430 likes | 611 Vues
2. Plan. Introduction : TCP/IP(v4|v6)Comparaison des en-ttesArchitecture d'adressage v6DNS / DHCPIPv6 en 2005. 3. Introduction : TCP/IP. Recherches finances par USA (projet DARPA)1983 : adopt Military Standard => RFC739 (TCP) et RFC791 (IP)TCP : Transport Control Protocol Protocole l
E N D
1. 1 Introduction IPv6
Alexis Dorais-Joncas
2. 2 Plan
Introduction : TCP/IP(v4|v6)
Comparaison des en-ttes
Architecture dadressage v6
DNS / DHCP
IPv6 en 2005
3. 3 Introduction : TCP/IP Recherches finances par USA (projet DARPA)
1983 : adopt Military Standard => RFC739 (TCP) et RFC791 (IP)
TCP : Transport Control Protocol
Protocole layer-4 (transport)
tablissement de sessions (3-way handshake)
Fiable (retransmission, checksum)
4. 4 Introduction : TCP/IP IPv4 : Internet Protocol v4
Protocole layer-3 (rseau)
Permet dchanger un payload travers diffrents rseaux
Adressage sur 32 bits (4,294,967,296 adresses max, en omettant les pertes dues au sous-rseautage)
5. 5 Introduction : TCP/IP IPv6 : Internet Protocol v6
Objectifs similaires IPv4
Adressage 128 bits : 340,282,366,920,938,463,463,374,607,431,768,211,456 adresses total
Objectif : ISP devrait dlguer un /64 pour chaque client => 18,446,744,073,709,551,616 adresses dans votre maison
6. 6 (Parenthse : Notation CIDR) Classless Inter Domain Routing (CIDR)
Permet lagrgation des prfixes
diminution de la taille de la table de routage
Assignation des adresses plus granulaire
limination des barrires des classes de sous-rseaux prtablies (p. ex. : classe A -> 255.0.0.0, classe B -> 255.255.0.0, etc)
7. 7 Notation standard :
10.0.0.0 netmask 255.0.0.0
Notation CIDR (utilise systmatiquement avec IPv6) :
10.0.0.0/8
3ffe:b00::/32
Consiste identifier le nombre de bits rservs ladresse du rseau, autrement dit identifier la taille du rseau
(Parenthse : Notation CIDR)
8. 8 Introduction : TCP/IP Quelques avantages de IPv6
Autoconfiguration (RA/RS=>SA, DAD, PMTU discovery)
limination du NAT (pansement sur un problme)
Sans NAT => vraie communication end-to-end (IPSec)
MobileIPv6 -> limination du routage en triangle
9. 9 Introduction : TCP/IP
10. 10 Introduction : TCP/IP IPv5 ??
Protocole Stream (SP)
Exprimental
Objectif : Rservation de ressources
Mme systme dadressage que IPv4, devait coexister avec IPv4 et non le remplacer
Protocole mort-n
11. 11 Comparaison des en-ttes
12. 12 Comparaison des en-ttes
13. 13 Comparaison des en-ttes
14. 14 Comparaison des en-ttes Extension Headers
Hop-By-Hop Options Header
Destination Options Header
Routing Header
Fragment Header
Authentication Header
Encapsulating Security Payload Header
Peu utiliss, permettent lextension du protocole
15. 15 Comparaison des en-ttes Notes gnrales
MTU : Maximum Transmission Unit, taille des donnes pouvant tre transmises par la couche infrieure Liaison (link-layer, p. ex. Ethernet)
68 octets <= IPv4 MTU <= 65536 octets
1280 octets <= IPv6 MTU <= 65536 octets
Possibilit dun MTU maximal de 4 GB avec loption Jumbogram du Hop-By-Hop extension header
videmment utile seulement avec un protocole link-layer qui supporte cette taille de MTU (futures technologies)
16. 16 Adressage IPv6
17. 17 Adressage IPv6 Reprsentation hexadcimale plutt que dcimale :
3ffe:0501:0000:0000:babe:97ff:dead:beef
Suite de 0 => :: (3ffe:0501::babe:)
IPv6 dans un URL : entre []
http://[3ffe:1:5::35b1]:80/index.html
Long et facile se tromper
Utilisation des FQDN recommande
18. 18 Adressage IPv6 Types dadresses :
Unicast
Multicast
Anycast
Noter la disparition du broadcast
19. 19 Adressage IPv6 Unicast
Global (publique(=)
20. 20 Adressage IPv6 Link-Local
Configuration automatique
Scope strictement limit au sous-rseau
Utilise pour
Neighbor discovery, router discovery, protocoles de routage
Prfixe : FE80::/10
21. 21 Adressage IPv6 Special purposes
Unspecified => :: (tous des 0)
Utilise pour la requte DHCP initiale et DAD
Loopback => ::1 (mme fonction que 127.0.0.1)
22. 22 Adressage IPv6 Multicast
Identifie un groupe dinterfaces
Une interface peut faire partie de 0n groupes
Prfixe : FF00/8
Remplace le broadcast
23. 23 Adressage IPv6 Adresses multicast assignes :
FF02::1 => All link nodes
FF02::2 => All link routers
FF02::9 => All link RIP Routers
FF05::101 => All site NTP Servers
24. 24 Adressage IPv6 Anycast
Adresse assigne plus dune interface
Principe de un-au-plus-proche
Adresses alloues partir du bloc dadresses Unicast
Usage actuel trs limit
Router-subnet anycast
MobileIPv6 Home-Agent anycast
25. 25 Adressage IPv6 Adresses requises sur un nud :
Link-local pour chaque interface
Loopback
All-nodes multicast address
Unicast / anycast addresses au besoin
