BRKIPM-1261-rev Cisco and/or its affiliates. All rights reserved. Cisco Public

Transcription

1 1

2 Introduction to IP Multicast BRKIPM-1261 Beau Williamson CiscoLive Distinguished Speaker 2

3 Session Goals To provide you with an understanding of the concepts, mechanics and protocols used to build IP multicast networks. To enable you to ask the right questions, and make the correct architectural decisions in deploying and maintaining an IP Multicast enabled network. To prove that Multicast doesn t have to be: Hard Scary 3

4 Agenda Why Multicast? Multicast History and Fundamentals PIM Protocols RP Choices Multicast at Layer 2 Interdomain IP Multicast IPv6 Multicast Geekometer 4

5 Why Multicast? 5

6 Unicast vs. Multicast Scaling Unicast Server Number of Streams Router Multicast Server Router 6

7 Multicast History & Fundamentals 7

8 A Brief History of Multicast Steven Deering, 1985, Stanford University Yeah, he was way ahead of his time and too clever for all of us. A solution for layer2 applications in the growing layer3 campus network Think overlay broadcast domain Broadcast Domain all members receive all members can source members dynamically come and go 8

9 A Brief History of Multicast RFC966 A Multicast Extension to the Internet Protocol Circa 1985 Aka The Multicast Dead-Sea Scrolls by some people Multi-destination delivery is useful to several applications, including: distributed, replicated databases [6,9]. conferencing [11]. distributed parallel computation, including distributed gaming [2]. All inherently many-to-many applications No mention of one-to-many services such as Video/IPTV 9

10 A Brief History of Multicast Overlay Broadcast Domain Requirements Tree building and maintenance Network-based source discovery Source route information Overlay mechanism tunneling The first solution had it all Distance Vector Multicast Routing Protocol DVMRP, RFC

11 A Brief History of Multicast PIM Protocol Independent Multicast Independent of which unicast routing protocol you run It does require that you re running one. Uses local routing table to determine route to sources Router-to-router protocol to build and maintain distribution trees Source discovery handled one of two ways: 1. Flood-and-prune PIM-DM, Dense Mode 2. Explicit Join w/ Rendezvous Point (RP) PIM-SM Sparse Mode - The Current Standard 11

12 A Brief History of Multicast PIM-SM Protocol Independent Multicast Sparse Mode Tree building and maintenance Network-based source discovery Source route information Overlay mechanism tunneling Long, Sordid IETF history RFC (original draft was rewritten from scratch) Primary challenges to the final specification were in addressing Network-based source discovery. 12

13 A Brief History of Multicast Today s dominant applications are primarily one-to-many IPTV, Contribution video over IP, etc. Sources are well known SSM Source Specific Multicast RFC3569, RFC Tree building and maintenance Network-based source discovery Source route information Overlay mechanism tunneling Very simple and the preferred solution for one-to-many applications 13

14 Multicast Uses Any applications with multiple receivers One-to-many or many-to-many Live video distribution Collaborative groupware Periodic data delivery push technology Stock quotes, sports scores, magazines, newspapers, adverts Server/Website replication Reducing network/resource overhead More than multiple point-to-point flows Resource discovery Distributed interactive simulation (DIS) War games Virtual reality 2015 Cisco and/or its affiliates. All 14 rights reserved. Cisco Public 14

15 Multicast Considerations Multicast Is UDP-Based Best effort delivery: Drops are to be expected; multicast applications should not expect reliable delivery of data and should be designed accordingly; reliable multicast is still an area for much research; expect to see more developments in this area; PGM, FEC, QoS No congestion avoidance: Lack of TCP windowing and slow-start mechanisms can result in network congestion; if possible, multicast applications should attempt to detect and avoid congestion conditions Duplicates: Some multicast protocol mechanisms (e.g., asserts, registers, and SPT transitions) result in the occasional generation of duplicate packets; multicast applications should be designed to expect occasional duplicate packets Out of order delivery: Some protocol mechanisms may also result in out of order delivery of packets 15

