Network Working Group                                      S. Bortzmeyer
Internet-Draft                                                     AFNIC
Intended status: Informational                         November 27, 2013
Expires: May 31, 2014

                     DNS privacy problem statement


   This document describes the privacy issues associated with the use of
   the DNS by Internet users.  It is intended to be mostly a problem
   statement and it does not prescribe solutions (although Section 5
   suggests some possible improvments).

   Discussions of the document should take place on the dnsop mailing
   list [dnsop]

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

   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 May 31, 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
   ( 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

Bortzmeyer                Expires May 31, 2014                  [Page 1]

Internet-Draft                DNS privacy                  November 2013

   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

Table of Contents

   1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   2
   2.  Risks . . . . . . . . . . . . . . . . . . . . . . . . . . . .   3
     2.1.  Data in the DNS request . . . . . . . . . . . . . . . . .   3
     2.2.  On the wire . . . . . . . . . . . . . . . . . . . . . . .   4
     2.3.  In the servers  . . . . . . . . . . . . . . . . . . . . .   5
       2.3.1.  In the resolvers  . . . . . . . . . . . . . . . . . .   6
       2.3.2.  In the authoritative name servers . . . . . . . . . .   6
       2.3.3.  Rogue servers . . . . . . . . . . . . . . . . . . . .   7
   3.  Actual "attacks"  . . . . . . . . . . . . . . . . . . . . . .   7
   4.  Legalities  . . . . . . . . . . . . . . . . . . . . . . . . .   7
   5.  Possible technical solutions  . . . . . . . . . . . . . . . .   7
     5.1.  On the wire . . . . . . . . . . . . . . . . . . . . . . .   8
       5.1.1.  Reducing the attack surface . . . . . . . . . . . . .   8
       5.1.2.  Encrypting the DNS traffic  . . . . . . . . . . . . .   8
     5.2.  In the servers  . . . . . . . . . . . . . . . . . . . . .  10
       5.2.1.  In the resolvers  . . . . . . . . . . . . . . . . . .  10
       5.2.2.  In the authoritative name servers . . . . . . . . . .  10
       5.2.3.  Rogue servers . . . . . . . . . . . . . . . . . . . .  11
   6.  Security considerations . . . . . . . . . . . . . . . . . . .  11
   7.  Acknowledgments . . . . . . . . . . . . . . . . . . . . . . .  11
   8.  References  . . . . . . . . . . . . . . . . . . . . . . . . .  11
     8.1.  Normative References  . . . . . . . . . . . . . . . . . .  12
     8.2.  Informative References  . . . . . . . . . . . . . . . . .  12
   Author's Address  . . . . . . . . . . . . . . . . . . . . . . . .  13

1.  Introduction

   The Domain Name System is specified in [RFC1034] and [RFC1035].  It
   is one of the most important infrastructure components of the
   Internet and one of the most often ignored or misunderstood.  Almost
   every activity on the Internet starts with a DNS query (and often
   several).  Its use has many privacy implications and we try to give
   here a comprehensive and accurate list.

   Let us start with a small reminder of the way the DNS works (with
   some simplifications).  A client, the stub resolver, issues a DNS
   query to a server, the resolver (also called caching resolver or full
   resolver or recursive name server).  For instance, the query is "What
   are the AAAA records for".  AAAA is the qtype (Query
   Type) and the qname (Query Name).  To get the answer,
   the resolver will query first the root nameservers, which will, most
   of the times, send a referral.  Here, the referral will be to .com
   nameservers.  In turn, they will send a referral to the

Bortzmeyer                Expires May 31, 2014                  [Page 2]

Internet-Draft                DNS privacy                  November 2013

   nameservers, which will provide the answer.  The root name servers,
   the name servers of .com and those of 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.
   Unlike what many "DNS for dummies" articles say, the question sent to
   the root name servers is "What are the AAAA records for", not "What are the name servers of .com?".  So, the
   DNS leaks more information than it should.

