Introduction to IPv6

Outline
 Protocol Background
 Technology Highlights
 Enhanced Capabilities
 Transition Issues
 Next Steps
Background
Why a New IP?
 1991 – ALE WG studied projections about
address consumption rate showed exhaustion
by 2008.

 Bake-off in mid-1994 selected approach of a

new protocol over multiple layers of
encapsulation.
What Ever Happened to IPv5?
0 IP March 1977 version
(deprecated)
1 IP January 1978 version (deprecated)
2 IP February 1978 version A (deprecated)
3 IP February 1978 version B (deprecated)
4 IPv4 September 1981 version (current widespread)
5 ST Stream Transport (not a new IP, little
use)
6 IPv6 December 1998 version (formerly SIP, SIPP)
7 CATNIP IPng evaluation (formerly TP/IX; deprecated)
8 Pip IPng evaluation (deprecated)
9 TUBA IPng evaluation (deprecated)
10-15 unassigned
What about technologies & efforts
to slow the consumption rate?
 Dial-access / PPP / DHCP
 Provides temporary allocation aligned with actual endpoint use.
 Strict allocation policies
 Reduced allocation rates by policy of ‘current-need’ vs. previous
policy based on ‘projected-maximum-size’.
 CIDR
 Aligns routing table size with needs-based address allocation policy.
Additional enforced aggregation actually lowered routing table
growth rate to linear for a few years.
 NAT
 Hides many nodes behind limited set of public addresses.
What did intense conservation efforts
of the last 5 years buy us?
 Actual allocation history
 1981 – IPv4 protocol published
 1985 ~ 1/16 total space
 1990 ~ 1/8 total space
 1995 ~ 1/4 total space
 2000 ~ 1/2 total space

 The lifetime-extending efforts & technologies

delivered the ability to absorb the dramatic
growth in consumer demand during the late 90’s.
In short they bought – TIME –
Would increased use of
NATs be adequate?
 NO!
 NAT enforces a ‘client-server’ application model where the server has
topological constraints.
 They won’t work for peer-to-peer or devices that are “called” by others
(e.g., IP phones)
 They inhibit deployment of new applications and services, because all
NATs in the path have to be upgraded BEFORE the application can
be deployed.
 NAT compromises the performance, robustness, and security of the
Internet.
 NAT increases complexity and reduces manageability of the local
network.
 Public address consumption is still rising even with current NAT
deployments.
What were the goals of a
new IP design?
 Expectation of a resurgence of “always-on”
technologies
 xDSL, cable, Ethernet-to-the-home, Cell-phones, etc.
 Expectation of new users with multiple devices.
 China, India, etc. as new growth
 Consumer appliances as network devices
 (1015 endpoints)
 Expectation of millions of new networks.
 Expanded competition and structured delegation.
 (1012 sites)
Return to an End-to-End
Architecture
New Technologies/Applications for Home Users
‘Always-on’—Cable, DSL, Ethernet@home, Wireless,…

Always-on Devices
Need an Address
When You Call Them

Global
Addressing
Realm
Why is a larger address space
needed?
 Overall Internet is still growing its user base
 ~320 million users in 2000 : ~550 million users by 2005
 Users expanding their connected device count
 405 million mobile phones in 2000, over 1 billion by 2005
 UMTS Release 5 is Internet Mobility, ~ 300M new Internet connected
 ~1 Billion cars in 2010
 15% likely to use GPS and locality based Yellow Page services
 Billions of new Internet appliances for Home users
 Always-On ; Consumer simplicity required
 Emerging population/geopolitical & economic drivers
 MIT, Xerox, & Apple each have more address space than all of China
 Moving to an e-Economy requires Global Internet accessibility
 Why Was 128 Bits Chosen
as the IPv6 Address Size?
Proposals for fixed-length, 64-bit addresses
 Accommodates 1012 sites, 1015 nodes, at .0001 allocation efficiency (3
orders of mag. more than IPng requirement)
 Minimizes growth of per-packet header overhead
 Efficient for software processing on current CPU hardware
Proposals for variable-length, up to 160 bits
 Compatible with deployed OSI NSAP addressing plans
 Accommodates auto-configuration using IEEE 802 addresses
 Sufficient structure for projected number of service providers
Settled on fixed-length, 128-bit addresses
 (340,282,366,920,938,463,463,374,607,431,768,211,456 in all!)
Benefits of
128 bit Addresses
 Room for many levels of structured hierarchy
and routing aggregation
 Easy address auto-configuration
 Easier address management and delegation
than IPv4
 Ability to deploy end-to-end IPsec
(NATs removed as unnecessary)
Incidental Benefits of
New Deployment
 Chance to eliminate some complexity in IP
header
 improve per-hop processing
 Chance to upgrade functionality
 multicast, QoS, mobility
 Chance to include new features
 binding updates
Summary of Main IPv6 Benefits
 Expanded addressing capabilities
 Structured hierarchy to manage routing table growth
 Serverless autoconfiguration and reconfiguration
 Streamlined header format and flow identification
 Improved support for options / extensions
IPv6 Advanced Features
 Source address selection
 Mobility - More efficient and robust mechanisms
 Security - Built-in, strong IP-layer encryption and
authentication
 Quality of Service
 Privacy Extensions for Stateless Address
Autoconfiguration (RFC 3041)
IPv6 Markets
 Home Networking
 Set-top box/Cable/xDSL/Ether@Home
 Residential Voice over IP gateway
 Gaming (10B$ market)
 Sony, Sega, Nintendo, Microsoft
 Mobile devices
 Consumer PC
 Consumer Devices
 Sony (Mar/01 - …energetically introducing IPv6 technology into hardware products …)
 Enterprise PC
 Service Providers
 Regional ISP, Carriers, Mobile ISP, and Greenfield ISP’s
IPv6 Markets
 Academic NRN:
 Internet-II (Abilene, vBNS+), Canarie*3, Renater-II, Surfnet, DFN,
CERNET,… 6REN/6TAP
 Geographies & Politics:
 Prime Minister of Japan called for IPv6 (taxes reduction)
 EEC summit PR advertised IPv6 as the way to go for Europe
 China Vice minister of MII deploying IPv6 with the intent to take a
leadership position and create a market force
 Wireless (PDA, Mobile, Car,...):
 Multiple phases before deployment
 RFP -> Integration -> trial -> commercial
 Requires ‘client devices’, eg. IPv6 handset ?
Outline
 Protocol Background
 Technology Highlights
 Enhanced Capabilities
 Transition Issues
 Next Steps
A new Header
The IPv6 Header
40 Octets, 8 fields