16 Multicast Components Cisco End-to-End Architecture ISP A Multicast Source X DR RP MSDP RP ISP B ISP B Multicast Source Y IGMP Snooping PIM Snooping ISP A MBGP End stations (hosts-to-routers) IGMP, MLD, AMT Campus Multicast Switches (Layer 2 optimization) IGMP snooping IGMP AMT Routers (multicast forwarding protocol) PIM sparse mode or bidirectional PIM DR PIM-SM: ASM, SSM, BiDir MVPN, BIER DR Multicast routing across domains MBGP Interdomain Multicast Multicast source discovery MSDP with PIM-SM Source Specific Multicast SSM 16

17 Multicast Fundamentals 17

18 Multicast Fundamentals Multicast Myth Busters Multicast is complicated, scary and hard to understand! 18

19 Unicast vs. Multicast Addressing src addr: A unique packet addressed to each destination IP Address src addr: Same packet addressed to Group destination address... Multicast Group Address e.g

20 Unicast vs. Multicast Addressing src addr: A unique packet addressed to each destination IP Address src addr: replicated at each node along the tree. Multicast Group Address e.g

21 Multicast Addressing Multicast Addressing IPv4 Header Version IHL Type of Service Total Length Identification Flags Fragment Offset Time to Live Protocol Header Checksum Source Source Source Always Addressthe unique unicast origin address of (Class A, B, C) the packet same as unicast Destination Destination Address (Class D) Multicast Group Address Range Options Padding 21

22 Multicast Addressing Class D Group addresses 224/4 Multicast Group addresses are NOT in the unicast route table. A separate multicast route table is maintained for active multicast trees. Multicast trees are initiated by receivers signaling their request to join a group. Sources do not need to join, they just send. Multicast routing protocols build the trees: Hop-by-hop, from the receivers (tree leaves) to the source (tree root). Tree path follows the unicast route table backward to the source using source address. Multicast relies on a dependable unicast infrastructure. 22

23 Multicast Addressing 224/4 Reserved link-local addresses Transmitted with TTL = 1 Examples All systems on this subnet All routers on this subnet OSPF routers PIMv2 routers IGMPv3 Other IANA reserved addresses Not local in scope (transmitted with TTL > 1) Examples NTP (Network Time Protocol) Mtrace routers Tibco Multicast1 23

24 Multicast Addressing 224/4 Administratively scoped addresses Private address space Similar to RFC1918 unicast addresses Not used for global Internet traffic scoped traffic SSM (Source Specific Multicast) range Primarily targeted for Internet-style broadcast 24

25 Multicast Addressing IP Multicast MAC Address Mapping 5 Bits Lost Bits 28 Bits e-7f Bits 23 Bits 48 Bits 25

26 Multicast Addressing IP Multicast MAC Address Mapping Be Aware of the 32:1 Address Overlap 32 IP Multicast Addresses Multicast MAC Address 0x0100.5E

27 How are Multicast Flows Identified Every Multicast Flow can be identified by two components: Source IP Address The address of the Sender Multicast Group Address 224/4 (Class D) IP Address Multicast Flow from Source to Group Example ( , ), 3w1d/00:02:40, flags: s Incoming interface: Ethernet 0/0, RPF nbr Outgoing interface list: Ethernet 1/0, Forward/Sparse, 3w1d/00:02:40 Ethernet 2/0, Forward/Sparse, 2w0d/00:02:33 How do Hosts Signal to Routers which flow they want? IPv4: IGMP IPv6: MLD 27

28 Host-Router Signaling: IGMP IGMP Version 3 is current version RFC3376 Oct 2002 (Over 10 years old) Uses (IGMPv3 routers) Link-Local Multicast Address All IGMP hosts send Membership Reports to this address All IGMP routers listen to this address Hosts do not listen or respond to this address (unlike previous IGMP versions) Membership Reports Sent by Hosts Contain list of Multicast (Source, Group) pairs to Include/Exclude (Join/Leave) Membership Queries Sent by Routers to refresh/maintain list of Multicast traffic to deliver. 28

