IPv6 IPv6, MPLS History Next generation IP (AKA IPng) Intended to extend address space and routing limitations of IPv4 Requires header change Attempted to include everything new in one change IETF moderated Based on Simple Internet Protocol Plus (SIPP) IPv6 IPv6 Addresses Wish list 128-bit 128-bit addresses 3.4 x 10 38 addresses (as compared to 4 x 10 9 ) Multicast traffic Classless addressing/routing (similar to CIDR) Mobility Address notation Real-time traffic/quality of service guarantees String of eight 16-bit hex values separated by colons Authentication and security 5CFA:0002:0000:0000:CF07:1234:5678:FFCD Autoconfiguration for local IP addresses Set of contiguous 0’s can be elided 5CFA:0002::CF07:1234:5678:FFCD End-to-end fragmentation Address assignment Protocol extensions Provider-based Smooth transition! geographic Note 3 m n o p 125-m-n-o-p Many of these functionalities have been retrofit into IPv4 010 Region ID Provider ID Subscriber ID Subnet Host IPv6 IPv4 Packet Format Prefix Address type 20 Byte minimum 0000 0000 Reserved (includes transition addresses) Mandatory fields are not always used 0000 0001 ISO NSAP (Network Service Point) Allocation e.g. fragmentation 0000 010 Novell IPX allocation Options are an unordered list of (name, value) pairs 010 Provider-based unicast 0 8 16 31 version hdr len TOS length 100 Geographic multicast ident flags offset 1111 1110 10 Link local address TTL protocol checksum 1111 1110 11 Site local address source address 1111 1111 Multicast address destination address options (variable) pad (variable) Other unassigned 1
IPv6 Packet Format IPv6 Packet Format 40 Byte minimum 0 8 16 31 Mandatory fields (almost) always used version priority flow label payload length next header hop limit Strict order on options reduces processing time source address word 1 No need to parse irrelevant options source address word 2 source address word 3 0 8 16 31 source address word 4 version priority flow label destination address word 1 destination address word 2 payload length next header hop limit source address 4 words destination address word 3 destination address 4 words destination address word 4 options (variable number, usually fixed length) options (variable number, usually fixed length) IPv6 Packet Format IPv6 Extension Headers Must appear in order Version 6 Hop-by-hop options Priority and Flow Label Miscellaneous information for routers Routing Support service guarantees Full/partial route to follow Allow “fair” bandwidth allocation Fragmentation Payload Length IP fragmentation info Header not included Authentication Next Header Sender identification Combines options and protocol Encrypted security payload Linked list of options Information about contents Ends with higher-level protocol header (e.g. TCP) Destination options Hop Limit Information for destination TTL renamed to match usage IPv6 Extension Headers IPv6 Extension Headers Hop-by-Hop extension 0 8 16 31 next header 0 # of addresses next address Length is in bytes beyond mandatory 8 strict/loose routing bitmap 0 8 16 31 next header length type 1 – 24 addresses value Jumbogram option (packet longer than 65,535 Routing extension bytes) Up to 24 “anycast” addresses target AS’s/providers Next address tracks current target Payload length in main header set to 0 Strict routing requires direct link 0 8 16 31 Loose routing allows intermediate nodes 0 0 next header 194 Payload length in bytes 2
IPv6 Extension Headers IPv6 Extension Headers Authentication extension 0 8 16 31 next header reserved offset reserved M Designed to be very flexible ident Includes Security parameters index (SPI) Fragmentation extension Authentication data Similar to IPv4 fragmentation Encryption Extension Called encapsulating security payload (ESP) 13-bit offset Includes an SPI Last fragment mark (M) All headers and data after ESP are encrypted Larger fragment identification field IPv6 Design Controversies IPv6 Design Controversies Address length Hop limit 65,535 8 byte 32 hop paths are common now Might run out in a few decades In a decade, we may see much longer paths Less header overhead 255 16 byte Objective is to limit lost packet lifetime More overhead Good network design makes long paths unlikely Good for foreseeable future Source to backbone 20 byte Across backbone Even more overhead Backbone to destination Compatible with OSI Variable length IPv6 Design Controversies IPv6 Design Controversies Greater than 64KB data Keep checksum Good for supercomputer/high bandwidth Removing checksum from IP is applications analogous to removing brakes from a car Too much overhead to fragment large data Light and faster packets Unprepared for the unexpected 64 KB data Remove checksum More compatible with low-bandwidth lines Typically duplicated in data link and 1 MB packet ties up a 1.5MBps line for more transport layers than 5 seconds Inconveniences interactive users Very expensive in IPv4 3
IPv6 Design Controversies IPv6 Design Controversies Security Mobile hosts Direct or indirect connectivity Where? Reconnect directly using canonical address Network layer Use home and foreign agents to forward traffic A standard service Mobility introduces asymmetry Application layer No viable standard Base station signal is strong, heard by mobile units Application susceptible to errors in network Mobile unit signal is weak and susceptible to interference, implementation may not be heard by base station Expensive to turn on and off How? Political import/export issues Cryptographic strength issues Tunneling Transition From IPv4 To IPv6 Not all routers can be upgraded E F A B simultaneous tunnel Logical view: IPv6 IPv6 IPv6 IPv6 no “flag days” E F How will the network operate with mixed A B Physical view: IPv4 and IPv6 routers? IPv6 IPv6 IPv6 IPv6 IPv4 IPv4 Tunneling: IPv6 carried as payload in IPv4 datagram among IPv4 routers Tunneling Multiprotocol label switching (MPLS) E F A B tunnel Logical view: IPv6 IPv6 IPv6 IPv6 initial goal: speed up IP forwarding by using fixed length label (instead of IP address) to F A B C D E Physical view: do forwarding IPv6 IPv6 IPv6 IPv4 IPv4 IPv6 borrowing ideas from Virtual Circuit (VC) Src:B Src:B Flow: X Flow: X approach Src: A Dest: E Dest: E Src: A Dest: F Dest: F PPP or Ethernet Flow: X but IP datagram still keeps IP address! Flow: X MPLS header IP header remainder of link-layer frame Src: A Src: A header Dest: F data Dest: F data data data label Exp S TTL A-to-B: E-to-F: B-to-C: B-to-C: 1 5 IPv6 IPv6 20 3 IPv6 inside IPv6 inside IPv4 IPv4 4
MPLS capable routers MPLS forwarding tables a.k.a. label-switched router in out out label label dest interface forwards packets to outgoing interface based 10 A 0 in out out only on label value (don’t inspect IP address) label label dest interface 12 D 0 10 6 A 1 8 A 1 MPLS forwarding table distinct from IP forwarding 12 9 D 0 tables R6 signaling protocol needed to set up forwarding 0 0 D RSVP-TE 1 1 R3 R4 forwarding possible along paths that IP alone would R5 not allow (e.g., source-specific routing) !! 0 0 A use MPLS for traffic engineering R2 in out out R1 label label dest interface must co-exist with IP-only routers in out out label label dest interface 6 - A 0 8 6 A 0 5
Recommend
More recommend