0 4 12 16 24 31

Version Class Flow Label

Payload Length Next Header Hop Limit

128 bit Source Address

128 bit Destination Address

The IPv4 Header
20 octets + options : 13 fields, including 3 flag bits

0 4 8 16 24 31

Ver IHL Service Type Total Length

Identifier Flags Fragment Offset

Time to Live Protocol Header Checksum

32 bit Source Address

32 bit Destination Address

Options and Padding

shaded fields are absent from IPv6 header

Summary of Header Changes
between
Streamlined IPv4 & IPv6
 Fragmentation fields moved out of base header
 IP options moved out of base header
 Header Checksum eliminated
 Header Length field eliminated
 Length field excludes IPv6 header
 Alignment changed from 32 to 64 bits
 Revised
 Time to Live ’ Hop Limit
 Protocol ’ Next Header
 Precedence & TOS ’ Traffic Class
 Addresses increased 32 bits ’ 128 bits
 Extended
 Flow Label field added
Extension Headers
IPv6 header TCP header + data
next header =
TCP

IPv6 header Routing header TCP header + data

next header = next header =
Routing TCP

IPv6 header Routing header Fragment header fragment of TCP

next header = next header = next header = header + data
Routing Fragment TCP
Extension Headers (cont.)
 Generally processed only by node identified in
IPv6 Destination Address field => much lower
overhead than IPv4 options processing
 exception: Hop-by-Hop Options header
 Eliminated IPv4’s 40-byte limit on options
 in IPv6, limit is total packet size,
or Path MTU in some cases
 Currently defined extension headers:
 Hop-by-Hop Options, Routing, Fragment,
Authentication, Encryption, Destination Options
Fragment Header

Next Header Reserved Fragment Offset 00M

Original Packet Identifier

 though discouraged, can use IPv6 Fragment

header to support upper layers that do not (yet)
do path MTU discovery
 IPv6 frag. & reasm. is an end-to-end function;
routers do not fragment packets en-route if too
big—they send ICMP “packet too big” instead
Routing Header
Routing
 Same “longest-prefix match” routing as IPv4
CIDR
 Straightforward changes to existing IPv4 routing
protocols to handle bigger addresses
 unicast: OSPF, RIP-II, IS-IS, BGP4+, …
 multicast: MOSPF, PIM, …
 Use of Routing header with anycast addresses
allows routing packets through particular regions
 e.g., for provider selection, policy, performance, etc.
Routing Header
Next Header Hdr Ext Len Routing Type Segments Left
Reserved

Address[0]

Address[1]

•
•
•
Example of Using the Routing
Header
S A

D
Addressing
Some Terminology
node a protocol module that implements IPv6
router a node that forwards IPv6 packets not explicitly
addressed to itself
host any node that is not a router
link a communication facility or medium over which
nodes can communicate at the link layer,
i.e., the layer immediately below IPv6
neighbors nodes attached to the same link
interface a node’s attachment to a link
address an IPv6-layer identifier for an interface or a set
of interfaces
Text Representation of
Addresses
“Preferred” form: 1080:0:FF:0:8:800:200C:417A