   Because the DNS uses caching heavily, not all questions are sent to
   the authoritative name servers.  If the stub resolver, a few seconds
   later, asks to the resolver "What are the SRV records of _xmpp-", the resolver will remember that it knows
   the name servers of and will just query them, bypassing
   the root and .com.  Because there is typically no caching in the stub
   resolver, the resolver, unlike the authoritative servers, sees

   Almost all the DNS queries are today sent over UDP, and this has
   practical consequences if someone thinks of encrypting this traffic
   (some encryption solutions are typically done for TCP, not UDP).

   I should be noted to that DNS resolvers sometimes forward requests to
   bigger machines, with a larger and more shared cache, the forwarders.
   From the point of view of privacy, forwarders are like resolvers,
   except that the caching in the resolver before them decreases the
   amount of data they can see.

   We will use here the terminology of [RFC6973].

2.  Risks

   This draft is limited to the study of privacy risks for the end-user
   (the one performing DNS requests).  Privacy risks for the holder of a
   zone (the risk that someone gets the data) are discussed in [RFC5936]
   and in [I-D.koch-perpass-dns-confidentiality].  Non-privacy risks
   (such as cache poisoning) are out of scope.

2.1.  Data in the DNS request

   The DNS request includes many fields but two of them seem specially
   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 +
   may be source port", because the port is also in the request and can
   be used to sort out several users sharing an IP address (CGN for

Bortzmeyer                Expires May 31, 2014                  [Page 3]

Internet-Draft                DNS privacy                  November 2013

   The qname is the full name sent by the original user.  It gives
   information about what the user does ("What are the MX records of"  means he probably wants to send email to someone at, which may be a domain used by only a few persons and
   therefore very revealing).  Some qnames are more sensitive than
   others.  For instance, querying the A record of
   reveals very little (everybody visits Web sites which use Google
   Analytics) but querying the A record of www.verybad.example where
   verybad.example is the domain of an illegal or very offensive
   organization may create more problems for the user.  Another example
   is when the qname embeds the software one uses.  For instance, some
   BitTorrent clients query a SRV record for _bittorrent-

   For the communication between the stub resolver and the resolver, the
   source IP address is the one of the user's machine.  Therefore, all
   the issues and warnings about collection of IP addresses apply here.
   For the communication between the 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 a HTTP connection.  It
   is now the IP address of the resolver which, in a way "hides" the
   real user.  However, it does not always work.  Sometimes
   [I-D.vandergaast-edns-client-subnet] is used.  Sometimes the end user
   has a personal resolver on her machine.  In that case, the IP address
   is as sensitive as it is for HTTP.

2.2.  On the wire

   DNS traffic can be seen by an eavesdropper like any other traffic.
   It is typically not encrypted.  (DNSSEC, specified in [RFC4033]
   explicitely excludes confidentiality from its goals.)  So, if an
   initiator starts a HTTPS communication with a recipient, while the
   HTTP traffic will be encrypted, the DNS exchange prior to it will not
   be.  When the other protocols will become more or more privacy-aware
   and secured against surveillance, the DNS risks to become "the
   weakest link" in privacy.

   What also makes the DNS traffic different 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 resolver, or to the authoritative name servers.

   The best place, from an eavesdropper's point of view, is clearly
   between the stub resolvers and the resolvers, because he is not
   limited by DNS caching.

Bortzmeyer                Expires May 31, 2014                  [Page 4]

Internet-Draft                DNS privacy                  November 2013

   The attack surface between the stub resolver and the rest of the
   world can vary widely depending upon how the end user's computer is
   configured.  By order of increasing attack surface:

   The resolver can be on the end user's computer.  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 stub resolver to caching resolver connection is limited to
   that single machine.

   The resolver can be in the IAP (Internet Access Provider) premises.
   For most residential users and potentially other networks the typical
   case is for the end user's computer to be configured (typically
   automatically through DHCP) with the addresses of the DNS resolver 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 resolvers.

