ICANN/DNSO
DNSO Mailling lists archives

[ga]


<<< Chronological Index >>>    <<< Thread Index >>>

[ga] Overcoming IPv6 Security Threat


Thanks to everyone who helped out.

cheers
joe baptista


>http://www.circleid.com/articles/2533.asp
>
>Overcoming IPv6 Security Threat
>
>September 12, 2002  |  By Joe Baptista
>
>Technology rags and industry pundits see IPv6 (Internet Protocol version
>6) as the future of networking, but Daniel Golding a participant of the
>North American Network Operators' Group (NANOG) thinks it's a "solution in
>search of a problem". Many others have argued IPv6 is a problem in itself
>and it is unlikely the protocol will gain wide acceptance in the short
>term.
>
>IPv6 does solve many of the problems with the current version of IPv4
>(Internet Protocol version 4). Its purpose is to expand address space and
>fix the IPv4 address depletion problem, which many techies claim, was due
>to mismanagement. The industry's goal is to use the very large address
>allocation pool in IPv6 to expand the capabilities of the Internet to
>enable a variety of peer-to-peer and mobile applications including
>cellular phone technology and home networking.
>
>IPv6, a suite of protocols for the network layer, uses IPv4 gateways to
>interconnect IPv6 nodes and comes prepackaged with some popular operating
>systems. This includes almost all Unix flavors, some Windows versions and
>Mac OS. Some vendors offer upgrades to older operating systems. Trumpet
>Software International in Tasmania Australia manufactures a Trumpet
>Winsock version that upgrades old Windows 95/98 and NT systems to the
>current IPv6 standard.
>
>IPv6 has suffered bad press over privacy issues. Jim Fleming, the inventor
>of IPv8, a competing protocol, sees many hazards and privacy flaws in
>existing IPv6 implementations. IPv6 address space in some cases uses an ID
>(identifier) derived from your hardware or phone "that allows your packets
>to be traced back to your PC or cell-phone" said Fleming. Potential abuse
>to user privacy exists as a hardware ID wired into the IPv6 protocol can
>be used to determine the manufacturer, make and model number, and value of
>the hardware equipment being used. Fleming warns users to think twice
>before they buy themselves a used Laptop computer and inherit all the
>prior surfing history of the previous user!
>
>IPv6 uses 128 bits to provide addressing, routing, and identification
>information on a computer interface or network card. The 128 bits are
>divided into the left 64 and the right 64. Some IPv6 systems use the right
>64 bits to store an IEEE defined global identifier (EUI64). This
>identifier is composed of company id value assigned to a manufacturer by
>the IEEE Registration Authority. The 64-bit identifier is a concatenation
>of the 24-bit company identification value and a 40-bit extension
>identifier assigned by the organization with that company identification
>assignment. The 48-bit MAC address of your network interface card may also
>be used to make up the EUI64.
>
>In the early stages of IPv6 development, Bill Frezza a General Partner
>with the venture capital firm, Adams Capital Management warned software
>developers that if privacy issues are not properly addressed, the
>migration to IPv6 "will blow up in their face"! Leah Gallegos agrees that
>while "expanding the address space is necessary the use of the address for
>ID and tracking is horrific". Gallegos the operator of the top-level
>domain .BIZ and a Director of the Top Level Domain Association cautions
>network administrators that they should refuse to implement IPv6 unless
>these issues are properly addressed.
>
>Privacy concerns prompted the creation of new standards, which provide
>privacy extensions to IPv6 devices. Thomas Narten and Track Draves of
>Microsoft Research published a procedure to ensure privacy of IPv6 users.
>Narten, IBM's technical lead on IPv6 and an Area Director for the Internet
>Engineering Task Force (IETF), agrees "IPv6 address can, in some cases,
>include an identifier derived from a hardware address". But Narten points
>out that a hardware address is not required. "In cases where using a
>permanent identifier is a problem", said Narten "RFC 3041 addresses should
>be used".
>
>RFC 3041 titled "Privacy Extensions for Stateless Address
>Autoconfiguration in IPv6" was published this past January 2001 by the
>IETF. It is an algorithm developed jointly by Narten and Draves which
>generates randomized interface identifiers and temporary addressees during
>a user session. This would eliminate the concerns privacy advocates have
>with IPv6.
>
>Unfortunately RFC 3041 is not widely implemented. But Narten expects major
>vendors to incorporate his privacy standard and offered that Microsoft
>implemented privacy extensions "and apparently intends to make it part of
>their standard stuff". Narten also assisted in the drafting of
>recommendations for some second and third generation cellular phones
>recently approved for publication by the Internet Engineering Steering
>Group. That document recommends that RFC 3041 be implemented as part of
>cellular phone technology but he did not know what direction cell phones
>manufacturers were taking. "I suspect that client vendors will generally
>implement it because of the potential bad PR if they don't" said Narten.
>
>Another obstacle raised by NANOG operators is that there is currently no
>commercial demand for IPv6 at this time. Dave Israel, a Data Network
>Engineer and regular participant on NANOG lists, sees no immediate demand
>for IPv6 services. "The only people who ask me about IPv6", said Israel
>"are people who have heard something about it from some tech-magazine and
>want the newest thing". Israel says he sees no commercial demand for a v6
>backbone.
>
>Daniel Golding, another NANOG participant agrees, "v6 deployment is being
>encouraged by some countries, and the spread of 3G (cellular technology)
>is helping things along, but we have yet to see really widespread v6
>deployments anywhere". Golding sees major backbone networks deploying IPv6
>when it makes economic sense for them to do so. "Right now", said Golding
>"there is no demand and no revenue upside. I don't expect this to change
>in the near future".
>
>Most on NANOG agree the roadblock seems to be a lack of ISPs that offer
>IPv6 services. Stephen Sprunk, a Network Design Consultant with Cisco's
>Advanced Services group sees the "greater adoption of always-on broadband
>access will be the necessary push" to get IPv6 off the ground. "Enterprise
>networks will not be the driver for ISPs to go to IPv6" said Sprunk and
>"NAT is too entrenched". Network Address Translation (NAT) is a method of
>connecting multiple computers to the Internet (or any other IP network)
>using one IPv4 address.
>
>Vint Cerf senior vice president of architecture & technology at WorldCom
>has been using IPv6 for about four years. IPv6 has been a key element for
>some of WorldCom's Government customers. Cerf thinks IPv6 supporters have
>a lot of work ahead to achieve successful deployment of the protocol. He
>expects "that over the next several years we will see a lot of consumer
>devices set up to work with IPv6" and "cell phones are likely candidates,
>as are radio-enabled PDAs".
>
>-EOF

The dot.GOD Registry, Limited
http://www.dot-god.com/


<<< Chronological Index >>>    <<< Thread Index >>>