29 IGMPv3 Membership Report Packet Format Type = 0x Reserved Reserved Group Record [1] Group Record [2]... Group Record [M] Checksum # of Group Records (M) # of Group Records (M) Number of Group Records in Report Group Records 1 - M Group address plus list of zero or more sources to Include/Exclude (See Group Record format) Record Type Record Type Include, Exclude, Chg-to-Include, Chg-to-Exclude, Allow New Srcs, Block Old Srcs # of Sources (N) Number of Sources in Record Source Address 1- N Address of Source Aux Data Len # of Sources (N) Multicast Group Address Source Address [1] Source Address [2].. Source Address [N] Auxiliary Data 29

30 IGMPv3 Query Packet Format Type = 0x11 IGMP Query Max. Resp. Time Max. time to send a response if < 128, Time in 1/10 secs if > 128, FP value ( secs) Group Address: Multicast Group Address ( for General Queries) S Flag Suppresses processing by routers QRV (Querier Robustness Value) Affects timers and # of retries QQIC (Querier s Query Interval) Same format as Max. Resp. Time Number of Sources (N) (Non-zero for Group-and-Source Query) Source Address Address of Source Type = 0x11 S QRV Max. Resp. Code Group Address QQIC Source Address [1] Source Address [2]... Source Address [N] Checksum Number of Sources (N) 30

31 IGMPv3 Joining Group G Source S H1 H2 H3 Type: Allow New (5) Group: Source: { } Report ( ) show ip igmp groups detail Flags: L - Local, U - User, SG - Static Group, VG - Virtual Group, SS - Static Source, VS - Virtual Source, Ac - Group accounted towards access control limit Interface: GigabitEthernet3/3 Group: Flags: SSM Uptime: 00:01:14 Group mode: INCLUDE Last reporter: Group source list: (C - Cisco Src Report, U - URD, R - Remote, S - Static, V - Virtual, M - SSM Mapping, L - Local, Ac - Channel accounted towards access control limit) Source Address Uptime v3 Exp CSR Exp Fwd Flags :01:14 00:02:08 stopped Yes R 31

32 IGMPv3 Maintaining State H1 H2 H3 Type: Include (1) Group: Source: { } Report ( ) Type: Include (1) Group: Source: { } Report ( ) Type: Include (1) Group: Source: { } Report ( ) Router sends periodic General Queries to All Hosts General Query: Group=0, #Srcs=0 All IGMP members respond Reports can contain multiple Group State records ( ) Query Group: Source: {} Hn Member Group: Source: (Destination IP Address) 32

33 IGMPv3 Leaving Group G Source S H1 1 H2 H3 Type: Block Old (6) Group: Source: { } Report ( ) 1. H2 leaves Group-Source 2. Sends Block Old Membership Report 3. Router sends Group-Source Query Group-Source Query: Group=G, #Srcs=N ( ) Query Group: Source: { } Hn 3 Member Group: Source: (Destination IP Address) 33

34 IGMPv3 Leaving Group G Source S H1 H2 H3 Type: Include Group: Source: { } Report ( ) H2 leaves Group-Source 2. Sends Block Old Membership Report 3. Router sends Group-Source Query 4. A remaining member host sends report 5. Group-Source flow remains active Hn Member Group: Source: (Destination IP Address) 34

35 IGMPv3 Leaving Group G Source S H1 H2 6 H3 Type: Block Old (6) Group: Source: { } Report 7 ( ) 6. H3 leaves Group-Source 7. Sends Block Old Membership Report 8. Router sends Group-Source Query ( ) Query Group: Source: { } Hn 8 Member Group: Source: (Destination IP Address) 35

