy properties of the HTTPS stack and as a consequence
introduces new privacy considerations when compared with DNS over UDP, TCP, or
TLS . describes the privacy considerations in
the server of the DoH protocol.
A brief summary of some of the issues includes the following:
- HTTPS presents new considerations for correlation, such as explicit HTTP
cookies and implicit fingerprinting of the unique set and ordering of HTTP
request header fields.
- The User-Agent and Accept-Language request header fields often convey specific
information about the client version or locale.
- Utilizing the full set of HTTP features enables DoH to be more than an HTTP
tunnel, but it is at the cost of opening up implementations to the full set of
privacy considerations of HTTP.
- Implementations are advised to expose the minimal set of data needed to
achieve the desired feature set.
specifically makes selection of HTTPS functionality vs. privacy an
implementation choice. At the extremes, there may be implementations that
attempt to achieve parity with DoT from a privacy perspective at the cost of
using no identifiable HTTP headers, and there might be others that provide feature-rich data flows where the low-level origin of the DNS query is easily
identifiable. Some implementations have, in fact, chosen to restrict the use of the User-Agent header so that resolver operators cannot identify the specific
application that is originating the DNS queries.
Privacy-focused users should be aware of the potential for additional client
identifiers in DoH compared to DoT and may want to only use DoH client
implementations that provide clear guidance on what identifiers they add.
In the Authoritative Name Servers
Unlike what happens for recursive resolvers, the observation capabilities of
authoritative name servers are limited by caching; they see only the requests
for which the answer was not in the cache. For aggregated statistics ("What
is the percentage of LOC queries?"), this is sufficient, but it prevents an
observer from seeing everything. Similarly, the increasing deployment of QNAME
minimization reduces the data visible at the
authoritative name server. Still, the authoritative name servers see a part
of the traffic, and this subset may be sufficient to violate some privacy
expectations.
Also, the user often has some legal/contractual link with the
recursive resolver (they have chosen the IAP, or they have chosen to use a
given public resolver) while having no control and perhaps no
awareness of the role of the authoritative name servers and their
observation abilities.
As noted before, using a local resolver or a resolver close to the
machine decreases the attack surface for an on-the-wire eavesdropper.
But it may decrease privacy against an observer located on an
authoritative name server. This authoritative name server will see
the IP address of the end client instead of the address of a big
recursive resolver shared by many users.
This "protection", when using a large resolver with many clients, is
no longer present if ECS is used because, in this case,
the authoritative name server sees the original IP address (or
prefix, depending on the setup).
As of today, all the instances of one root name server, L-root,
receive together around 50,000 queries per second. While most of it
is "junk" (errors on the Top-Level Domain (TLD) name), it gives an
idea of the amount of big data that pours into name servers. (And
even "junk" can leak information; for instance, if there is a typing
error in the TLD, the user will send data to a TLD that is not the
usual one.)
Many domains, including TLDs, are partially hosted by third-party
servers, sometimes in a different country. The contracts between the
domain manager and these servers may or may not take privacy into
account. Whatever the contract, the third-party hoster may or may not be honest; in any case, it will have to follow its local laws. For
example,
requests to a given ccTLD may go to servers managed by organizations
outside of the ccTLD's country. Users may not anticipate that
when doing a security analysis.
Also, it seems (see the survey described in ) that there is a
strong concentration of authoritative name servers among "popular" domains
(such as the Alexa Top N list). For instance, among the Alexa Top
100K, one DNS provider hosts 10% of
the domains today. The ten most important DNS providers together host one-third of
all domains. With the control (or the ability to sniff the traffic) of a few
name servers, you can gather a lot of information.
Other Risks
Re-identification and Other Inferences
An observer has access not only to the data they directly collect but also
to the results of various inferences about this data. The term "observer" here is used very generally; for example, the observer might
passively observe cleartext DNS traffic or be in the network
that is actively attacking the user by redirecting DNS resolution, or it might be a
local or remote resolver operator.
For instance, a user can be re-identified via DNS queries. If the
adversary knows a user's identity and can watch their DNS queries for
a period, then that same adversary may be able to re-identify the
user solely based on their pattern of DNS queries later on regardless
of the location from which the user makes those queries. For
example, one study found that such re-identification is possible so that "73.1% of all day-to-day links
were correctly established, i.e. user u was either re-identified
unambiguously (1) or the classifier correctly reported that u was not
present on day t + 1 any more (2)". While that study related to web
browsing behavior, equally characteristic patterns may be produced
even in machine-to-machine communications or without a user taking
specific actions, e.g., at reboot time if a characteristic set of
services are accessed by the device.
For instance, one could imagine that an intelligence agency
identifies people going to a site by putting in a very long DNS name
and looking for queries of a specific length. Such traffic analysis
could weaken some privacy solutions.
The IAB Privacy and Security Program also has a document
that considers such inference-based attacks in a more
general framework.
More Information
Useful background information can also be found in (regarding the risk of privacy leaks through DNS) and in a few academic papers:
, , , and
.
Actual "Attacks"
A very quick examination of DNS traffic may lead to the false conclusion that
extracting the needle from the haystack is difficult. "Interesting" primary
DNS requests are mixed with useless (for the eavesdropper) secondary and
tertiary requests (see the terminology in ). But, in
this time of "big data" processing, powerful techniques now exist to get from
the raw data to what the eavesdropper is actually interested in.
