Network Working Group                                           J. Arkko
Internet-Draft                                                A. Keranen
Intended status: Informational                                  Ericsson
Expires: January 7, 2011                                    July 6, 2010

                 Experiences from an IPv6-Only Network


   This document discusses our experiences from moving a small number of
   users to an IPv6-only network, with access to the IPv4-only parts of
   the Internet via a NAT64 device.  The document covers practical
   experiences as well as road blocks and opportunities for this type of
   a network setup.  The document also makes some recommendations about
   where such networks are applicable and what should be taken into
   account in the network design.  The document also discusses further
   work that is needed to make IPv6-only networking applicable in all

Status of this Memo

   This Internet-Draft is submitted to IETF in full conformance with the
   provisions of BCP 78 and BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as Internet-

   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."

   The list of current Internet-Drafts can be accessed at

   The list of Internet-Draft Shadow Directories can be accessed at

   This Internet-Draft will expire on January 7, 2011.

Copyright Notice

   Copyright (c) 2010 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

Arkko & Keranen          Expires January 7, 2011                [Page 1]

Internet-Draft            IPv6-Only Experiences                July 2010

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

Table of Contents

   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  3
   2.  Technology and Terminology . . . . . . . . . . . . . . . . . .  3
   3.  Network Setup  . . . . . . . . . . . . . . . . . . . . . . . .  4
     3.1.  The IPv6-Only Network  . . . . . . . . . . . . . . . . . .  5
     3.2.  DNS Operation  . . . . . . . . . . . . . . . . . . . . . .  5
   4.  General Experiences  . . . . . . . . . . . . . . . . . . . . .  7
   5.  Experiences with IPv6-Only Networking  . . . . . . . . . . . .  8
     5.1.  Operating Systems  . . . . . . . . . . . . . . . . . . . .  9
     5.2.  Instant Messaging and VoIP . . . . . . . . . . . . . . . .  9
     5.3.  Gaming . . . . . . . . . . . . . . . . . . . . . . . . . . 10
     5.4.  Appliances . . . . . . . . . . . . . . . . . . . . . . . . 11
     5.5.  Other Differences  . . . . . . . . . . . . . . . . . . . . 11
   6.  Experiences with NAT64 . . . . . . . . . . . . . . . . . . . . 11
     6.1.  IPv4 Address Literals  . . . . . . . . . . . . . . . . . . 11
     6.2.  Comparison of Web Access via NAT64 to Other Methods  . . . 12
     6.3.  Response Times . . . . . . . . . . . . . . . . . . . . . . 13
   7.  Conclusions and Recommendations  . . . . . . . . . . . . . . . 13
   8.  Security Considerations  . . . . . . . . . . . . . . . . . . . 15
   9.  IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 15
   10. References . . . . . . . . . . . . . . . . . . . . . . . . . . 16
     10.1. Normative References . . . . . . . . . . . . . . . . . . . 16
     10.2. Informative References . . . . . . . . . . . . . . . . . . 16
   Appendix A.  Acknowledgments . . . . . . . . . . . . . . . . . . . 17
   Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 17

Arkko & Keranen          Expires January 7, 2011                [Page 2]

Internet-Draft            IPv6-Only Experiences                July 2010

1.  Introduction

   This document discusses our experiences from moving a small number of
   users to an IPv6-only network, with access to the IPv4-only parts of
   the Internet via a NAT64 device.  This arrangement has been done with
   a permanent change in mind rather than as a temporary experiment,
   involves both office and home users, heterogeneous computing
   equipment, and varied applications.  We have learned both practical
   details, road blocks and opportunities, as well as more general
   understanding of when such a configuration can be recommended and
   what should be taken into account in the network design.

   The networks involved in this setup have been in dual stack mode for
   considerable amount of time, in one case for over ten years.  Our
   IPv6 connectivity is stable and in constant use with no significant
   problems.  Given that the IETF is working on technology such as NAT64
   [I-D.ietf-behave-v6v4-framework] and several network providers are
   discussing the possibility of employing IPv6-only networking, we
   decided to take our network beyond the "comfort zone" and make sure
   that we understand the implications of having no IPv4 connectivity at
   all.  This also allowed us to test a NAT64 device that is being
   developed by Ericsson.