   The resolver may also be at the local network edge.  For many/most
   enterprise networks and for some residential users 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 IAP network referenced above.

   The resolver can be a public DNS service.  Some end users may be
   configured to use public DNS resolvers such as those operated by
   Google Public DNS or OpenDNS.  The end user may have configured their
   machine to use these DNS resolvers themselves - or their IAP may
   choose 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 end user's connection and the public DNS

2.3.  In the servers

   Using the terminology of [RFC6973], the DNS servers (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 [RFC6973], "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
   [RFC6973] parlance, enablers become observers when they start
   collecting data.

Bortzmeyer                Expires May 31, 2014                  [Page 5]

Internet-Draft                DNS privacy                  November 2013

   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 TODO reference and DNSmezzo TODO
   reference.  The organization managing the DNS server can use this
   data itself or it can be part of a surveillance program like PRISM
   [prism] and pass data to an outside attacker.

   Sometimes, these data are kept for a long time and/or distributed to
   third parties, for research purposes [ditl], for security analysis,
   or for surveillance tasks.  Also, there are observation points in the
   network which gather DNS data and then make it accessible to third-
   parties for research or security purposes ("passive DNS

2.3.1.  In the resolvers

   The resolvers see the entire traffic since there is typically no
   caching before them.  They are therefore well situated to observe the
   traffic.  To summarize: your resolver knows a lot about you.  The
   resolver of a large IAP, or a large public resolver can collect data
   from many users.  You may get an idea of the data collected by
   reading the privacy policy of a big public resolver [1].

2.3.2.  In the authoritative name servers

   Unlike the resolvers, they are limited by caching.  They see only a
   part of the requests.  For aggregated statistics ("what is the
   percentage of LOC queries?"), it is sufficient but it may prevent an
   observer to observe everything.  Nevertheless, the authoritative name
   servers sees a part of the traffic and this sample may be sufficient
   to defeat some privacy expectations.

   Also, the end user has typically some legal/contractual link with the
   resolver (he has chosen the IAP, or he has chosen to use a given
   public resolver) while he is often not even aware of the role of the
   authoritative name servers and their observation abilities.

   It is an interesting question whether the privacy issues are bigger
   in the root or in a large TLD.  The root sees the traffic for all the
   TLDs (and the huge amount of traffic for non-existing TLD) but a
   large TLD has less caching before it.

   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 since the authoritative name server will
   see the IP address of the end client, and not the address of a big
   resolver shared by many users.  This is no longer true if

Bortzmeyer                Expires May 31, 2014                  [Page 6]

Internet-Draft                DNS privacy                  November 2013

   [I-D.vandergaast-edns-client-subnet] is used because, in this case,
   the authoritative name server sees the original IP prefix or address
   (depending on the setup).

   As of today, all the instances of one root name server, L-root,
   receive together around 20 000 queries per second.  While most of it
   is junk (errors on the TLD name), it gives an idea of the amount of
   big data which pours into name servers.

   Many domains, including TLD, 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.  But it may be surprising for an end-user that requests to a
   given ccTLD may go to servers managed by organisations outside of the

2.3.3.  Rogue servers

   A rogue DHCP server can direct you to a rogue resolver.  Most of the
   times, it seems to be done to divert traffic, by providing lies for
   some domain names.  But it could be used just to capture the traffic
   and gather information about you.  Same thing for malwares like
   DNSchanger[dnschanger] which changes the resolver in the machine's

3.  Actual "attacks"

   Many research papers about malware detection use DNS traffic to
   detect "abnormal" behaviour that can be traced back to the activity
   of malware on infected machines.  Yes, this reasearch was done for
   the good but, technically, it is a privacy attack and it demonstrates
   the power of the observation of DNS traffic.  See [dns-footprint],
   [dagon-malware] and [darkreading-dns].

4.  Legalities

   To our knowledge, there are no specific privacy laws for DNS data.
   Interpreting general privacy laws like [data-protection-directive]
   (European Union) in the context of DNS traffic data is not an easy
   task and it seems there is no court precedent here.