Compressed form: FF01:0:0:0:0:0:0:43

becomes FF01::43

IPv4-compatible: 0:0:0:0:0:0:13.1.68.3
or ::13.1.68.3
IPv6 - Addressing Model
 Addresses are assigned to interfaces
No change from IPv4 Model
 Interface ‘expected’ to have multiple addresses

 Addresses have scope

Link Local
Site Local
Global Site-Local Link-Local
Global

 Addresses have lifetime

Valid and Preferred lifetime
Types of IPv6 Addresses
 Unicast
 Address of a single interface
 Delivery to single interface

 Multicast
 Address of a set of interfaces
 Delivery to all interfaces in the set

 Anycast
 Address of a set of interfaces
 Delivery to a single interface in the set

 No more broadcast addresses

Interface Address set
 Loopback
 (only assigned to a single virtual interface per node)

 Link local
 Site local
 Auto-configured 6to4
 (if IPv4 public is address available)

 Auto-configured IPv4 compatible

 (operationally discouraged)

 Solicited node Multicast

 All node multicast
 Global anonymous
 Global published
Source Address Selection Rules
 Rule 1: Prefer same address
 Rule 2: Prefer appropriate scope
 Smallest matching scope
 Rule 3: Avoid deprecated addresses
 Rule 4: Prefer home addresses
 Rule 5: Prefer outgoing interface
 Rule 6: Prefer matching label from policy table
 Native IPv6 source > native IPv6 destination
 6to4 source > 6to4 destination
 IPv4-compatible source > IPv4-compatible destination
 IPv4-mapped source> IPv4-mapped destination
 Rule 7: Prefer temporary addresses
 Rule 8: Use longest matching prefix

 Local policy may override

Destination Address Selection
Rules
 Rule 1: Avoid unusable destinations
 Rule 2: Prefer matching scope
 Rule 3: Avoid dst with matching deprecated src address
 Rule 4: Prefer home addresses
 Rule 5: Prefer matching label from policy table
 Native IPv6 source > native IPv6 destination
 6to4 source > 6to4 destination
 IPv4-compatible source > IPv4-compatible destination
 IPv4-mapped source> IPv4-mapped destination
 Rule 6: Prefer higher precedence
 Rule 7: Prefer smaller scope
 Rule 8: Use longest matching prefix
 Rule 9: Order returned by DNS

 Local policy may override

Address Type Prefixes
Address type Binary prefix
IPv4-compatible 0000...0 (96 zero bits)
global unicast 001
link-local unicast 1111 1110 10
site-local unicast 1111 1110 11
multicast 1111 1111
 all other prefixes reserved (approx. 7/8ths of total)
 anycast addresses allocated from unicast prefixes
Global Unicast Addresses
001 TLA NLA* SLA* interface ID

public site interface

topology topology identifier
(45 bits) (16 bits) (64 bits)

 TLA = Top-Level Aggregator

NLA* = Next-Level Aggregator(s)
SLA* = Site-Level Aggregator(s)
 all subfields variable-length, non-self-encoding (like CIDR)
 TLAs may be assigned to providers or exchanges
Link-Local & Site-Local Unicast
Addresses
Link-local addresses for use during auto-configuration
and when no routers are present:
1111111010 0 interface ID

Site-local addresses for independence from changes

of TLA / NLA*:

1111111011 0 SLA* interface ID

Interface IDs

Lowest-order 64-bit field of unicast address

may be assigned in several different ways:
 auto-configured from a 64-bit EUI-64, or expanded
from a 48-bit MAC address (e.g., Ethernet address)
 auto-generated pseudo-random number
(to address privacy concerns)
 assigned via DHCP
 manually configured
 possibly other methods in the future
Some Special-Purpose Unicast
Addresses

 The unspecified address, used as a placeholder when

no address is available:

0:0:0:0:0:0:0:0

 The loopback address, for sending packets to self:

0:0:0:0:0:0:0:1
Multicast Address Format
FP  (8bits) Flags (4bits) Scope (4bits) RESERVED (80bits) Group ID (32bits)

11111111 000T Lcl/Sit/Gbl MUST be 0 Locally administered

 flag field
 low-order bit indicates permanent/transient group
 (three other flags reserved)
 scope field:
 1 - node local 8 - organization-local
 2 - link-local B - community-local
 5 - site-local E - global
 (all other values reserved)
 map IPv6 multicast addresses directly into low order 32
bits of the IEEE 802 MAC
Multicast Address Format
Unicast-Prefix based
Network Prefix
FP  (8bits) Flags (4bits) Scope (4bits) reserved (8bits) plen (8bits) Group ID (32bits)
(64bits)
Locally
11111111 00PT Lcl/Sit/Gbl MUST be 0 Unicast prefix Auto configured
administered

