Overview of IPv6 DNS configuration Routing protocols Transition strategies Router configurations Host installation and configuration How to connect to the IPv6 IPv6 deployment on the Internet IPv6 industry support and trends
Viag谷nie,March 2000 2
Why IPv6 ?
Problems with IPv4 IPv4 has been designed early in the 70s Many add-ons to the protocol : Mobileip QoS Security (IPsec)
Others Using one add-ons -> easy Using two at the same time -> difficult Using three or more -> acrobatic !!!!
Viag谷nie,March 2000 3
Why IPv6 ?
Problems with IPv4 During the 80s, addresses delegation without optimisation and without aggregation Possible solution : IP renumbering and unused address space redistribution Consequences :Large routing table on the backbone Unthinkable for some sites
Viag谷nie,March 2000 4
Why IPv6 ?
IPv4 address shortage (current situation)
Fact #1 : Few consequence in North America Internet heaven !
Fact #2 : Major problem for every other countries around the world China requested addresses to connect 60 000 schools and got one class B Several countries in Europe, Africa and Asia are using one class C for a whole country
Viag谷nie,March 2000 5
Why IPv6 ?
IPv4 address shortage (current situation)
Some ISP in these countries are providing private addresses to their clients (Suedish ISP using NAT)
Internet users move from PPP connectivity to xDSL/cable modem ( ratio users by IP address is changing from 10:1 to 1:1)
ISP are delegating only few address space to their corporate client s Temporary solution ——> NAT (but unfortunatly permanent)
Viag谷nie,March 2000 6
Why IPv6 ?
IPv4 address shortage in the future Internet growth in some regions : Asia (2.5 billions people)
Eastern Europe (250 millions)
Africa (800 millions)
South and Central America (500 millions)
Growth of the applications that need IP addresses globally scoped, unique and routable (VOIP, videoconferencing, games)
Viag谷nie,March 2000 7
Why IPv6 ?
NAT hinders Internet applications deployment Unidirectionnal concept (from Intranets to Internet)
How to reach a VOIP application with a private address ? -> Impossible !
VoIP Application 192.168.3.x Segment A VOIP Application 192.168.3.100
ISP/Internet 192.168.2.x Segment B
205.123.41.10
192.168.1.x Segment C Router NAT support
Viag谷nie,March 2000 8
Why IPv6 ?
NAT hinders Internet applications deployment Comunication, security and game applications need bidirectionnel support VOIP (RTP/RTCP)
Videoconferencing (RTP/RTCP)
IPsec Network game (Quake multiplayer)
RFC 2775 about Internet Transparency by Brian Carpenter
Viag谷nie,March 2000 9
Home gaming IPv6 setup
QUAKE server(IPv6) Quake IPv6 client
ISP/Internet local subnet IPv6 backbone (IPv4) ROUTER Quake IPv4 /w NAT IPv6 over IPv4 tunnel
Quake IPv6 client
Viag谷nie,March 2000 10
Why IPv6 ?
NAT hinders Internet applications deployment Several protocols don *t pass throught NAT IPsec -> NAT changes address in the packet header -> lost of integrity Kerboros -> NAT changes address in the packet header -> K needs the source address RTP/RTCP -> use UDP with dynamic ports assignation -> NAT is not able to support this translation during a session (except proxy)
Multicast is not easy to set-up !!!
Viag谷nie,March 2000 11
Why IPv6 ?
Communications technologies need permanent addresses to get connected to the Internet Cellulars (500 millions )
Standard phones (900 millions)
Radio/TV (++ hundred millions)
Industrials devices (billions of IP addresses)
Any electronics device (walkman to download MP3 files, bulgar alarm to send e-mail to the police station #)
Viag谷nie,March 2000 12
Why IPv6 ?
CONCLUSION :The true question is not :Do we need and do we believe in IPv6 ?
Not, the right one is :Are we interested in a network that allows any IP electronic devices to communicate transparently to each other regarless its location on THE global net ?
- Viag谷nie
Viag谷nie,March 2000 13
IPv6 Features
Larger Address Space Aggregation-based address hierarchy Efficient backbone routing Efficient and Extensible IP datagram No fragmentation by routers 64 bits field alignement Simpler basic header Autoconfiguration Security IP Renumbering part of the protocol
Viag谷nie,March 2000 14
History
TUBA (1992)
TCP and UDP over Bigger Addresses Uses ISO CLNP (Connection-Less Network Protocol)
Dropped SIPP (1993)
Simple IP Plus Merge of Sip and Pip 64 bits addresses IPng adopted SIPP in 1994 Changed address size to 128 bits Changed to IPv6
Viag谷nie,March 2000 15
Design criterias for IPv6
Number of addresses Efficiency in routers low and very high bandwidth (100G/bytes++)
Security Mobility Autoconfig Seamless transition Don*t require a day X for switching to IPv6 No need to change hardware
Viag谷nie,March 2000 16
Basic specifications
IPv4 packet description (20 bytes + options)
Ver. header TOS total length identification flag fragment offset TTL Protocol Checksum 32 bit Source Address 32 bit Destination Address
removed changed
Viag谷nie,March 2000 17
Basic specifications
RFC2460 IPv6 packet description (40 bytes)
Ver. TrafficClass Flow Label Payload Length Next Header Hop Limit
128 bit Source Address
128 bit Destination Address
Viag谷nie,March 2000 18
Basic specifications
Version (4 bits)
6 for IPv6 Traffic Class (8 bits)
~= TOS in IPv4 Identifies and distinguishes between different classes or priorities (diffserv)
Flow Label (20 bits)
Experimental Used by a source node to label sequences of packets Payload Length ~= Total length in IPv4
Viag谷nie,March 2000 19
Basic specifications
Next Header (8 bits)
Used for extension headers ~= Protocol field in IPv4 Most not processed by routers in the path Hop-by-hop options (0)
information that must be examined by every node along the path Routing (43)
similar to IPv4's Loose Source and Record Route option Fragment (44)
used by source node (routers don*t fragment anymore !)
