ipv6 @ cisco

1、IPv6 @ Cisco Systems,Patrick GrosseteteCisco IOS® IPv6 Product Managerpgrosset@cisco.com,Agenda,IPv6 Rationales IPng Protocols & Standards IPv6 Integration & Co-Existence Cisco IOS IPv6 roadmap IPv6 De

2、ployment,A Need for IPv6?,IETF IPv6 WG began in early 90s to solve addressing growth issues butCIDR, NAT,… were developedIPv4 32 bits address = 4 billion hosts~40% of the IPv4 address space is still unused… BUTIP is

3、everywhereData, Voice, Audio and Video integration is a RealityRegional Registries apply a strict allocation controlAddressing scheme is not optimum as for anySo, let’s play with Numbers and focus on Applications,Why

4、 a Larger Address Space is Needed,Overall Internet traffic is still growing at 400%/year worldwide~320 million users in 2000, ~550 million by 2005Emerging population/geopolitical & Address spaceChina, India, Japa

5、n, Korea needs global IP addressesHow to move to e-Economy without Global Internet access?405 million mobile phones sold in 2000, over 1 billion by 2005UMTS Release 5 is Internet Mobility, 1/3 of 1B should get connect

6、ed~1 Billion cars in 2010, 15% should get GPS and Yellow Page servicesBillion of new Internet appliances for Home users and industrial environments will always be ON,Explosion of New Internet Appliances,,Coming Back to

7、 an End-to-End Architecture,GlobalAddressingRealm,,,,,,New Technologies/Applications for Home Users‘Always-on’—Cable, DSL, Ethernet@home, Wireless,…,Internet started with End to End connectivity for any application

8、s Today, NAT and Application-Layer Gateways connecting disparate networksAlways-on Devices Need an Address When You Call Them, eg.- Mobile Phones- Gaming- Residential Voiceover IP gateway,IPv6 Markets,Academic

9、NRNInternet-II (Abilene, vBNS+), Canarie*3, Renater-II, Surfnet, DFN, Ukerna, CERNET,… 6REN/6TAPGeographies & PoliticsPrime Minister of Japan called for IPv6, Korea adopts similar positionEEC e-Europe document ad

10、vertised IPv6 as the way to go for EuropeWireless (PDA, 3G Mobile Phone networks, Car,...)Multiple phases before deploymentRFP -> Integration -> trial -> commercial Requires ‘client devices’, eg. IPv6 handse

11、t ?,IPv6 Markets,Home NetworkingSet-top box/Cable/xDSL/Ether@HomeResidential Voice over IP gatewayGaming (10B$ market)Sony, (Sega), Nintendo, Microsoft Consumer PCEnterpriseRequires IPv6 support by O.S. & Appl

12、icationsSUN Solaris 8, BSD 4.x, Linux, Microsoft Windows XP Pro,...Service ProvidersRegional ISP, Carriers, Mobile ISP, and Greenfield ISP’s,IPv6 AddressesBootstrap Phase,Where to get address space?Real IPv6 address

13、 space now allocated by APNIC, ARIN and RIPE NCCAPNIC2001:0200::/23ARIN2001:0400::/23RIPE NCC2001:0600::/236Bone3FFE::/16Have a look at www.cisco.com/ipv6 for further information,IPv6 Address SpaceCurrent

14、Allocations,May 25th, 2001,IPv6 Address SpaceCurrent Allocations,May 25th, 2001,IPv6 Address SpaceCurrent Allocations,March 26th, 2001,The Benefits ofIntegrating IPv6,Large Address Space,“Plug-and-Play”,Enhanced Mobil

15、ity,The UbiquitousInternet,Agenda,IPv6 Rationales IPv6 Protocols & Standards IPv6 Integration & Co-Existence Cisco IOS IPv6 roadmap IPv6 Deployment,IPv6 – So What’s Really Changed?,Defined by RFC 2460Addres

16、s length quadrupled to 16 bytesFixed length(Optional headers daisy chained)No checksumming (Done by Link Layer)No hop-by-hop segmentation (Path MTU discovery)Flow label/class(Integrated QoS support)Concatenated