Many research papers about malware detection use DNS traffic to
detect "abnormal" behavior that can be traced back to the activity of
malware on infected machines.
Yes, this research was done for the greater good, but technically it is a privacy attack and it demonstrates the
power of the observation of DNS traffic. See ,
, and .
Passive DNS services allow reconstruction of the data of sometimes an entire zone. Well-known passive DNS services keep only the DNS
responses and not the source IP address of the client, precisely for
privacy reasons. Other passive DNS services may not be so careful.
And there are still potential problems with revealing QNAMEs.
The revelations from the Edward Snowden documents, which were leaked from the
National Security Agency (NSA), provide evidence of the use of the DNS in mass
surveillance operations . For example, the MORECOWBELL
surveillance program uses a dedicated covert monitoring infrastructure
to actively query DNS servers and perform HTTP requests to obtain meta-information about services and to check their availability. Also, the
QUANTUMTHEORY
project, which includes detecting lookups for certain addresses and injecting
bogus replies, is another good example showing that the lack of privacy
protections in the DNS is actively exploited.
Legalities
To our knowledge, there are no specific privacy laws for DNS data in any
country. Interpreting general privacy laws, like the European Union's
or GDPR, in the context of DNS traffic data is not an easy task, and
there is no known court precedent. See an interesting analysis in
.
Security Considerations
This document is entirely about security -- more precisely, privacy. It just
lays out the problem; it does not try to set requirements (with the choices
and compromises they imply), much less define solutions. Possible solutions
to the issues described here are discussed in other documents (currently too
many to all be mentioned); see, for instance, "Recommendations for DNS
Privacy Operators" .
IANA Considerations
This document has no IANA actions.
References
Normative References
Informative References
DNS-DNS: DNS-based De-NAT Scheme
Encrypted DNS Deployment Initiative
EDDI
Vie privée: et le DNS alors? [Privacy: what about DNS?]
DNS Cache snooping - should I be concerned?
ISC
Anonymous Resolution of DNS Queries
Lecture Notes in Computer Science, Vol. 5332
Cloud Computing: Centralization and Data Sovereignty
European Journal of Law and Technology, Vol. 3, No. 2
Corrupted DNS Resolution Paths: The Rise of a Malicious
Resolution Authority
ISC/OARC Workshop
Got Malware? Three Signs Revealed In DNS Traffic
Directive 95/46/EC of the European Parliament and of the Council of 24 October 1995 on the protection of individuals with regard to the processing of personal data and on the free movement of such data
European Parliament
Official Journal L 281, pp. 31-50
A Day at the Root of the Internet
ACM SIGCOMM Computer Communication Review, Vol. 38, No. 5
Security and privacy issues of edns-client-subnet
A Day in the Life of the Internet (DITL)
CAIDA
DNS Footprint of Malware
OARC Workshop
An End-to-End, Large-Scale Measurement of DNS-over-Encryption: How Far Have We Come?
IMC '19: Proceedings of the Internet Measurement Conference, pp. 22-35
DNSmezzo
Analysis of Privacy Disclosure in DNS Query
MUE '07: Proceedings of the 2007 International Conference on Multimedia and Ubiquitous Engineering
pp. 952-957
Privacy-Preserving DNS: Analysis of Broadcast, Range Queries and Mix-based Protection Methods
ESORICS 2011, pp. 665-683
getdns
Cache Snooping or Snooping the Cache for Fun and
Profit
Analyzing Characteristic Host Access Patterns for Re-Identification of
Web User Sessions
Lecture Notes in Computer Science, Vol. 7127
NSA's MORECOWBELL: Knell for DNS
A tool that provides a basic SQL-frontend to PCAP-files
DNS-OARC
Release 1.4.3
commit 29a8288
Passive DNS Replication
17th Annual FIRST Conference
Pretty Bad Privacy: Pitfalls of DNS Encryption
WPES '14: Proceedings of the 13th Workshop on Privacy in the Electronic Society, pp. 191-200
PRISM (surveillance program)
Wikipedia
Making the DNS More Private with QNAME Minimisation
A privacy framework for 'DNS big data' applications
An Analysis of TCP Traffic in Root Server DITL Data
DNS-OARC 2014 Fall Workshop
Tor FAQs: I keep seeing these warnings about SOCKS and DNS information leaks. Should I worry?
Tor
Towards Plugging Privacy Leaks in Domain Name System
Updates since RFC 7626
Many references were updated. Discussions of encrypted transports, including
DoT and DoH, and sections on DNS payload, authentication of servers, and blocking of services were added.
With the publishing of
on QNAME minimization, text, references, and initial attempts to
measure deployment were added to reflect this. The text and references on the
Snowden revelations were updated.
The "Risks Overview" section was changed to "Scope" to help clarify the risks
being considered. Text on cellular network DNS, blocking, and
security was added. Considerations for recursive resolvers were collected and placed
together. A discussion on resolver selection was added.
Acknowledgments
Thanks to and to the CENTR members for the original work
that led to this document. Thanks to for the interesting
discussions. Thanks to and for proofreading and
to , , , ,
, and for proofreading, providing technical
remarks, and making many readability improvements. Thanks to ,
, , , , , and
for good written contributions. Thanks to and
for a detailed review of the -bis. And thanks to the IESG
members for the last remarks.
Contributions
and were the original authors of the
document, and their contribution to the initial draft of this document is greatly appreciated.