   The main conclusion is that it is possible to employ IPv6-only
   networking, though there are a number of issues such as lack of IPv6
   support in some applications or bugs in untested parts of code.  As a
   result, dual-stack [RFC4213] remains as our recommended model for
   general purpose networking at this time, but IPv6-only networking can
   be employed by early adopters or highly controlled networks.  The
   document also suggests actions to make IPv6-only networking
   applicable in all environments.  In particular, resolving problems
   with a few key applications would have a significant impact for
   enabling IPv6-only networking for large classes of users and
   networks.  It is important that the Internet community understands
   these deployment barriers and works to remove them.

   The rest of this document is organized as follows.  Section 2
   introduces some relevant technology and terms, Section 3 describes
   the network setup, Section 4 discusses our general experiences,
   Section 5 discusses experiences related to having only IPv6
   networking available, and Section 6 discusses experiences related to
   NAT64 use.  Finally, Section 7 draws some conclusions and makes
   recommendations on when and how one should employ IPv6-only networks.

2.  Technology and Terminology

   In this document, the following terms are used.  "NAT44" refers to

Arkko & Keranen          Expires January 7, 2011                [Page 3]

Internet-Draft            IPv6-Only Experiences                July 2010

   any IPv4-to-IPv4 network address translation algorithm, both "Basic
   NAT" and "Network Address/Port Translator (NAPT)", as defined by

   "Dual Stack" refers to a technique for providing complete support for
   both Internet protocols -- IPv4 and IPv6 -- in hosts and routers

   "NAT64" refers to a Network Address Translator - Protocol Translator
   defined in [I-D.ietf-behave-v6v4-framework]
   [I-D.ietf-behave-v6v4-xlate] [I-D.ietf-behave-v6v4-xlate-stateful]
   [I-D.ietf-behave-address-format] [I-D.ietf-behave-dns64]

3.  Network Setup

   We have tested IPv6-only networking in two different network
   environments: office and home.  In both environments all hosts had
   normal dual stack native IPv4 and IPv6 Internet access already in
   place.  The networks were also already employing IPv6 in their
   servers and DNS records.  Similarly, the network was a part of white
   listing arrangement to ensure that IPv6-capable content providers
   would be able to serve their content to the network over IPv6.

   The office environment has heterogeneous hardware with PCs, laptops,
   and routers running Linux, Mac OS X, and Microsoft Windows operating
   systems.  Common uses of the network include e-mail, Secure Shell
   (SSH), web browsing, and various instant messaging (IM) and Voice
   over IP (VoIP) applications.  The hardware in the home environment
   consists of PCs, laptops and a number of server, camera, and sensor
   appliances.  The primary operating systems in this environment are
   Linux and Microsoft Windows operating systems.  Common applications
   include web browsing, streaming, instant messaging and VoIP
   applications, gaming, file storage, and various home control
   applications.  Both environments employ extensive firewalling
   practices, and filtering is applied for both IPv4 and IPv6 traffic.
   Firewall capabilities dictate some differences between the filtering
   applied for IPv4 and IPv6, however.  In addition, in the home
   environment the individual devices are directly accessible from the
   Internet on IPv6 (on select protocols such as SSH) but not on IPv4
   due to lack of available public IPv4 addresses.

   In both environments, volunteers had the possibility to opt-in for
   the IPv6-only network.  The number of users is small, there are
   roughly five permanent users and a dozen users who have been in the
   network at least for some amount of time.  Each user had to connect
   to the IPv6-only wired or wireless network, and depending on their

Arkko & Keranen          Expires January 7, 2011                [Page 4]

Internet-Draft            IPv6-Only Experiences                July 2010

   software, possibly configure their computer by indicating that there
   is no IPv4 and/or setting DNS server addresses.  The users were also
   asked to report their experiences back to the organizers.