 P = 1 indicates a multicast address that is assigned based on the network

prefix
 plen indicates the actual length of the network prefix
 Source-specific multicast addresses is accomplished by setting
 P=1
 plen = 0
 network prefix = 0

 draft-ietf-ipngwg-uni-based-mcast-01.txt
Outline
 Protocol Background
 Technology Highlights
 Enhanced Capabilities
 Transition Issues
 Next Steps
Security
IPv6 Security

 All implementations required to support

authentication and encryption headers (“IPsec”)
 Authentication separate from encryption for use
in situations where encryption is prohibited or
prohibitively expensive
 Key distribution protocols are under
development (independent of IP v4/v6)
 Support for manual key configuration required
Authentication Header
Next Header Hdr Ext Len Reserved
Security Parameters Index (SPI)
Sequence Number

Authentication Data

 Destination Address + SPI identifies security

association state (key, lifetime, algorithm, etc.)
 Provides authentication and data integrity for all fields
of IPv6 packet that do not change en-route
 Default algorithm is Keyed MD5
Encapsulating Security Payload
(ESP)
Security Parameters Index (SPI)
Sequence Number

Payload

Padding
Padding Length Next Header

Authentication Data
Quality of Service
IP Quality of Service Approaches
Two basic approaches developed by IETF:
 “Integrated Service” (int-serv)
 fine-grain (per-flow), quantitative promises
(e.g., x bits per second), uses RSVP signaling
 “Differentiated Service” (diff-serv)
 coarse-grain (per-class), qualitative promises
(e.g., higher priority), no explicit signaling
IPv6 Support for Int-Serv
20-bit Flow Label field to identify specific
flows needing special QoS
 each source chooses its own Flow Label
values; routers use Source Addr + Flow
Label to identify distinct flows
 Flow Label value of 0 used when no special
QoS requested (the common case today)
 this part of IPv6 is not standardized yet, and
may well change semantics in the future
IPv6 Support for Diff-Serv
8-bit Traffic Class field to identify specific
classes of packets needing special QoS
 same as new definition of IPv4 Type-of-
Service byte
 may be initialized by source or by router
enroute; may be rewritten by routers enroute
 traffic Class value of 0 used when no special
QoS requested (the common case today)
Compromise
 Signaled diff-serv (RFC 2998)
 uses RSVP for signaling with course-grained
qualitative aggregate markings
 allows for policy control without requiring per-router
state overhead
Mobility
IPv6 Mobility
 Mobile hosts have one or more home address
 relatively stable; associated with host name in DNS
 A Host will acquire a foreign address when it discovers it
is in a foreign subnet (i.e., not its home subnet)
 uses auto-configuration to get the address
 registers the foreign address with a home agent,
i.e, a router on its home subnet
 Packets sent to the mobile’s home address(es) are
intercepted by home agent and forwarded to the foreign
address, using encapsulation
 Mobile IPv6 hosts will send binding-updates to
correspondent to remove home agent from flow
Mobile IP (v4 version)
mobile host

correspondent foreign agent

host

home agent

home location of mobile host

Mobile IP (v4 version)
mobile host

correspondent foreign agent

host

home agent

home location of mobile host

Mobile IP (v4 version)
mobile host

correspondent foreign agent

host

home agent

home location of mobile host

Mobile IP (v4 version)
mobile host

correspondent foreign agent

host

home agent

home location of mobile host

Mobile IP (v6 version)
mobile host

correspondent
host

home agent

home location of mobile host

Mobile IP (v6 version)
mobile host

correspondent
host

home agent

home location of mobile host

Mobile IP (v6 version)
mobile host

correspondent
host

home agent

home location of mobile host

Mobile IP (v6 version)
mobile host

correspondent
host

home agent

home location of mobile host

Mobile IP (v6 version)
mobile host

correspondent
host

home agent

home location of mobile host

IPv6 Routing
RIPng
 RIPv2, supports split-horizon with
poisoned reverse
 RFC2080
BGP4+ Overview
 Added IPv6 address-family
 Added IPv6 transport
 Runs within the same process - only one
AS supported
 All generic BGP functionality works as for
IPv4
 Added functionality to route-maps and
prefix-lists
IPv6 routing
 OSPF & ISIS updated for IPv6
Outline
 Protocol Background
 Technology Highlights
 Enhanced Capabilities
 Transition Issues
 Next Steps
Porting Issues
Effects on higher layers
 Changes TCP/UDP checksum “pseudo-header”
 Affects anything that reads/writes/stores/passes IP
addresses (just about every higher protocol)
 Packet lifetime no longer limited by IP layer
(it never was, anyway!)
 Bigger IP header must be taken into account when
computing max payload sizes
 New DNS record type: AAAA and (new) A6
 …
Sockets API Changes
 Name to Address Translation Functions
 Address Conversion Functions
 Address Data Structures
 Wildcard Addresses
 Constant Additions
 Core Sockets Functions
 Socket Options
 New Macros
Core Sockets Functions
 Core APIs
Use IPv6 Family and Address Structures
socket() Uses PF_INET6