5.  Possible technical solutions

   We mention here only the solutions that could be deployed in the
   current Internet.  Disruptive solutions, like replacing the DNS with
   a completely new resolution protocol, are interesting but are kept
   for a future work.  Remember that the focus of this document is on
   describing the threats, not in detailing solutions.  This section is

Bortzmeyer                Expires May 31, 2014                  [Page 7]

Internet-Draft                DNS privacy                  November 2013

   therefore non-normative and is NOT a technical specification of
   solutions.  For the same reason, there are not yet actual
   recommendations in this document.

   Raising seriously the bar against the eavesdropper will require
   SEVERAL actions.  Not one is decisive by itself but, together, they
   can have an effect.  The most important suggested here are:

      qname minimization,

      encryption of DNS traffic,

      padding (sending random queries from time to time).

   We detail some of these actions later, classified by the kind of
   observer (on the wire, in a server, etc).  Some actions will help
   against several kinds of observers.  For instance, padding, sending
   gratuitous queries from time to time (queries where you're not
   interested in the replies, just to disturb the analysis), is useful
   against all sorts of observers.  It is a costly technique, because it
   increases the traffic on the network but it seriously blurs the
   picture for the observer.

5.1.  On the wire

5.1.1.  Reducing the attack surface

   See Section 5.2.1 since the solution described there apply against
   on-the-wire eavesdropping as well as against observation by the

5.1.2.  Encrypting the DNS traffic

   To really defeat an eavesdropper, there is only one solution:
   encryption.  But, from the end user point of view, even if you check
   that your communication between your stub resolver and the resolver
   is encrypted, you have no way to ensure that the communication
   between the resolver and the autoritative name servers will be.
   There are two different cases, communication between the stub
   resolver and the resolver (no caching but only two parties so
   solutions which rely on an agreement may work) and communication
   between the resolver and the authoritative servers (less data because
   of caching, but many parties involved, so any solution has to scale
   well).  Encrypting the "last mile", between the user's stub resolver
   and the resolver may be sufficient since the biggest danger for
   privacy is between the stub resolver and the resolver, because there
   is no caching involved there.

Bortzmeyer                Expires May 31, 2014                  [Page 8]

Internet-Draft                DNS privacy                  November 2013

   The only encryption mechanism available for DNS which is today an
   IETF standard is IPsec in ESP mode.  Its deployment in the wide
   Internet is very limited, for reasons which are out of scope here.
   Still, it may be a solution for "the last mile" and, indeed, many VPN
   solutions use it this way, encrypting the whole traffic, including
   DNS to the safe resolver.  In the IETF standards, a possible
   alternative could be DTLS [RFC6347].  It enjoyed very little actual
   deployment and its interaction with the DNS has never been
   considered, studied or of course implemented.  There are also non
   standard encryption techniques like DNScrypt [dnscrypt] for the stub
   resolver <-> resolver communication or DNScurve [dnscurve] for the
   resolver <-> authoritative server communication.  It seems today that
   the possibility of massive encryption of DNS traffic is very remote.

   A last "pervasive encryption" solution for the DNS could be
   "Confidential DNS" by W.  Wijngaards which is not published yet but
   seems promising.

   Another solution would be to use more TCP for the queries, together
   with TLS [RFC5246].  DNS can run over TCP and it provides a good way
   to leverage the software and experience of the TLS world.  There have
   been discussions to use more TCP for the DNS, in light of reflection
   attacks (based on the spoofing of the source IP address, which is
   much more difficult with TCP).  For instance, a stub resolver could
   open a TCP connection with the resolver at startup and keep it open
   to send queries and receive responses.  The server would of course be
   free to tear down these connections at will (when it is under stress,
   for instance) and the client could reestablish them when necessary.
   Remember that TLS sessions can survive TCP connections so there is no
   need to restart the TLS negociation each time.  This DNS-over-TLS-
   over-TCP is already implemented in the Unbound resolver.  It is safe
   only if pipelining multiple questions over the same channel.  Name
   compression should also be disabled, or CRIME-style [crime] attacks
   can apply.