17、extension headers…,IPv4 Header,IPv6 Header,IPv6 Main Features/Functionality,Expanded Address Space Header Format Simplification Auto-configuration and Multi-Homing Mobile IP without triangular routing Class of Servic

18、e/Multimedia support Authentication and Privacy Capabilities No more broadcast Multicast IPv4 IPv6 Transition Strategy,,,IPv6 Addressing,IPv6 Addressing rules are covered by multiples RFC’sArchitecture de

19、fined by RFC 2373Address Types are :Unicast : One to One (Global, Link local, Site local, Compatible)Anycast : One to Nearest (Allocated from Unicast)Multicast : One to ManyReservedA single interface may be assigne

20、d multiple IPv6 addresses of any type (unicast, anycast, multicast)No Broadcast Address -> Use Multicast,IPv6 Addressing,Prefix Format (PF) AllocationPF = 0000 0000 : ReservedPF = 0000 001 : Reserved for OSI NSAP A

21、llocation (see RFC 1888)PF = 0000 010 : Was reserved for IPX Allocation (no use)PF = 001 : Aggregatable Global Unicast AddressPF = 1111 1110 10 : Link Local Use AddressesPF = 1111 1110 11 : Site Local Use AddressesP

22、F = 1111 1111 : Multicast AddressesOther values are currently Unassigned (approx. 7/8th of total)All Prefix Formats have to have EUI-64 bits Interface IDBut Multicast,Text Representation of IPv6 Addresses,“preferred”

23、form:1080:0:FF:0:8:800:200C:417Acompressed form:FF01:0:0:0:0:0:0:43becomes FF01::43IPv4-compatible:0:0:0:0:0:0:13.1.68.3or ::13.1.68.3 RFC 2732: Preferred format for literal IPv6 address in URL,Global

24、Unicast Addresses (RFC 2374),Aggregatable Global Unicast Format - RFC2374Address hierarchy matches Internet Service Provider hierarchy Terminology:FP - Format Prefix: Unicast (001), Multicast, AnycastTLA - Top Level

25、 Aggregator ? Global ISPNLA - Next Level Aggregator ? ISPSLA - Site Level Aggregator ? “Customer”Interface ID - Host,Link-local addresses for use during auto-configuration and when no routers are present: Site-loca

26、l addresses for independence from changes of TLA / NLA*:,Link-Local & Site-Local Unicast Addresses,Multicast Addresses (RFC 2375),low-order flag indicates permanent / transient group; three other flags reservedscope

27、 field:1 - node local2 - link-local5 - site-local8 - organization-localB - community-localE - global(all other values reserved),4,112 bits,8,group ID,scope,flags,,,,11111111,,,,4,,IPv6 Addressing Examples,Global u

28、nicast address(es) is :2001:420:101:1::E0:F726:4E58, subnet is 2001:420:101:1::0/64Link-local address is FE80::E0:F726:4E58Unspecified Address is 0:0:0:0:0:0:0:0 or ::Loopback Address is 0:0:0:0:0:0:0:1 or ::1Group

29、 Addresses (Multicast), ie: FF02::9 for RIPv6Joined group address(es):FF02:0:0:0:0:1:FF:xxxx (solicited Node Multicast)Unicast : 4037::01:800:200E:8C6C is FF02::1:FF0E:8C6C,More on IPv6 Addressing,Bootstrap process -

30、RFC2450Definitions:TLA - special TLA 0x0001subTLA - Top Level Aggregator ? Transit ISPNLA - Next Level Aggregator ? ISPSLA - Site Level Aggregator ? “Customer”Interface ID - Host,IPv6 AddressesBootstrap Phase,IP

31、v6 AddressesBootstrap Phase,Minimum assignment to ISP is a /35ISP creates own NLA boundary - or -ISP assigns /48 SLAs to each customer16 bits for subnetworks65536 subnetworks per site64 bits for hosts18446744073

32、710 million hosts per subnetwork!!,IPv6 AddressesBootstrap phase,subTLA holder ISP allocates SLAs to end-customerssubTLA holder ISP creates its own NLA boundary for customer ISPs,IPv6 Header Options (RFC 2460),Proces