36 IGMPv3 Leaving Group G Source S H1 H2 H H3 leaves Group-Source 7. Sends Remove Membership Report 8. Router sends Group-Source specific query 9. State times out. Group-Source flow pruned. Hn Member Group: Source: (Destination IP Address) 36

37 Unicast vs. Multicast Routing/Forwarding Unicast Routing/Forwarding Destination IP address directly indicates where to forward packet Unicast Routing protocols build a table of destination/interface/next-hop triples Unicast Forwarding is hop-by-hop simply based on these entries Unicast routing table determines interface and next-hop router to forward packet 37

38 Unicast vs. Multicast Routing/Forwarding Multicast Routing & Forwarding Destination Group address doesn t directly indicate where to forward packet. Multicast Routing is Backwards from Unicast Routing Multicast Routing builds a tree backwards from the receivers to the source. Concerned with Where the packet will come from? More specifically, What s the route back to the Source? Trees are built via connection requests called Joins sent toward the source. Joins follow the unicast routing table backwards toward the source. Joins create Multicast tree/forwarding state in the routers along the tree. Trees are rebuilt dynamically in case of network topology changes. Only when a tree is completely built from receiver backwards to the source can source traffic flow down the tree to the receivers. Say that over and over to yourself when working with Multicast! 38

39 Unicast vs. Multicast Routing/Forwarding Multicast Routing & Forwarding All of this can easily lead to thinking with your Unicast Lizard Brain! If you ever get confused by Multicast, just remember to... Stand on your head! 39

40 Multicast Tree Building 1. Multicast packet s source address is checked against the unicast routing table 2. Determines interface & next-hop multicast router in the direction of the source Which is where the Joins are sent 3. This interface becomes the Incoming interface Often referred to as the RPF (Reverse Path Forwarding) interface A router forwards a multicast datagram only if received on the Incoming/RPF interface A bit of History The term RPF is actually a left-over from early Dense mode Multicast days Multicast traffic was flooded everywhere (i.e. no explicit Join signaling to build trees) Traffic was only accepted on the RPF interface to avoid loops We still tend to use the term to indicate the calculation of the Incoming interface. 40

41 Multicast Routing & Forwarding Traffic to Source Receiver 41

42 Multicast Routing & Forwarding Traffic to Source IGMP Join ( , ) Receiver Mroute Entry 42

43 Multicast Routing & Forwarding Traffic to Source PIM Join ( , ) Mroute Entry Receiver Mroute Entry 43

44 Multicast Routing & Forwarding Traffic to PIM Join ( , ) Source Mroute Entry Mroute Entry Receiver Mroute Entry 44

45 Multicast Routing & Forwarding Traffic to Source Mroute Entry Mroute Entry Receiver Mroute Entry 45

46 Multicast Routing & Forwarding Traffic to IGMP Join ( , ) Source Mroute Entry Mroute Entry Receiver Mroute Entry Receiver Mroute Entry 46

47 Multicast Routing & Forwarding Traffic to PIM Join ( , ) Source Mroute Entry Mroute Entry Receiver Mroute Entry Receiver Mroute Entry 47

48 Multicast Routing & Forwarding Traffic to Source Mroute Entry Mroute Entry Receiver Mroute Entry Receiver Mroute Entry 48

49 Multicast Routing & Forwarding Mroute Entry Source Traffic to show ip mroute ( , ), 3w1d/00:02:40, flags: s Incoming interface: Ethernet 0/0, RPF nbr Outgoing interface list: Ethernet 1/0, Forward/Sparse, 3w1d/00:02:40 Ethernet 2/0, Forward/Sparse, 2w0d/00:02:33 Receiver Receiver 49

50 Multicast Routing & Forwarding Mroute Entry Source Traffic to show ip mroute ( , ), 3w1d/00:02:40, flags: s Incoming interface: Ethernet 0/0, RPF nbr Outgoing interface list: Ethernet 1/0, Forward/Sparse, 3w1d/00:02:40 Ethernet 2/0, Forward/Sparse, 2w0d/00:02:33 Receiver Receiver This type of Multicast has a special name: Source Specific Multicast (SSM) 50