   Encryption alone does not guarantee perfect privacy, because of the
   available metadata.  For instance, the size of questions and
   responses, even encrypted, provide hints about what queries have been
   sent.  (DNScrypt uses random-length padding, and a 64 bytes block
   size, to limit this risk, but this raises other issues, for instance
   during amplification attacks.  Other security protocols use similar
   techniques, for instance ESPv3.)  Observing the periodicity of
   encrypted questions/responses also discloses the TTL, which is yet
   another hint about the queries.  Non-cached responses are disclosing
   the RTT between the resolver and authoritative servers.  This is a
   very useful indication to guess where authoritative servers are
   located.  Web pages are made of many resources, leading to multiple
   requests, whose number and timing fingerprint which web site is being

Bortzmeyer                Expires May 31, 2014                  [Page 9]

Internet-Draft                DNS privacy                  November 2013

   browsed.  So, observing encrypted traffic is not enough to recover
   any plaintext queries, but is enough to answer the question "is one
   of my employees browsing Facebook?".  Finally, attackers can perform
   a denial-of-service attack on possible targets, check if this makes a
   difference on the encrypted traffic they observe, and infer what a
   query was.

5.2.  In the servers

5.2.1.  In the resolvers

   It does not seem there is a possible solution against a leaky
   resolver.  A resolver has to see the entire DNS traffic in clear.

   The best approach to limit the problem is to have local resolvers
   whose caching will limit the leak.  Local networks should have a
   local caching resolver (even if it forwards the unanswered questions
   to a forwarder) and individual laptops can have their very own
   resolver, too.

   One mechanism to potentially mitigate on the wire attacks between
   stub resolvers and caching resolvers is to determine if the network
   location of the caching resolver can be moved closer to the end
   user's computer (reducing the attack surface).  As noted earlier in
   Section 2.2, if an end user's computer is configured with a caching
   resolver on the edge of the local network, an attacker would need to
   gain access to that local network in order to successfully execute an
   on the wire attack against the stub resolver.  On the other hand, if
   the end user's computer is configured to use a public DNS service as
   the caching resolver, the attacker needs to simply get in the network
   path between the end user and the public DNS server and so there is a
   much greater opportunity for a successful attack.  Configuring a
   caching resolver closer to the end user can also reduce the
   possibility of on the wire attacks.

5.2.2.  In the authoritative name servers

   A possible solution would be to minimize the amount of data sent from
   the resolver.  When a resolver receives the query "What is the AAAA
   record for", it sends to the root (assuming a cold
   resolver, whose cache is empty) the very same question.  Sending
   "What are the NS records for .com?"  would be sufficient (since it
   will be the answer from the root anyway).  To do so would be
   compatible with the current DNS system and therefore could be
   deployable, since it is an unilateral change to the resolvers.

   To do so, the resolver needs to know the zone cut [RFC2181].  There
   is not a zone cut at every label boundary.  If we take the name

Bortzmeyer                Expires May 31, 2014                 [Page 10]

Internet-Draft                DNS privacy                  November 2013, it is possible that there is a zone cut between
   "foo" and "bar" but not between "bar" and "example".  So, assuming
   the resolver already knows the name servers of .example, when it
   receives the query "What is the AAAA record of",
   it does not always know if the request should be sent to the name
   servers of bar.example or to those of example.  [RFC2181] suggests an
   algorithm to find the zone cut, so resolvers may try it.

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

   It can be noted that minimizing the amount of data sent also
   partially addresses the case of a wire sniffer.