3.1.  The IPv6-Only Network

   The IPv6-only network was provided as a parallel network on the side
   of the already existing dual stack network.  It was important to
   retain the dual stack network for the benefit of those users that did
   not decide to opt-in.  It was also important as we knew there were
   IPv4-only devices.  A separate wired access network was created using
   Virtual Local Area Networks (VLANs).  This network had its own IPv6
   prefix.  A separate wireless network was also created, bridged to the
   wired network.  With the devices that were used in our environment
   the new wireless network required additional access point hardware in
   order to accommodate advertising multiple wireless networks.  The
   simple access point model that we employed in these networks did not
   allow this on a single device.  All the secondary infrastructure
   resulted in some additional management burden and cost, however.  An
   added complexity was that the home network already employed two types
   of infrastructure, one for family members and another one for
   visitors.  In order to duplicate this model for the IPv6-only network
   there are now four separate networks, with several access points on

   A NAT64 with integrated DNS64 was installed on the edge of the IPv6-
   only networks.  No IPv4 routing or DHCP was offered on these
   networks.  The NAT64 device sends Router Advertisements (RAs)
   [RFC4861] from where the hosts learn the IPv6 prefix and can
   automatically configure IPv6 addresses for them.  Each new IPv6-only
   network needed one new /64 prefix to be used in these advertisements.
   In addition, the each NAT64 device needed another /64 prefix to be
   used for the representation of IPv4 destinations in the IPv6-only
   network.  As a result, one IPv6-only network requires an additional
   /63 of address space.  This space was easily available in our
   networks, as IPv6 allocations are on purpose made in sufficiently
   large blocks.  Additional address space needs can be accommodated
   from the existing block without registry involvement.  However, the
   additional prefixes have to be listed in the intra-domain routing
   system so that they can be reached.  In one case the increase from
   one block to multiple also made it necessary to employ an improved
   routing configuration.  In addition to routing, the new prefixes have
   to be listed in the appropriate firewall rules.

3.2.  DNS Operation

   The RAs are also used to carry DNS Configuration options [RFC5006],
   listing the DNS64 as the DNS server the hosts should use.  In

Arkko & Keranen          Expires January 7, 2011                [Page 5]

Internet-Draft            IPv6-Only Experiences                July 2010

   addition, aliases were added to the DNS64 device to allow it to
   receive packets on the well-known DNS server addresses that Windows
   operating systems use (fec0:0:0:ffff::1, fec0:0:0:ffff::2, and fec0:
   0:0:ffff::3).  DHCPv6 was not enabled in our networks due to lack of
   time, but we do recommend enabling RFC 5006, well-known addresses,
   and DHCPv6 in order to maximize the likelihood of IPv6-only hosts
   being able to use DNS without manual configuration.  DNS server
   discovery was never a problem in dual stack networks, because DNS
   servers on the IPv4 side can easily provide IPv6 information (AAAA
   records) as well.  With IPv6-only networking, it becomes crucial that
   the local DNS server can be reached via IPv6 as well.

   When a host served by the DNS64 asks for a domain name that does not
   have an AAAA (IPv6 address) record, but has an A (IPv4 address)
   record, an AAAA record is synthesized from the A record (as defined
   in DNS64 [I-D.ietf-behave-dns64]) and sent in the DNS response to the
   host.  IP packets sent to this synthesized address are routed via the
   NAT64, translated to IPv4 by the NAT64, and forwarded to the queried
   host's IPv4 address; return traffic is translated back from IPv4 to
   IPv6 and forwarded to the host behind the NAT64 (as described in
   [I-D.ietf-behave-v6v4-framework]).  This allows the hosts in the
   IPv6-only network to contact any host in the IPv4 Internet as long as
   the hosts in the IPv4 Internet have DNS address records.

   The NAT64 devices have standard dual stack connectivity and their
   DNS64 function can use both IPv4 and IPv6 when requesting information
   from DNS.  A destination that has both an A and AAAA records is not
   treated in any special manner, because the hosts in the IPv6-only
   network can contact the destination over IPv6.  Destinations with
   only an A record will be given a synthesized AAAA record as explained
   above.  However, in one of our open visitor networks we needed a
   special arrangement.  Currently, the home network gets its IPv6
   connectivity through a tunnel via the office network, and it is
   undesirable to allow outsiders to generate traffic through the office
   network, even if the traffic is just passing by and forwarded to the
   IPv6 Internet.  As a result, in the visitor network there is a
   special IPv6-only to IPv4-only configuration where the DNS64 never
   asks for AAAA records and always generates synthesized records.

      Note: This configuration may also be useful for other purposes.
      For instance, one drawback of standard behavior is that if a
      destination publishes AAAA records but has bad IPv6 connectivity,
      the hosts in the IPv6-only network have no fallback.  In the dual
      stack model a host can always try IPv4 if the IPv6 connection
      fails.  In the special configuration IPv6 is only used internally
      at the site but never across the Internet, eliminating this
      problem.  This is not a recommended mode of operation, but it is
      interesting to note that it may solve some issues.