Solicited-node multicast adress pour chaque adresse (any|uni)cast (Neighbour discovery, rempalce ARP)
Gnralement, un nud possde donc plus dune adresse
Force dIPv6 : espace dadressage suffisant permettant un design de protocoles simples et efficaces
26. 26
DNS / DHCP
27. 27 DNS
28. 28 DHCPv6 Version amliore de DHCPv4
Permet un meilleur contrle que les mcanismes dautoconfiguration
Permet la configuration dans un environnement sans routeur
Dynamic DNS updates
Utilisation typique
Router sollicitation
Si la rponse mentionne DHCP ou aucun router nest dcouvert : DHCP-Sollicit
Multicast utilis :
FF02::1:2 => all-dhcp-agents (serveur DHCP ou relais)
FF05::1:3 => all-dhcp-servers (site-local scope)
29. 29
IPv6 en 2005
30. 30 IPv6 en 2005
31. 31 IPv6 en 2005 Prfixes annoncs dans la table de routage globale (2002) :
32. 32 IPv6 en 2005 Forte activit en Asie => plus touchs par le manque dadresses IPv4
Cellulaires v6-enabled sont monnaie courante en Asie
IPv6 rsidentiel via tunneling est en phase de tests
Amrique du Nord : intrt grandissant mais rien de concrt
Hexago (www.hexago.com)
Qubcois
Leader nord-amricain en migration v4/v6
Fournisseur du tunnel broker Freenet6 (www.freenet6.net)
33. 33 IPv6 en 2005 Systmes dexploitation v6-enabled:
*BSD (Kame stack)
MacOS X
Linux (Usagi)
Solaris 8+
Windows 2000 : Advanced Networking pack (experimental)
Windows XP : Advanced Networking pack disponible dans le SP1 seulement
34. 34 IPv6 en 2005 Comment avoir un rseau IPv6 aujourdhui Qubec?
Tunnel broker (freenet6, sixxs, he.net)
Routeur (recommandation : Linux ou BSD)
Clients v6-enabled
35. 35 IPv6 en 2005
36. 36 IPv6 en 2005
37. 37
Merci !
Questions ?
38. 38 Solicited-Node Multicast Addresses
In addition to the regular multicast addresses, each unicast address has a special multicast address called its solicited-node address. This address is created through a special mapping from the devices unicast address. Solicited-node addresses are used by the IPv6 Neighbor Discovery (ND) protocol to provide more efficient address resolution than the ARP technique used in IPv4.
All solicited-node addresses have their T flag set to zero and a scope ID of 2, so they start with FF02. The 112-bit group ID is broken down as follows
39. 39
40. 40 80 bits consisting of 79 zeroes followed by a single one; this means that in colon hexadecimal notation, the next five hexadecimal values are 0000:0000:0000:0000:0001, or more succinctly, 0:0:0:0:1.
8 ones: FF.
24 bits taken from the bottom 24 bits of its unicast address.
So, these addresses start with FF02:0:0:0:0:1:FF followed by the bottom 24 bits of the unicast address. So, the node with IP address 805B:2D9D:DC28:0:0:FC57:D4C8:1FFF would have a solicited-node address of FF02:0:0:0:0:1:FFC8:1FFF (or FF02::1:FFC8:1FFF).
Key Concept: Each unicast address has an equivalent solicited-node multicast address, which is created from the unicast address and used when other devices need to reach it on the local network.
41. 41 3.3. IPv6 multicast over Ethernet
To send an IPv6 multicast packet over Ethernet, one simply takes the last 32 bits of the destination IPv6 address, prepends 33-33- and uses that as the destination Ethernet address. Thus, an IPv6 packet addressed to FF02::1:FF68:12CB would be sent to the Ethernet address 33-33-FF-68-12-CB. Any host which is interested in packets for that IPv6 address is expected to be listening for the corresponding Ethernet address.
42. 42 3.4. Neighbour discovery (RFC 2461)
Where IPv4 has ARP, IPv6 has NDP, the neighbour discovery protocol. For simple purposes, NDP and ARP are very similar: one node sends out a request packet (called a neighbour solicitation in NDP), and the node it was looking for sends back a reply (neighbour advertisement) giving its link-layer address. NDP is part of ICMPv6, unlike ARP, which doesn't even run over IP. NDP also uses multicast rather than broadcast packets, and that deserves a little more explanation.
43. 43 For each unicast address it responds to, each host listens on a solicited-node multicast address. The solicited-node multicast address for a given unicast address is constructed by taking the last three octets of the unicast address and prepending FF02::1:FF00:0000/104. Thus, the solicited-node multicast address of 2001:630:200:8100:02C0:4FFF:FE68:12CB is FF02::1:FF68:12CB. It's the solicited-node multicast address that a node uses as the destination of a neighbour solicitation packet. This use of multicast means that most hosts don't get disturbed by neighbour solicitations that aren't either for them or for a host with a very similar IPv6 address.