   One should note that the behaviour suggested here (minimizing the
   amount of data sent in qnames) is NOT forbidden by the [RFC1034]
   (section 5.3.3) or [RFC1035] (section 7.2).  Sending the full qname
   to the authoritative name server is a tradition, not a protocol

   Another note is that the answer to the NS query, unlike the referral
   sent when the question is a full qname, is in the Answer section, not
   in the Authoritative section.  It has probably no practical

5.2.3.  Rogue servers

   Traditional security measures (do not let malware change the system
   configuration) are of course a must.  A protection against rogue
   servers announced by DHCP could be to have a local resolver, and to
   always use it, ignoring DHCP.

6.  Security considerations

   Hey, man, the entire document is about security!

7.  Acknowledgments

   Thanks to Nathalie Boulvard and to the CENTR members for the original
   work which leaded to this draft.  Thanks to Olaf Kolkman, Francis
   Dupont and Ondrej Sury for the interesting discussions.  Thanks to
   Mohsen Souissi for proofreading.  Thanks to Dan York and Frank Denis
   for good written contributions.

8.  References

Bortzmeyer                Expires May 31, 2014                 [Page 11]

Internet-Draft                DNS privacy                  November 2013

8.1.  Normative References

   [RFC1034]  Mockapetris, P., "Domain names - concepts and facilities",
              STD 13, RFC 1034, November 1987.

   [RFC1035]  Mockapetris, P., "Domain names - implementation and
              specification", STD 13, RFC 1035, November 1987.

   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119, March 1997.

   [RFC6973]  Cooper, A., Tschofenig, H., Aboba, B., Peterson, J.,
              Morris, J., Hansen, M., and R. Smith, "Privacy
              Considerations for Internet Protocols", RFC 6973, July

8.2.  Informative References

   [RFC2181]  Elz, R. and R. Bush, "Clarifications to the DNS
              Specification", RFC 2181, July 1997.

   [RFC4033]  Arends, R., Austein, R., Larson, M., Massey, D., and S.
              Rose, "DNS Security Introduction and Requirements", RFC
              4033, March 2005.

   [RFC5246]  Dierks, T. and E. Rescorla, "The Transport Layer Security
              (TLS) Protocol Version 1.2", RFC 5246, August 2008.

   [RFC5936]  Lewis, E. and A. Hoenes, "DNS Zone Transfer Protocol
              (AXFR)", RFC 5936, June 2010.

   [RFC6347]  Rescorla, E. and N. Modadugu, "Datagram Transport Layer
              Security Version 1.2", RFC 6347, January 2012.

              Koch, P., "Confidentiality Aspects of DNS Data,
              Publication, and Resolution", draft-koch-perpass-dns-
              confidentiality-00 (work in progress), November 2013.

              Contavalli, C., Gaast, W., Leach, S., and E. Lewis,
              "Client Subnet in DNS Requests", draft-vandergaast-edns-
              client-subnet-02 (work in progress), July 2013.

   [dnsop]    IETF, , "The dnsop mailing list", October 2013.


Bortzmeyer                Expires May 31, 2014                 [Page 12]

Internet-Draft                DNS privacy                  November 2013

              Dagon, D., "Corrupted DNS Resolution Paths: The Rise of a
              Malicious Resolution Authority", 2007.

              Stoner, E., "DNS footprint of malware", October 2010.

              Lemos, R., "Got Malware? Three Signs Revealed In DNS
              Traffic", May 2013.

              Wikipedia, , "DNSchanger", November 2011.

              Denis, F., "DNSCrypt", .

              Bernstein, D., "DNScurve", .

   [prism]    NSA, , "PRISM", 2007.

   [crime]    Rizzo, J. and T. Dong, "The CRIME attack against TLS",

   [ditl]     , "A Day in the Life of the Internet (DITL)", 2002.

              , "European directive 95/46/EC on the protection of
              individuals with regard to the processing of personal data
              and on the free movement of such data", November 1995.

              Weimer, F., "Passive DNS Replication", April 2005.

Author's Address

   Stephane Bortzmeyer
   Immeuble International
   Saint-Quentin-en-Yvelines  78181

   Phone: +33 1 39 30 83 46

Bortzmeyer                Expires May 31, 2014                 [Page 13]