Arkko & Keranen          Expires January 7, 2011                [Page 6]

Internet-Draft            IPv6-Only Experiences                July 2010

   Note that in NAT64 (unlike in its older variant [RFC4966]) it is
   possible to decouple the packet translation, IPv6 routing, and DNS64
   functions.  This was used by one of our users, as he is outside our
   physical network and wants to communicate directly on IPv6 where it
   is possible without having to go through our central network
   equipment.  However, to establish communications to an IPv4
   destination, our central DNS64 is used, and if there is a need to
   translate some packets, these packets find the translator device
   through normal IPv6 routing means.

4.  General Experiences

   Based on our experiences, it is possible to live (and work) using an
   IPv6-only network.  For instance, one of the authors has been in this
   network for about three months and has been able to do all his work.
   Most things work well in the new environment.  We have been unable to
   spot any practical difference in the web browsing experience, for
   instance.  E-mail, software upgrades, operating system services, many
   chat systems and media streaming work well.  On certain mobile
   handsets that we tried all applications work flawlessly even on an
   IPv6-only network.

   However, there is some pain involved and thus it is not suitable for
   everyone just yet.  Switching IPv4 off does break many things as
   well.  Some of the users in our environment left due to these issues,
   as they missed some key feature that they needed from their computing
   environment.  These issues fall in several categories:


      We saw many issues that can be classified as bugs, likely related
      to so few people having tried the software in question in an IPv6-
      only network.  For instance, some operating system facilities
      support IPv6 but have annoying problems that are only uncovered in
      IPv6-only networking.

   Lack of IPv6 Support

      We also saw many applications that do not support IPv6 at all.
      These range from minor, old tools (such as the Unix dict(1)
      command) to major applications that are important to our users
      (such as Skype) and even to entire classes of applications (many
      games have issues).

Arkko & Keranen          Expires January 7, 2011                [Page 7]

Internet-Draft            IPv6-Only Experiences                July 2010

   Protocol, Format, and Content Problems

      There are a number of protocols that carry addresses in them, and
      using these protocols through a translator can lead to problems.
      In our current network setup we did not employ any Application
      Layer Gateways (ALGs) except for FTP [I-D.ietf-behave-ftp64].
      However, we have observed a number of protocol issues with IPv4
      addresses.  For instance, some instant messaging services do not
      work due to this.  Finally, content on some web pages may refer to
      IPv4 address literals (i.e., plain IP addresses instead of host
      and domain names).  This renders some links inaccessible in an
      IPv6-only network.  While this problem is easily quantifiable in
      measurements, the authors have only run into it once during real-
      life web browsing.

   Firewall Issues

      We also saw a number of issues related to lack of features in IPv6
      support in firewalls.  In particular, while we did not experience
      any MTU and fragmentation problems in our networks, there is
      potential for generating problems, as the support for IPv6
      fragment headers is not complete in all firewalls and the NAT64
      specifications call for use the fragment header (even in
      situations where fragmentation has not yet occurred).

   In general, most of the issues relate to poor testing and lack of
   IPv6 support in some applications.  IPv6 itself and NAT64 did not
   cause any major issues for us, once our setup and NAT64 software was
   stable.  In general, the authors feel that with the exception of some
   applications, our experience with translation to reach the IPv4
   Internet has been equal to our past experiences with NAT44-based
   Internet access.  While translation implies loss of end-to-end
   connectivity, in practice direct connectivity has not been available
   to the authors in the IPv4 either Internet for a number of years.

   It should be noted that the experience with a properly configured set
   of ALGs and work-arounds such as proxies may be different.  Some of
   the problems we encountered can be solved through these means.  For
   instance, a problematic application can be configured to use a proxy
   that in turn has both IPv4 and IPv6 access.

5.  Experiences with IPv6-Only Networking

   The overall experience was as explained above.  The remainder of this
   section discusses specific issues.