 Functions that pass addresses

bind()

connect()

sendmsg()

sendto()

 Functions that return addresses

accept()

recvfrom()

recvmsg()

getpeername()

getsockname()
 Name

to Address
getaddrinfo()
Translation
 Pass in nodename and/or servicename string
 Can Be Address and/or Port
 Optional Hints for Family, Type and Protocol
 Flags – AI_PASSIVE, AI_CANNONNAME, AI_NUMERICHOST,
AI_NUMERICSERV, AI_V4MAPPED, AI_ALL, AI_ADDRCONFIG
 Pointer to Linked List of addrinfo structures Returned
 Multiple Addresses to Choose From
 freeaddrinfo()
struct
struct addrinfo
addrinfo {{
int
int ai_flags;
ai_flags;
int ai_family;
int ai_family;
int
int getaddrinfo(
getaddrinfo( int
int ai_socktype;
ai_socktype;
IN const char FAR * nodename,
IN const char FAR * nodename, int ai_protocol;
int ai_protocol;
IN
IN const
const char
char FAR
FAR ** servname,
servname, size_t
size_t ai_addrlen;
ai_addrlen;
IN
IN const
const struct
struct addrinfo
addrinfo FARFAR ** hints,
hints, char *ai_canonname;
char *ai_canonname;
OUT
OUT struct
struct addrinfo
addrinfo FARFAR ** FAR
FAR ** res
res struct
struct sockaddr
sockaddr *ai_addr;
*ai_addr;
);
); struct addrinfo *ai_next;
struct addrinfo *ai_next;
};
};
Address
 to Name Translation
getnameinfo()
 Pass in address (v4 or v6) and port
 Size Indicated by salen
 Also Size for Name and Service buffers (NI_MAXHOST, NI_MAXSERV)
 Flags
 NI_NOFQDN
 NI_NUMERICHOST
 NI_NAMEREQD
 NI_NUMERICSERV int
int getnameinfo(
getnameinfo(
IN
IN const
const struct
struct sockaddr
sockaddr FAR
FAR ** sa,
sa,
 NI_DGRAM IN socklen_t salen,
IN socklen_t salen,
OUT
OUT char
char FAR
FAR ** host,
host,
IN size_t hostlen,
IN size_t hostlen,
OUT
OUT char
char FAR
FAR ** serv,
serv,
IN size_t servlen,
IN size_t servlen,
IN
IN int
int flags
flags
);
);
Porting Environments
 Node Types
 IPv4-only
 IPv6-only
 IPv6/IPv4

 Application Types
 IPv6-unaware
 IPv6-capable
 IPv6-required

 IPv4 Mapped Addresses

Porting Issues
 Running on ANY System
 Including IPv4-only

 Address Size Issues

 New IPv6 APIs for IPv4/IPv6
 Ordering of API Calls
 User Interface Issues
 Higher Layer Protocol Changes
Specific things to look for
 Storing IP address in 4 bytes of an array.
 Use of explicit dotted decimal format in UI.
 Obsolete / New:
 AF_INET replaced by AF_INET6
 SOCKADDR_IN replaced by SOCKADDR_STORAGE
 IPPROTO_IP replaced by IPPROTO_IPV6
 IP_MULTICAST_LOOP replaced by SIO_MULTIPOINT_LOOPBACK
 gethostbyname replaced by getaddrinfo
 gethostbyaddr replaced by getnameinfo
IPv6 literal addresses in URL’s
 From RFC 2732
Literal IPv6 Address Format in URL's Syntax To use a literal IPv6 address in a
URL, the literal address should be enclosed in "[" and "]" characters. For
example the following literal IPv6 addresses:
FEDC:BA98:7654:3210:FEDC:BA98:7654:3210
3ffe:2a00:100:7031::1
::192.9.5.5
2010:836B:4179::836B:4179

would be represented as in the following example URLs:

http://[FEDC:BA98:7654:3210:FEDC:BA98:7654:3210]:80/index.html
http://[3ffe:2a00:100:7031::1]
http://[::192.9.5.5]/ipng
http://[2010:836B:4179::836B:4179]
Other Issues
 Renumbering & Mobility routinely result in changing IP
Addresses –
 Use Names and Resolve, Don’t Cache

 Multihomed Servers
 More Common with IPv6
 Try All Addresses Returned

 Using New IPv6 Functionality

Porting Steps -Summary
 Use IPv4/IPv6 Protocol/Address Family

 Fix Address Structures

in6_addr

sockaddr_in6

sockaddr_storage to allocate storage

 Fix Wildcard Address Use

in6addr_any, IN6ADDR_ANY_INIT
in6addr_loopback, IN6ADDR_LOOPBACK_INIT

 Use IPv6 Socket Options

IPPROTO_IPV6, Options as Needed

 Use getaddrinfo()
For Address Resolution
 IPv4 - IPv6
Co-Existence / Transition
IPv6 Timeline
(A pragmatic projection)

2000 2001 2002 2003 2004 2005 2006 2007

Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4

• Early adopter

• Application porting <= Duration 3+ years =>

• ISP adoption <= Duration 3+ years =>

• Consumer adoption <= Duration 5+ years =>

• Enterprise adoption <= Duration 3+ years =>

Deployments
 IPv6 deployments will occur piecewise
from the edge.
 Core infrastructure only moving when
significant customer usage demands it.