33、sed only by node identified in IPv6 Destination Address field => much loweroverhead than IPv4 optionsexception: Hop-by-Hop Options header Eliminated IPv4’s 40-octet limit on optionsin IPv6, limit is total packet s

34、ize, or Path MTU in some cases,IPv6 Header Options (RFC2460),Currently defined Headers should appear in the following order :IPv6 headerHop-by-Hop Options headerDestination Options headerRouting headerFragment heade

35、rAuthentication header (RFC 1826)Encapsulating Security Payload header (RFC 1827)Destination Options headerupper-layer header,MTU Issues,Minimum link MTU for IPv6 is 1280 octets(versus 68 octets for IPv4)=> on l

36、inks with MTU < 1280, link-specificfragmentation and reassembly must be usedImplementations are expected to perform path MTU discovery to send packets bigger than 1280Minimal implementation can omit PMTU discovery

37、as long as all packets kept ≤ 1280 octetsA Hop-by-Hop Option supports transmission of “jumbograms” with up to 232 octets of payload,Fragment Header,Though discouraged, can use IPv6 Fragment header to support upper layer

38、s that do not (yet) do path MTU discoveryIPv6 frag. & reas. is an end-to-end function; routers do not fragment packets en-route if too big—they send ICMP “packet too big” instead,Next Header,Original Packet Identifi

39、er,Next Header,Fragment Offset,0 0 M,Neighbour Discovery (RFC 2461),Protocol built on top of ICMPv6 (RFC 2463) combination of IPv4 protocols (ARP, ICMP,…)Fully dynamic, interactive between Hosts & Routers defines

40、5 ICMPv6 packet typesRouter Solicitation / Router AdvertisementsNeighbour Solicitation / Neighbour AdvertisementsRedirect,Neighbour Discovery (RFC 2461),Defined mechanisms between nodes attached on the same linkRoute

41、r discoveryPrefix discoveryParameters discovery, ie: link MTU, hop limit,…Address autoconfigurationAddress Resolution (same function as ARP)Next-hop determinationNeighbor Unreachability Detection (useful for defaul

42、t routers)Duplicate Address DetectionRedirect,,(Single Subnet Scope, Formed fromReserved Prefix and Link Layer Address),SUBNET PREFIX,IPv6 Auto-Configuration,Stateless (RFC2462)Host autonomously configures its own

43、addressLink local addressingi.e.: FE80::E0:F726:4E58StatefulDHCPv6Addressing lifetimeFacilitates graceful renumberingAddresses defined as valid, deprecated or invalid,,,,,,,SUBNET PREFIX + MAC ADDRESS,,SUBNET PREF

44、IX + MAC ADDRESS,,SUBNET PREFIX + MAC ADDRESS,,SUBNET PREFIX + MAC ADDRESS,Serverless Autoconfiguration(“Plug-n-Play”),IPv6 Hosts can construct their own addresses:subnet prefix(es) learned from periodic multicast adve

45、rtisements from neighboring router(s)interface IDs generated locally, e.g., using MAC addressesOther IP-layer parameters also learned from router adverts (e.g., router addresses, recommended hop limit, etc.)Higher-la

46、yer info (e.g., DNS server and NTP server addresses) discovered by multicast / anycast-based service-location protocol [details still to be decided],Auto-Reconfiguration(“Renumbering”),New address prefixes can be intro

47、duced,and old ones withdrawnwe assume some overlap period between old and new,i.e., no “flash cut-over”hosts learn prefix lifetimes and preferability from router advertisementsold TCP connections can survive until e

48、nd of overlap;new TCP connections can survive beyond overlapRouter renumbering protocol, to allow domain-interior routers to learn of prefix introduction / withdrawalNew DNS structure to facilitate prefix changes,Uses

49、 same “l(fā)ongest-prefix match” routing as IPv4 CIDR Key to scalable routing—hierarchical addressingAssignment of production IPv6 Sub-TLA address prefixes obtainable from Registries (RIPE-NCC, APNIC, ARIN) since 1999Exis

50、ting routing protocols require extensions for IPv6Neighbor discovery—dynamic host routerCan use Routing header with anycast addresses to route packets through particular regionse.g., for provider selection, policy, p