Arkko & Keranen          Expires January 7, 2011                [Page 8]

Internet-Draft            IPv6-Only Experiences                July 2010

5.1.  Operating Systems

   Even operating systems have some minor problems with IPv6.  For
   example, in Linux RA information was not automatically updated when
   the network changes while the computer is on, in Ubuntu and Mac OS X
   the network manager needed to be explicitly told to not expect IPv4,
   RA-based DNS discovery does not come as default setting in Ubuntu,
   and on Microsoft Windows 7 we experienced problems when relying on
   default, well-known DNS server addresses.  Without manual
   configuration, the host was unable to use the DNS addresses, even
   though the system displays them as current DNS server addresses.

5.2.  Instant Messaging and VoIP

   By far the biggest complaint in our group of users was that Skype
   stopped working.  In some environments even Skype can be made to work
   through a proxy configuration, and this was verified in our setting
   but not used as a permanent solution.  More generally, we tested a
   number of instance messaging applications on IPv6 and the test
   results can be found from Table 1.

     SYSTEM                                 STATUS

     Facebook on the web (http)               OK
     Facebook via a client (xmpp)             OK chat service (xmpp)           OK
     Gmail chat on the web (http)             OK
     Gmail chat via a client (xmpp)           OK
     Gtalk client                           NOT OK
     AIM (AOL)                              NOT OK
     ICQ (AOL)                              NOT OK
     Skype                                  NOT OK
     MSN                                    NOT OK

   Table 1. Instant Messaging Applications in an IPv6-Only Network

   Packet tracing revealed that the issues in AIM, ICQ, and MSN appear
   to be related to passing literal IPv4 addresses in the protocol.  It
   remains to be determined whether this can be solved through
   configuration, proxies, or ALGs.  The problem with the Gtalk client
   is that the software does not support IPv6 connections at this
   moment.  We are continuing our tests with additional applications
   such as Webex and Sametime, but have either not completed the work or
   have so far inconclusive results.  One problem in running these tests
   is to ensure that we can distinguish IPv6 and NAT64 issues from other
   issues, such as a generic issue on a given operating system platform.

Arkko & Keranen          Expires January 7, 2011                [Page 9]

Internet-Draft            IPv6-Only Experiences                July 2010

5.3.  Gaming

   Another class of applications that we tried was games.  We tried both
   web-based gaming and standalone gaming applications that have a
   "network" / "Internet" or "LAN" gaming modes.  The results are shown
   in Table 2.

     SYSTEM                                           STATUS

     Web-based (e.g. armorgames)                        OK
     Runescape (on the web)                           NOT OK
     Flat out 2                                       NOT OK
     Battlefield                                      NOT OK
     Secondlife                                       NOT OK
     Guild Wars                                       NOT OK
     Age of Empires                                   NOT OK
     Star Wars: Empire at War                         NOT OK
     Crysis                                           NOT OK
     Lord of the Rings: Conquest                      NOT OK
     Rome Total War                                   NOT OK
     Lord of the Rings: Battle for Middle Earth 2     NOT OK

   Table 2. Gaming Applications in an IPv6-Only Network

   Most web-based games worked well, as expected from our earlier good
   general web experience.  However, we were also able to find one web-
   based game that failed to work (Runescape).  This particular game is
   a Java application that fails on an attempt to perform a HTTP GET
   request.  The reason remains unclear, but a likely theory is the use
   of an IPv4-literal in the application itself.

   The experience with standalone games was far more discouraging.
   Without exception all games failed to enable either connections to
   ongoing games in the Internet or even LAN-based connections to other
   computers in the same IPv6-only LAN segment.  This is somewhat
   surprising, and the result require further verification.
   Unfortunately, the games provide no diagnostics about their
   operation, so it is hard to guess what is going on.  It is possible
   that their networking code employs older APIs that cannot use IPv6
   addresses [RFC4038].  The inability to provide any LAN-based
   connectivity is even more surprising, as this must mean that they are
   unable to use IPv4 link local connectivity, which should have been
   available to the devices.  (IPv4 was not blocked; just that no DHCP
   answers were provided on IPv4.)

Arkko & Keranen          Expires January 7, 2011               [Page 10]

Internet-Draft            IPv6-Only Experiences                July 2010

5.4.  Appliances

   There are also problems with different appliances such as webcams.
   Many of them do not support IPv6 and hence will not work in an IPv6-
   only network.  Also not all firewalls support IPv6.  Or even if they
   do, they may still experience issues with some aspects of IPv6 such
   as fragments.