 Platforms and products that are updated first

need to address the lack of ubiquity.
Whenever possible, devices and applications
should be capable of both IPv4 & IPv6, to
minimize the delays and potential failures
inherent in translation points.
Impediments to IPv6 deployment
 Applications
 Applications
 Applications

 Move to the new APIs NOW

Transition / Co-Existence Techniques
A wide range of techniques have been identified and implemented,
basically falling into three categories:
 (1) dual-stack techniques, to allow IPv4 and IPv6 to
co-exist in the same devices and networks
 (2) tunneling techniques, to avoid order
dependencies when upgrading hosts, routers, or
regions
 (3) translation techniques, to allow IPv6-only
devices to communicate with IPv4-only devices
Expect all of these to be used, in combination
Dual-Stack Approach
 When adding IPv6 to a system, do not delete IPv4
 this multi-protocol approach is familiar and
well-understood (e.g., for AppleTalk, IPX, etc.)
 note: in most cases, IPv6 will be bundled with
new OS releases, not an extra-cost add-on
 Applications (or libraries) choose IP version to use
 when initiating, based on DNS response:
Prefer scope match first, when equal IPv6 over IPv4
 when responding, based on version of initiating packet
 This allows indefinite co-existence of IPv4 and IPv6, and gradual
app-by-app upgrades to IPv6 usage
Tunnels to Get Through
IPv6-Ignorant Routers
 Encapsulate IPv6 packets inside IPv4 packets
(or MPLS frames)
 Many methods exist for establishing tunnels:
 manual configuration
 “tunnel brokers” (using web-based service to create a tunnel)
 automatic (depricated, using IPv4 as low 32bits of IPv6)
 “6-over-4” (intra-domain, using IPv4 multicast as virtual LAN)
 “6-to-4” (inter-domain, using IPv4 addr as IPv6 site prefix)
 Can view this as:
 IPv6 using IPv4 as a virtual NBMA link-layer, or
 an IPv6 VPN (virtual public network), over the IPv4 Internet
Translation
 May prefer to use IPv6-IPv4 protocol translation for:
 new kinds of Internet devices (e.g., cell phones, cars,
appliances)
 benefits of shedding IPv4 stack (e.g., serverless
autoconfig)
 This is a simple extension to NAT techniques, to
translate header format as well as addresses
 IPv6 nodes behind a translator get full IPv6 functionality
when talking to other IPv6 nodes located anywhere
 they get the normal (i.e., degraded) NAT functionality
when talking to IPv4 devices
 drawback : minimal gain over IPv4/IPv4 NAT approach
Tunnels

 6to4
 Configured
 Automatic
6to4 tunnels
FP  (3bits) TLA  (13bits) IPv4 Address  (32bits) SLA ID  (16bits) Interface ID (64bits)

001 0x0002 ISP assigned Locally administered Auto configured

2002:8243:1::/48

2002:947A:1::/48

IPv6 IPv4 IPv6

130.67.0.1 148.122.0.1

11.0.0.1
6to4 prefix is 2002::/16 + IPv4 address.
2002:a.b.c.d::/48 IPv6 Internet
6to4 relay
2002:B00:1::1
Announces 2002::/16 to the IPv6 Internet
6to4 tunnels II
Pros Cons
Minimal configuration All issues that NMBA
networks have.
Only site border router Requires relay router to
needs to know about 6to4 reach native IPv6 Internet

Works without adjacent Has to use 6to4

native IPv6 routers addresses, not native.

NB: there is a draft describing how to use IPv4 anycast to reach the relay router.
(This is already supported, by our implementation...)
 Configured tunnels
3ffe:c00:2::/48
3ffe:c00:1::/48

IPv4
IPv6 IPv6
130.67.0.1 148.122.0.1

--------------------------------------
|IPv4 header|IPv6 header IPv6 payload|
--------------------------------------
IPv4 protocol type = 41
Configured tunnels II

Pros Cons
As point to point links Has to be configured and
managed
Multicast Inefficient traffic patterns

Real addresses No keepalive mechanism,

interface is always up
Automatic tunnels
0 IPv4 Address  (32bits)