51 Multicast Tree Building RPF Calculation Based on source address SRC Best path to source found in unicast route table A Determines where to send join Joins continue towards source to build multicast tree Multicast data flows down tree B Join Join E0 E1 C D Unicast Route Table Network Interface /24 E0 E E2 R Cisco and/or its affiliates. All 51 rights reserved. Cisco Public 51

52 Multicast Tree Building RPF Calculation Based on source address SRC Best path to source found in unicast route table Determines where to send join Joins continue towards source to build multicast tree Multicast data flows down tree B E0 A E1 Join C Join D R2 E E2 R Cisco and/or its affiliates. All 52 rights reserved. Cisco Public 52

53 Multicast Tree Building RPF Calculation What if we have equal-cost paths? We can t use both Tie-breaker Use highest next-hop IP address B SRC A C D Join E0 E1 E Unicast Route Table Network Intfc Nxt-Hop /24 E /24 E R1 F E Cisco and/or its affiliates. All 53 rights reserved. Cisco Public 53

54 Multicast State Multicast route entries are in (S,G) form. rtr-a#show ip mroute ( , ), 3w1d/00:02:40, flags: s Incoming interface: Ethernet 0/0, RPF nbr Outgoing interface list: Ethernet 1/0, Forward/Sparse, 3w1d/00:02:40 Ethernet 2/0, Forward/Sparse, 2w0d/00:02:33 Incoming interface points upstream toward the root of the tree (i.e. Source) rtr-b#show ip route % Network not in table Multicast Group addresses are NEVER in the unicast route table. Outgoing interface list is where receivers have joined downstream and where packets will be replicated and forwarded downstream. 54

55 Multicast Routing & Forwarding Key Point If you ever get confused by Multicast just remember to Stand on your head. (Because Multicast is an Upside-down world where we are interested in where the packet came from, not its destination address.) 55

56 Multicast Fundamentals Multicast Myth Busters Multicast is complicated, scary and hard to understand! 56

57 Multicast Fundamentals See there that wasn t so hard, was it? 57

58 What makes Multicast Complicated? 58

59 Biggest Multicast Complicating Factor Network-Based Source Discovery Lazy One-to-Many Application Developers Let s just let the Network do all the work to keep track of Sources. Uses old and outdated IGMPv2 methods to join (*,G) only. Really!!!! IGMPv3 has been out for 10+ years!! Even Apple OS supports IGMPv3 Suffers from Capt. Midnight stream hijacking Complicates Multicast Address management/allocation Now you have to worry about what application uses what Multicast Address Requires complex Any-Source Multicast (ASM) & Rendezvous Point Engineering/Mgmt Or maybe BiDir Multicast Source Specific Multicast (SSM) avoided all of these network issues 59

60 Multicast Complicating Factor Network-Based Source Discovery Ad-Hoc Multicast Applications No good way to know or predict who will become a source. Sometimes you just have to support Network-Based Source Discovery Requires complex Any-Source Multicast (ASM) & Rendezvous Point Engineering/Mgmt Or maybe BiDir Multicast 60

61 Network-Based Source Discovery Requires Any-Source Multicast (ASM) Much, much more complicated than SSM or Bidir Requires physical Rendezvous Point (RP) router(s) & RP Redundancy methods Uses Shortest-Path Trees (ala SSM) to first deliver traffic to RP Then uses a common Shared Tree rooted at RP to deliver all Multicast traffic Routers w/directly connected receivers then learn about new sources via Shared Tree Then join Shortest-Path Tree to all the sources.... or use of Bidir Multicast (Bidir)... A bit more complicated than Source Specific Multicast but easier than ASM. True Broadcast like Multicast Uses a common Shared Tree to deliver all Multicast traffic Traffic flows both up and down the Tree to get to the Receivers (i.e. Bidirectional) Shared Tree rooted at some designated location in the network (not necessarily a router). 61