   Some of these issues are easily solved when the appliance works as a
   server, such as what most webcams and our sensor gateway devices do.
   We placed the appliance in the IPv4 part of the network, added its
   name to the local DNS, and simply allowed devices from the IPv6-only
   network reach it through NAT64.

5.5.  Other Differences

   One thing that becomes simplified in an IPv6-only network is source
   address selection [RFC3484].  As there is no IPv4 connectivity, the
   host only needs to consider its IPv6 source address.  For global
   communications there is typically just one possible source address.

6.  Experiences with NAT64

   After correcting some initial bugs and stability issues, the NAT64
   operation itself has been relatively problem free.  There have been
   no unexplained DNS problems or lost sessions.  With the exception of
   the specific applications mentioned above and IPv4 literals, the user
   experience has been in line with using IPv4 Internet through a NAT44
   device.  These failures with the specific applications are clearly
   very different from the IPv4 experience, however.

   The rest of this section discusses our measurements on specific

6.1.  IPv4 Address Literals

   While browsing in general works, IPv4 literals embedded in the HTML
   code may break some parts of the web pages when using IPv6-only
   access.  This happens because the DNS64 can not synthesize AAAA
   records for the literals since the addresses are not queried from the
   DNS.  Luckily, the IPv4 literals seem to be fairly rarely
   encountered, at least so that they would be noticed, with regular

   We have attempted to measure the likelihood of running into an IPv4
   literal in the web.  To do this, we took the top 1,000 and 10,000 web
   sites from the Alexa popular web site list.  With 1,000 top sites,

Arkko & Keranen          Expires January 7, 2011               [Page 11]

Internet-Draft            IPv6-Only Experiences                July 2010

   0.2% needed an IPv4 literal to render all components in their top
   page (i.e., images, videos, JavaScript and Cascading Style Sheet
   (CSS) files, etc.).  With 10,000 top sites, this number increases to

   However, it is not clear what conclusions can be made about this.  It
   is often the case that there are unresolvable or inaccessible
   components on a web page anyway for various reasons, and to
   understand the true impact we would have to know how "important" a
   given page component was.  Also, we did not measure the number of
   links with IPv4 literals on these pages, nor did we attempt to search
   the site in any thorough manner for these literals.

   As noted, personal anecdotal evidence says IPv4 literals are not a
   big problem.  But clearly, cleaning the most important parts of the
   web from IPv4 literals would be useful.  With tools such as the
   popular web site list, some user pressure, and co-operation from the
   content providers the most urgent part of the problem could hopefully
   be solved as a one-time effort.

6.2.  Comparison of Web Access via NAT64 to Other Methods

   We also compared how well the web works behind a NAT64 compared to
   IPv4-only and native IPv6 access.  For this purpose, we used wget to
   go through the same top web site lists as in section Section 6.1,
   again downloading everything needed to render their top page.  The
   tests were repeated and an average was calculated over all runs.
   Separate tests were done with an IPv4-only network, an IPv6-only
   network, and an IPv6-only network with NAT64.

   When accessed with an IPv4-only network, our tests show that 1.9% of
   the sites experienced some sort of error or failure.  The failure
   could be that the whole site is not accessible, or just that a single
   image (e.g., advertisement banner) was not loaded properly.  It
   should also be noted that access through wget is somewhat different
   to access via a regular browser: some web sites refuse to serve
   content to wget, browsers typically have DNS heuristics to fill in
   "www." in front of a domain name where needed, and so on.  In
   addition to missing advertisement banners, temporary routing glitches
   and other mistakes, these differences also help explain the reason
   for the high baseline error rate in this test.  It should also be
   noted that variations in wget configuration options produced highly
   different results, but we believe that the options we settled on bear
   closest resemblance to real world browsing.

   When we tried to access the same sites with native IPv6 (without
   NAT64), 96% of the sites failed to load correctly.  This was as
   expected, given that most of the Internet content is not available on

Arkko & Keranen          Expires January 7, 2011               [Page 12]

Internet-Draft            IPv6-Only Experiences                July 2010

   IPv6.  The few exceptions included, for instance, sites managed by

   When the sites were accessed from IPv6-only network via a NAT64
   device, the failure rate increased to 2.1%.  Most of these failures
   appear to be due to IPv4 address literals, and the increased failure
   rate matches that of IPv4 literal occurrence in the same set of top
   web sites.  With the top 10,000 sites the failure rate with NAT64
   increases similarly to our test on IPv4 address literals.