Defined ISP assigned

148.122.0.1
130.67.0.1
::148.122.0.1
::130.67.0.1

IPv6 IPv4 IPv6

Connects dual stacked nodes

Quite obsolete
IPv6 Internet
Automatic tunnels II
Pros Cons
Obsolete Difficult to reach the native
IPv6 Internet, without
injecting IPv4 routing
information in the IPv6
routing table
Useful for some other Has to use IPv4
mechanisms, like BGP compatible addresses
tunnels
Tunneling issues
 IPv4 fragmentation needs to be reconstructed at
tunnel endpoint.
 No translation of Path MTU messages between
IPv4 & IPv6.
 Translating IPv4 ICMP messages and pass back
to IPv6 originator.
 May result in an inefficient topology.
Tunneling issues II
 Tunnel interface is always up. Use routing
protocol to determine link failures.
 Be careful with using the same IPv4 source
address for several tunneling mechanisms.
Demultiplexing incoming packets is difficult.
Deployment scenarios
 Many ways to deliver IPv6 services to End Users
 Most important is End to End IPv6 traffic forwarding
 Service Providers and Enterprises may have
different deployment needs
 IPv6 over IPv4 tunnels
 Dedicated Data Link layers for native IPv6
 no impact on IPv4 traffic & revenues
 Dual stack Networks
 IPv6 over MPLS or IPv4-IPv6 Dual Stack Routers
Media - Interface Identifier
 IEEE interfaces - EUI-64
 MAC-address: 0050.a218.0c38

 Interface ID: 250:A2FF:FE18:C38

 P2P links (HDLC, PPP)

 Interface ID: 50:A218:C00:D

 48 bits from the first MAC address in the box + 16

bit interface index. U/L bit off
 IPv4 tunnels
 Interface ID: ::a.b.c.d
Outline
 Protocol Background
 Technology Highlights
 Enhanced Capabilities
 Transition Issues
 Next Steps
Current Status
Standards
 core IPv6 specifications are IETF Draft
Standards
=> well-tested & stable
 IPv6 base spec, ICMPv6, Neighbor Discovery, PMTU
Discovery, IPv6-over-Ethernet, IPv6-over-PPP,...
 other important specs are further behind on the
standards track, but in good shape
 mobile IPv6, header compression, A6 DNS support,...
 for up-to-date status: playground.sun.com/ipng
 UMTS R5 cellular wireless standards mandate
IPv6
Implementations
 Most IP stack vendors have an implementation
at some stage of completeness
 some are shipping supported product today,
e.g., 3Com, *BSD(KAME), Cisco, Epilogue,
Ericsson/Telebit, IBM, Hitachi, NEC, Nortel, Sun, Trumpet
 others have beta releases now, supported products soon,
e.g., Compaq, HP, Linux community, Microsoft
 others rumored to be implementing, but status unkown (to
me), e.g., Apple, Bull, Juniper, Mentat, Novell, SGI
 (see playground.sun.com/ipng for most recent status
reports)
 Good attendance at frequent testing events
IPv6 Addresses
Bootstrap phase
 Where to get address space?
 Real IPv6 address space now allocated
by APNIC, ARIN and RIPE NCC
 APNIC 2001:0200::/23
 ARIN 2001:0400::/23
 RIPE NCC 2001:0600::/23
 6Bone 3FFE::/16
Have a look at www.cisco.com/ipv6 for further information
 IPv6 Address Space
Current Allocations
 APNIC (whois.apnic.net) SONYTELECOM-JPNIC-JP-20001207 2001:298::/35
CONNECT-AU-19990916 2001:210::/35 TTNET-JPNIC-JP-20001208 2001:2A0::/35
WIDE-JP-19990813 2001:200::/35
CCCN-JPNIC-JP-20001228 2001:02A8::/35
IMNET-JPNIC-JP-20000314 2001:0248::/35
NUS-SG-19990827 2001:208::/35
KIX-KR-19991006 2001:220::/35 KORNET-KRNIC-KR-20010102 2001:02B0::/35
ETRI-KRNIC-KR-19991124 2001:230::/35  ARIN (whois.arin.net)
NTT-JP-19990922 2001:218::/35
ESNET-V6 2001:0400::/35
HINET-TW-20000208 2001:238::/35
ARIN-001 2001:0400::/23
IIJ-JPNIC-JP-20000308 2001:240::/35
VBNS-IPV6 2001:0408::/35
CERNET-CN-20000426 2001:250::/35
INFOWEB-JPNIC-JP-2000502 2001:258::/35
CANET3-IPV6 2001:0410::/35
JENS-JP-19991027 2001:228::/35 VRIO-IPV6-0 2001:0418::/35
BIGLOBE-JPNIC-JP-20000719 2001:260::/35 CISCO-IPV6-1 2001:0420::/35
6DION-JPNIC-JP-20000829 2001:268::/35 QWEST-IPV6-1 2001:0428::/35
DACOM-BORANET-20000908 2001:270::/35 DEFENSENET 2001:0430::/35
ODN-JPNIC-JP-20000915 2001:278::/35 ABOVENET-IPV6 2001:0438::/35
KOLNET-KRNIC-KR-20000927 2001:280::/35 SPRINT-V6 2001:0440::/35
HANANET-KRNIC-KR-20001030 2001:290::/35
UNAM-IPV6 2001:0448::/35
TANET-TWNIC-TW-20001006 2001:288::/35
GBLX-V6 2001:0450::/35