Viag谷nie,March 2000 20
Basic specifications
Next Header (8 bits) cont. Destination options (60)
used to carry optional information that need to be examined only by a packet's destination node(s)
Authentication (IPsec)
ESP (IPsec)
Hop Limit ~= TTL in IPv4 MTU must be at least 1280 bytes (1500+ recommended)。 Nodes should use Path MTU discovery. UDP checksum required
Viag谷nie,March 2000 21
IPv6 addresses
128 bits = 3,40 E 38 addresses Imagine Bill Gates* fortune is 85 billions $(8.5 E 10)
Take 1 trillion Bill Gateses Convert their fortune to pennies Assign 1 E 12 addresses to each pennies takes 8.5 E 36 addresses You*ve just assigned 2.5% of the entire IPv6 address space http://www.cnn.com/TECH/computing/9909/21/ip.crunch.idg/index.html
Viag谷nie,March 2000 22
IPv6 addressing
RFC2373 IP Version 6 Addressing Architecture
Reserved 0000 0000 1/256
Reserved for NSAP Allocation 0000 001 1/128 Reserved for IPX Allocation 0000 010 1/128
Aggregatable Global Unicast Addresses 001 1/8
Link-Local Unicast Addresses 1111 1110 10 1/1024 Site-Local Unicast Addresses 1111 1110 11 1/1024 Multicast Addresses 1111 1111 1/256
Total of about 15 % of address space reserved,but not necessarily assigned or allocated
Viag谷nie,March 2000 23
IPv6 address representation Format is x:x:x:x:x:x:x:x x is a 16 bit hexadecimal field FEDC:BA98:7654:3210:FEDC:BA98:7654:3210 Leading zeros in a field are optional:: can be used to represent multiple groups of 16 bits of zero :: can only be used once in an address FF01:0:0:0:0:0:0:101 = FF01::101 0:0:0:0:0:0:0:1 = ::1 0:0:0:0:0:0:0:0 = ::
Viag谷nie,March 2000 24
IPv6 address representation
RFC2732: Preferred Format for Literal IPv6 Addresses in URL
http://[1080::8:800:200C:417A]:80/index.html
Viag谷nie,March 2000 25
IPv6 addressing
Unicast address FE80::/10 Link-Local Unicast Address scope limited to local network automatically configured on all nodes using interface identifiers FE80::<interface id> used for neighbor discovery and router discovery. can also be used as a non-globally-routed IPv6 local network
Viag谷nie,March 2000 26
IPv6 addressing
Unicast address FEC0::/10 Site-Local Unicast Address confined to local site or organization configured using interface identifier and a pre- defined 16 bits subnet ID FEC0::<subnet id>:<interface id> what is a site??? (few drafts: draft-haberman-IPv6- site-route-00.txt, draft-ietf-ipngwg-site-prefixes- 02.txt )
Viag谷nie,March 2000 27
Aggregatable Global Unicast Addresses
RFC2374 Aggregatable Global Unicast Addresses 2000::/3
TLA: Top Level Aggregator Primary providers (default free)
NLA: Next Level Aggregator Can have multiple NLA as sub-NLA SLA: Site Level Aggregator Your site (16 bits)
Viag谷nie,March 2000 28
Aggregatable Global Unicast Addresses
3FFE:0B00:0C18:0001:0290:27FF:FE17:FC0F TLA NLA(s) SLA Interface ID
16 bits 32 bits 16 bits 64 bits
Addresses are allocated from your provider If you change provider, your prefix changes But renumbering (of hosts, routers and sites)
has been included in the IPv6 protocol
Viag谷nie,March 2000 29
IPv6 Addressing
::1 Loopback address (like 127.0.0.1 in IPv4)
:: Unspecified address::<IPv4 address> IPv4 compatible address Auto-tunnels (IPv6 over IPv4)
::FFFF:<IPv4 address> IPv4 mapped address (used by resolver library)
IPv6 representation of an IPv4 node 206.123.31.101 is mapped as ::FFFF:206.123.31.101
[1] [2] [3] [4] 下一页