6.3.  Response Times

   One important set of measurements remains for future work, however.
   It would be useful to understand the effect of DNS64 and NAT64 to
   response time and end-to-end communication delays.  Some users have
   anecdotal reports of slow web browsing response times, but we have
   been unable to determine if this was due to the IPv6-only network
   mechanisms or for some other reason.  Measurements on pure DNS
   response times and packet round-trip delays does not show a
   significant difference to a NAT44 environment.

   It would be particularly interesting to measure delays in the context
   of dual stack vs. NAT64-based IPv6-only networking.  With dual stack,
   broken IPv6 connectivity can be repaired by falling back to IPv4 use.
   With NAT64 this is not always possible as discussed in Section 3.2.

7.  Conclusions and Recommendations

   The main conclusion is that it is possible to employ IPv6-only
   networking.  For large classes of applications there are no downsides
   or the downsides are negligible.  We have been unable to spot any
   practical difference in the web browsing experience, for instance.
   And IPv6 usage -- be it in dual stack or IPv6-only form -- comes with
   inherent advantages, such as enabling direct end-to-end connectivity.
   In our case we employed this by enabling direct connectivity to
   devices in a home network from anywhere in the Internet.  There are,
   however, a number of issues as well such as lack of IPv6 support in
   some applications or bugs in untested parts of code.

   Our experience with IPv6-only networking confirms that dual stack
   should still be our recommended model for general purpose networking
   at this time.  But IPv6-only networking can be employed by early
   adopters or highly controlled networks.  One example such controlled
   networks are mobile networks with operator-driven selection of
   handsets.  For instance, on some handsets that we tested we were
   unable to see any functional difference between IPv4 and IPv6, today.

Arkko & Keranen          Expires January 7, 2011               [Page 13]

Internet-Draft            IPv6-Only Experiences                July 2010

   Our recommendations apply at the present time.  With effort and time,
   deployment barriers can be removed and IPv6-only networking becomes
   applicable in all networking situations.

   Some of the improvements are already in process in the form of new
   products and additional IPv6 support.  For instance, we expect that
   the handset market will have a much higher number of IPv6-capable
   devices next year.  But some of the changes do not come without the
   community spending additional effort.  We have identified a number of
   actions that should be taken to improve the state of IPv6-only
   networking.  These include:

   DNS Discovery

      The state of DNS discovery continues to be one of the main
      barriers for easy adoption of IPv6-only networking.  Since DNS
      discovery is not a problem in dual stack networking, there has
      been too little effort in testing and deploying the necessary
      components.  For instance, it would be useful if RA-based DNS
      discovery came as a standard feature and not as an option in
      Linux.  Our hope is that ongoing standardization of the RA-based
      DNS discovery at the IETF will help this happen.  Similar issues
      face other operating systems.  The authors believe that at this
      time, prudent operational practices call for maximizing the number
      of offered automatic configuration mechanisms in the network side.
      It might be useful for an IETF document to provide guidance on
      operating DNS in IPv6-only networks.

   Network Managers

      Another key software component are the various network management
      and attachment tools in operating systems.  These tools generally
      have the required functionality, but do not always appear to have
      been tested very extensively on IPv6 or let alone IPv6-only
      networks.  Further work is required here.

   Application Support

      But by far the most important action for at least our group of
      users would be to bring some key applications to a state where
      they can easily run on IPv6-only networks and behind a NAT64.  For
      instance, instant messaging and VoIP applications and games.  In
      some cases it may also be necessary add support for new types of

Arkko & Keranen          Expires January 7, 2011               [Page 14]

Internet-Draft            IPv6-Only Experiences                July 2010

   IPv4 Literals

      The web should be cleaned from IPv4 literals.

   Measurements and Analysis

      It is also important to continue with testing, measurements, and
      analysis of what Internet technology works in IPv6-only networks,
      to what extent, at what speed, and where the remaining problems


      It is also useful to provide guidance for network administrators
      and users on how to turn on IPv6-only networking.

   As can be seen from the above list, there are only minor things that
   can be done through standardization.  Most of the effort is practical
   and centers around improving various implementations.