January 5th, 2001

 IPv6 Address Space
Current Allocations
 RIPE (whois.ripe.net) EU-EUNET-20000403 2001:0670::/35

UK-BT-19990903 2001:0618::/35 DE-IPF-20000426 2001:0678::/35

CH-SWITCH-19990903 2001:0620::/35 DE-NACAMAR-20000403 2001:0668::/35

DE-XLINK-20000510 2001:0680::/35
AT-ACONET-19990920 2001:0628::/35
DE-ECRC-19991223 2001:0650::/35
UK-JANET-19991019 2001:0630::/35
FR-TELECOM-20000623 2001:0688::/35
DE-DFN-19991102 2001:0638::/35
PT-RCCN-20000623 2001:0690::/35
NL-SURFNET-19990819 2001:0610::/35
SE-SWIPNET-20000828 2001:0698::/35
RU-FREENET-19991115 2001:0640::/35
PL-ICM-20000905 2001:06A0::/35
GR-GRNET-19991208 2001:0648::/35
DE-SPACE-19990812 2001:0608::/35
EU-UUNET-19990810 2001:0600::/35 BE-BELNET-20001101 2001:06A8::/35
DE-TRMD-20000317 2001:0658::/35 SE-SUNET-20001218 2001:06B0::/35
FR-RENATER-20000321 2001:0660::/35 IT-CSELT-20001221 2001:06B8::/35

SE-TELIANET-20010102 2001:06C0::/35
Deployment
 experimental infrastructure: the 6bone
 for testing and debugging IPv6 protocols and
operations (see www.6bone.net)
 production infrastructure in support of education
and research: the 6ren
 CAIRN, Canarie, CERNET, Chunahwa Telecom,
Dante, ESnet, Internet 2, IPFNET, NTT, Renater,
Singren, Sprint, SURFnet, vBNS, WIDE
(see www.6ren.net, www.6tap.net)
 commercial infrastructure
 a few ISPs (IIJ, NTT, SURFnet, Trumpet,…) have
announced commercial IPv6 service or service trials
Deployment (cont.)
 IPv6 address allocation
 6bone procedure for test address space
 regional IP address registries (APNIC, ARIN,
RIPE-NCC) for production address space
 deployment advocacy (a.k.a. marketing)
 IPv6 Forum: www.ipv6forum.com
Much Still To Do
though IPv6 today has all the functional capability of IPv4,
 implementations are not as advanced
(e.g., with respect to performance, multicast support,
compactness, instrumentation, etc.)
 deployment has only just begun
 much work to be done moving application, middleware,
and management software to IPv6
 much training work to be done
(application developers, network administrators, sales
staff,…)
 many of the advanced features of IPv6 still need
specification, implementation, and deployment work
Recent IPv6 “Hot Topics” in the
IETF
 multihoming / address  enhanced router-to-host info
selection  site renumbering procedures
 address allocation  temp. addresses for privacy
 DNS discovery
 inter-domain multicast routing
 address propagation and AAA
 3GPP usage of IPv6 issues of different access
 anycast addressing scenarios
 scoped address architecture  (always-on, dial-up, mobile,…)
 and, of course, transition /
 flow-label semantics co-existence / interoperability
 API issues with IPv4
 (flow label, traffic class, PMTU
discovery, scoping,…)

Note: this indicates vitality, not incompleteness, of IPv6!

Next Steps
For More Information

 http://www.ietf.org/html.charters/ipngwg-
charter.html
 http://www.ietf.org/html.charters/ngtrans-
charter.html
 http://playground.sun.com/ipv6/
 http://www.6bone.net/ngtrans/
For More Information

 http://www.6bone.net
 http://www.ipv6forum.com
 http://www.ipv6.org
 http://www.cisco.com/ipv6/
 http://www.microsoft.com/windows2000/librar
y/howitworks/communications/networkbasics/
IPv6.asp
For More Information

 BGP4+ References
 RFC2858 Multiprotocol extension to BGP
RFC2545 BGP MP for IPv6
RFC2842 Capability negotiation
 RIPng RFC2080
Other Sources of Information
 Books
 IPv6, The New Internet Protocol
by Christian Huitema (Prentice Hall)
 Internetworking IPv6 with Cisco Routers
by Silvano Gai (McGraw-Hill)

 and many more... (14 hits at Amazon.com)

 Questions?

2213
1313_06_2000_c2 © 2000, Cisco Systems, Inc. 119
Cisco Systems