62 Types of Multicast Trees 62

63 Multicast Distribution Tree Types Shortest Path or Source Distribution Tree Source 1 A Notation: (S1, G) S = Source G = Group B D F Source 2 C E Receiver 3 Receiver 1 Receiver 2 63

64 Multicast Distribution Trees Shortest Path or Source Distribution Tree Source 1 Notation: (S2, G) S = Source G = Group A B D F Source 2 C E Receiver 3 Receiver 1 Receiver 2 These are the type of trees we ve been talking about so far. 64

65 Multicast Distribution Trees Shared Distribution Tree Source 1 Notation: (*, G) * = All Sources G = Group A B D F RP Source 2 C E (RP) PIM Rendezvous Point Shared Tree Receiver 1 Receiver 2 65

66 Multicast Distribution Trees Shared Distribution Tree Source 1 Notation: (*, G) * = All Sources G = Group A B D RP F Source 2 C Receiver 1 E Receiver 2 (RP) PIM Rendezvous Point Shared Tree Source Tree(s) (used to get traffic to RP) 66

67 Multicast Distribution Trees Characteristics of Distribution Trees Source or shortest path trees Uses more memory O (S x G) but you get optimal paths from source to all receivers; minimizes delay Shared trees Uses less memory O(G) but you may get suboptimal paths from source to all receivers; may introduce extra delay 67

68 Bidirectional Multicast i.e. Bidir PIM 68

69 Bidirectional (BiDir) PIM Idea: Use Shared Tree to bring traffic up from sources to the RP and then down to receivers Benefits: Less state in routers Only (*, G) state is used Source traffic follows the Shared Tree Flows up the Shared Tree to reach the RP Flows down the Shared Tree to reach all other receivers 69

70 PIM Modifications for BiDir Operation All trees rooted at the RP Data traveling from source toward RP is moving UPSTREAM Data traveling from RP toward receivers is moving DOWNSTREAM Designated Forwarders (DF) One DF per link Router with best path to the RP is elected DF Election mechanism insures all routers on link agree on who is DF Prevents route loops from forming BiDir (*,G) forwarding rules: DF is the only router that picks-up upstream traveling packets off the link to forward towards the RP 70

71 Bidir Forwarding/Tree Building RP E0 (DF) PIM (*,G) Join E0 E0 E E1 (DF) F E1 (DF) PIM (*,G) Join E0 E0 E0 E0 A B C E1 (DF) E1 (DF) E1 (DF) D E1 (DF) IGMP (*,G) Join Receiver 1 Branch of Shared Tree Is Now Built Down to Receiver 1 71

72 Bidir Forwarding/Tree Building RP E0 (DF) E0 E0 E E1 (DF) F E1 (DF) E0 E0 E0 E0 A B C E1 (DF) E1 (DF) E1 (DF) D E1 (DF) (*, ), 00:00:04/00:00:00, RP , flags: BC Bidir-Upstream: Ethernet0, RPF nbr Outgoing interface list: Ethernet0, Bidir-Upstream/Sparse-Dense, 00:00:04/00:00:00 Ethernet1, Forward/Sparse-Dense, 00:00:04/00:02:55 Example of Bidir State along Shared Tree Receiver 1 72

73 Bidir Forwarding/Tree Building RP E0 (DF) E0 E0 E E1 (DF) F E1 (DF) E0 E0 E0 E0 A B C E1 (DF) E1 (DF) E1 (DF) D E1 (DF) Source (*, ), 00:32:20/00:02:59, RP , flags: BP Bidir-Upstream: Ethernet0, RPF nbr Outgoing interface list: Receiver 1 Ethernet0, Bidir-Upstream/Sparse-Dense, 00:32:20/00:00:00 Arriving Traffic from Source Causes Router A to Create (*, G) State 73