51、erformance, etc.,IPv6 Routing,IPv6 Routing Protocols,Update to existing IPv4 routing protocols to handle bigger addressesRIPv6 (RFC 2080) - Similar to RIPv2BGP4+ - Multi-Protocols Extensions defined in RFC 2283, 2545

52、Integrated IS-IS - Large Address support facilitates IPv6 addressfamily. Draft-ietf-isis-ipv6-01.OSPFv3 (RFC 2740) New protocol implementation for IPv6IPv6 Multicast RoutingPIM, MOSPF, MBGP have IPv6 extensionsIPv

53、6 Multicast has larger address space removing potential IP addresses collision,Mobile IPv6 Terminology,,,,,,Internet,MN,HA,Access Router,,Home Agent (HA)Mobile Node (MN)Care of Address (COA)Correspondent Node (CN

54、),Overview of Mobile IPv6 Functionality,1. MN obtains IP address using stateless or stateful autoconfiguration2. MN registers with HA3. HA tunnels packets from CN to MN4. MN sends packets from CN directly or via

55、tunnel to HA,,,HA,,1.,,2.,MN,CN,,4.,,3.,,,Mobile IPv6,Plenty of addresses2 for Mobile Hosts>> 2 for Mobile NetworksNo dependency on specialized access network functionality (Foreign Agent, DHCP)NAT/PT/ALGs need

56、ed during migration from IPv4 to IPv6 but decrease with time,What does it do for:,SecurityNothing IP4 doesn’t do - IPSec runs in bothbut IPv6 mandates IPSecQoSNothing IP4 doesn’t do - Differentiated and Integrated S

57、ervices run in bothSo far, Flow label has no real use,IPv6 Technology Scope,IPv6 Standards Status,IPv6 documents are at various points in the standards process, core documents are doneDocument review for completeness,

58、followed by issues or additional work.To know more about IPv6 specificationswww.ietf.org/html.charters/ipngwg-charter.htmlMain covered areas are :Architecture, Addressing, Routing, Security, Transition, DNS, Manageme

59、nt, Discovery & Auto-Configuration, Mobility, Multicast, Applications API, ...,IPv6 Current Status - Standardisation,Several key components now on Standards Track: Specification (RFC2460)Neighbour Discover

60、y (RFC2461)ICMPv6 (RFC2463)IPv6 Addresses (RFC2373/4/5)RIP (RFC2080)BGP (RFC2545) IGMPv6 (RFC2710)OSPF (RFC2740)Router Alert (RFC2711)Jumbograms (RFC2675)Autoconfiguration (RFC2462)IPv6 over:PPP

61、(RFC2023)Ethernet (RFC2464)FDDI (RFC2467)Token Ring (RFC2470)NBMA(RFC2491)ATM (RFC2492)Frame Relay (RFC2590)ARCnet (RFC2549),Agenda,IPv6 Rationales IPng Protocols & Standards IPv6 Integration &am

62、p; Co-Existence Cisco IOS IPv6 roadmap IPv6 Deployment,IPv4-IPv6 Co-Existence / Transition,A wide range of techniques have been identified and implemented, basically falling into three categories:(1)dual-stack techni

63、ques, 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 dev

64、ices 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 IPv4this multi-protocol approach is familiar and well-understood

65、 (e.g., for AppleTalk, IPX, etc.)note: in most cases, IPv6 will be bundled with new OS releases, not an extra-cost add-onApplications (or libraries) choose IP version to usewhen initiating, based on DNS response:if (

66、dest has AAAA or A6 record) use IPv6, else use IPv4when responding, based on version of initiating packetThis allows indefinite co-existence of IPv4 and IPv6, and gradual, app-by-app upgrades to IPv6 usage,Tunnels to G

67、et ThroughIPv6-Ignorant Routers / Switches,Encapsulate IPv6 packets inside IPv4 packets(or MPLS frames)any methods exist for establishing tunnels:manual configuration“tunnel brokers” (using web-based service to crea

68、te a tunnel)“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 link-layer, oran IPv6 VPN (virtual

69、public network), over the IPv4 Internet(becoming “l(fā)ess virtual” over time, we hope),IPv6 Tunnelling,Configured tunnels—manual point-2-point linksAutomatic tunnels—via 6to4 mechanism 2