8.  Security Considerations

   The use of IPv6 instead of IPv4 by itself does not make a big
   security difference.  The main security requirement is that,
   naturally, network security devices need to be able to deal with IPv6
   in these networks.  This is though already required in all dual-stack
   networks.  As noted it is important to ensure firewall capabilities,
   for instance.

   In our experience many of the critical security functions in a
   network end up being on the dual stack part of the network anyway.
   For instance, our mail servers obviously still have to be able to
   communicate with both the IPv4 and IPv6 Internet, and as a result
   they and the associated spam & filtering components are not in the
   IPv6-only part of the network.

9.  IANA Considerations

   This document has no IANA implications.

10.  References

Arkko & Keranen          Expires January 7, 2011               [Page 15]

Internet-Draft            IPv6-Only Experiences                July 2010

10.1.  Normative References

   [RFC2663]  Srisuresh, P. and M. Holdrege, "IP Network Address
              Translator (NAT) Terminology and Considerations",
              RFC 2663, August 1999.

   [RFC3484]  Draves, R., "Default Address Selection for Internet
              Protocol version 6 (IPv6)", RFC 3484, February 2003.

   [RFC4213]  Nordmark, E. and R. Gilligan, "Basic Transition Mechanisms
              for IPv6 Hosts and Routers", RFC 4213, October 2005.

   [RFC5006]  Jeong, J., Park, S., Beloeil, L., and S. Madanapalli,
              "IPv6 Router Advertisement Option for DNS Configuration",
              RFC 5006, September 2007.

10.2.  Informative References

   [RFC4038]  Shin, M-K., Hong, Y-G., Hagino, J., Savola, P., and E.
              Castro, "Application Aspects of IPv6 Transition",
              RFC 4038, March 2005.

   [RFC4861]  Narten, T., Nordmark, E., Simpson, W., and H. Soliman,
              "Neighbor Discovery for IP version 6 (IPv6)", RFC 4861,
              September 2007.

   [RFC4966]  Aoun, C. and E. Davies, "Reasons to Move the Network
              Address Translator - Protocol Translator (NAT-PT) to
              Historic Status", RFC 4966, July 2007.

              Baker, F., Li, X., Bao, C., and K. Yin, "Framework for
              IPv4/IPv6 Translation",
              draft-ietf-behave-v6v4-framework-09 (work in progress),
              May 2010.

              Bao, C., Huitema, C., Bagnulo, M., Boucadair, M., and X.
              Li, "IPv6 Addressing of IPv4/IPv6 Translators",
              draft-ietf-behave-address-format-08 (work in progress),
              May 2010.

              Bagnulo, M., Sullivan, A., Matthews, P., and I. Beijnum,
              "DNS64: DNS extensions for Network Address Translation
              from IPv6 Clients to IPv4 Servers",
              draft-ietf-behave-dns64-09 (work in progress), March 2010.

Arkko & Keranen          Expires January 7, 2011               [Page 16]

Internet-Draft            IPv6-Only Experiences                July 2010

              Li, X., Bao, C., and F. Baker, "IP/ICMP Translation
              Algorithm", draft-ietf-behave-v6v4-xlate-20 (work in
              progress), May 2010.

              Bagnulo, M., Matthews, P., and I. Beijnum, "Stateful
              NAT64: Network Address and Protocol Translation from IPv6
              Clients to IPv4 Servers",
              draft-ietf-behave-v6v4-xlate-stateful-11 (work in
              progress), March 2010.

              Beijnum, I., "IPv6-to-IPv4 translation FTP
              considerations", draft-ietf-behave-ftp64-03 (work in
              progress), May 2010.

Appendix A.  Acknowledgments

   The authors would like to thank the many people who have engaged in
   discussions around this topic, and particularly the people who were
   involved in building some of the new tools used in our network, our
   users who were interested in going where only few had dared to
   venture before, or people who helped us in this effort.  In
   particular, we would like to thank Martti Kuparinen, Tero Kauppinen,
   Heikki Mahkonen, Jan Melen, Fredrik Garneij, Christian Gotare, Teemu
   Rinta-Aho, Petri Jokela, Mikko Sarela, Olli Arkko, Lasse Arkko, and
   Cameron Byrne.

Authors' Addresses

   Jari Arkko
   Jorvas  02420


   Ari Keranen
   Jorvas  02420


Arkko & Keranen          Expires January 7, 2011               [Page 17]