74 Bidir Forwarding/Tree Building RP E0 (DF) E0 E0 E E1 (DF) F E1 (DF) E0 E0 E0 E0 A B C E1 (DF) E1 (DF) E1 (DF) D E1 (DF) Source Receiver 1 Arriving Traffic Causes Router E & RP to Create (*, G) State 74

75 Any-Source Multicast i.e. ASM PIM 75

76 PIM-SM Shared Tree Join RP (*, G) Join Shared Tree (*, G) State Created Only Along the Shared Tree Receiver 76

77 PIM-SM Sender Registration Source RP Traffic Flow Shared Tree Source Tree (S, G) Register (S, G) Join (unicast) Receiver (S, G) State Created Only Along the Source Tree 77

78 PIM-SM Sender Registration Source RP Traffic Flow Shared Tree Source Tree (S, G) Register (S, G) Register-Stop (unicast) (unicast) Receiver (S, G) Traffic Begins Arriving at the RP via the Source Tree RP Sends a Register-Stop Back to the First-Hop Router to Stop the Register Process 78

79 PIM-SM Sender Registration Source RP Traffic Flow Shared Tree Source Tree Receiver Source Traffic Flows Natively Along SPT to RP From RP, Traffic Flows Down the Shared Tree to Receivers 79

80 PIM-SM SPT Switchover Source RP Traffic Flow Shared Tree Source Tree (S, G) Join Receiver Last-Hop Router Joins the Source Tree Additional (S, G) State Is Created Along New Part of the Source Tree 80

81 PIM-SM SPT Switchover Source RP Traffic Flow Shared Tree Source Tree (S, G)RP-bit Prune Receiver Traffic Begins Flowing Down the New Branch of the Source Tree Additional (S, G) State Is Created Along the Shared Tree to Prune Off (S, G) Traffic 81

82 PIM-SM SPT Switchover Source RP Traffic Flow Shared Tree Source Tree Receiver (S, G) Traffic Flow Is Now Pruned Off of the Shared Tree and Is Flowing to the Receiver via the Source Tree 82

83 PIM-SM SPT Switchover Source RP Traffic Flow Shared Tree Source Tree (S, G) Prune Receiver (S, G) Traffic Flow Is No Longer Needed by the RP so It Prunes the Flow of (S, G) Traffic 83

84 PIM-SM SPT Switchover Source RP Traffic Flow Shared Tree Source Tree Receiver (S, G) Traffic Flow Is Now Only Flowing to the Receiver via a Single Branch of the Source Tree 84

85 The default behavior of PIM-SM (ASM) is that routers with directly connected members will join the shortest path tree as soon as they detect a new multicast source. PIM Frequently Forgotten Fact 85

86 But what about PIM Dense Mode?? Fuggidaboudit! Fuggidaboudit! Click Source: to edit The source Wiseguys s Guide to IP Multicast, 2005, T. Soprano It does Flood & Prune without any Joins! It can meltdown your network and blackhole your traffic! 86

87 RP Choices 87

88 How Does the Network Learn RP Address? Static configuration Manually on every router in the PIM domain AutoRP Originally a Cisco solution Facilitated PIM-SM early transition BSR draft-ietf-pim-sm-bsr 88

89 Static RPs Hard-configured RP address When used, must be configured on every router All routers must have the same RP address RP failover not possible Exception: If Anycast RPs are used Command ip pim rp-address <address> [group-list <acl>] [override] Optional group list specifies group range Default: range = /4 (includes auto-rp groups!) Override keyword overrides auto-rp information Default: auto-rp learned info takes precedence 89

90 Announce Announce Announce Announce Auto-RP From 10,000 Feet MA MA A B Announce C-RP C Announce Announce D C-RP Announce RP-Announcements Multicast to the Cisco Announce ( ) Group Announce 90

91 Auto-RP From 10,000 Feet MA MA A B C-RP C D C-RP RP-Discoveries Multicast to the Cisco Discovery ( ) Group Discovery 91

92 BSR From 10,000 Feet BSR Election Process G C-BSR D C-BSR A C-BSR F B C BSR Msgs BSR Msgs Flooded Hop-by-Hop E 92

93 BSR From 10,000 Feet Highest Priority C-BSR Is Elected as BSR G D BSR A F B C E 93

94 BSR From 10,000 Feet G BSR D A F C-RP B C C-RP E 94

95 BSR From 10,000 Feet G BSR D A F C-RP B C C-RP BSR Msgs BSR Msgs Containing RP-SET Flooded Hop-by-Hop E 95

96 Multicast at Layer2 96

97 L2 Multicast Frame Switching Problem: Layer 2 Flooding of Multicast Frames Typical L2 switches treat multicast traffic as unknown or broadcast and must flood the frame to every port Static entries can sometimes be set to specify which ports should receive which group(s) of multicast traffic Dynamic configuration of these entries would cut down on user administration PIM Multicast M 97

98 L2 Multicast Frame Switching IGMP Snooping IGMPv3 reports sent to a special link-local group ( ) Switches listen to just this group Only contains IGMP control traffic no multicast data traffic Permits low-end switches to easily implement IGMP snooping By just joining/listening to that link-local Group Switch examines contents of IGMP messages To determine which ports want what Multicast traffic Adds entries to TCAM as appropriate All IGMPv3 hosts send complete update of desired flows i.e. No Host Suppression Enables switch to perform complete individual host/port tracking 98

99 L2 Multicast Frame Switching Older IGMPv1 v2 Snooping IGMP Join/Leave packets sent to same Group Address as Multicast data Not to a separate IGMP Control address (e.g ) Switch CPU must wade through all Multicast Data packets looking for IGMP packets Requires special ($$$) hardware/asics to avoid impact to the switch Impact on older low-end, Layer 2 switches w/o special hardware Must process all Layer 2 multicast packets Admin load increases with multicast traffic load Often results in switch meltdown It worked just fine when we tested it with just some Multicast pings! 99

100 Interdomain IP Multicast 100

101 MP-BGP Overview MP-BGP: Multiprotocol BGP Originally defined in RFC 2858 (extensions to BGP) Can carry different types of routes Unicast Multicast Both routes carried in same BGP session Does not propagate multicast state info That s PIM s job Same path selection and validation rules AS-Path, LocalPref, MED 101

102 MP-BGP Overview Separate BGP tables maintained Unicast prefixes for unicast forwarding Unicast prefixes for multicast RPF checking AFI = 1, Sub-AFI = 1 Contains unicast prefixes for unicast forwarding Populated with BGP unicast NLRI AFI = 1, Sub-AFI = 2 Contains unicast prefixes for RPF checking Populated with BGP multicast NLRI 102

103 MBGP Overview MBGP Allows Divergent Paths and Policies Same IP address holds dual significance Unicast routing information Multicast RPF information For same IPv4 address two different NLRI with different next-hops Can therefore support both congruent and incongruent topologies 103

104 Inter-domain SSM Multicast Simple! MP-BGP Peering Domain E Domain B Domain C Receiver Receiver Learns S and G Out of Band, i.e., Webpage Domain D Source in 232/8 Source S Domain A 104

105 Inter-domain SSM Multicast Simple! MP-BGP Peering Domain E Multicast Traffic Domain B Domain C Receiver Data flows natively along the interdomain source tree Domain D Source in 232/8 Source S Domain A 105

106 Inter-domain ASM Really?? 106

107 Inter-domain ASM Issues Global Group Address Allocation/Management With ASM we have to make sure that we use unique Groups Otherwise we start mixing up the Multicast flows Solution(?): GLOP Addressing Put your ASN in the middle two Octets Provides /24 group prefix per ASN How do we do Inter-domain Source Discovery? Can we all agree on what domain owns the RP? And for which Global Multicast Group?? GLOP Addressing? Why not have RP s in each domain share Source information? Solution: Multicast Source Discovery Protocol (MSDP) 107

108 MSDP Multicast Source Discovery Protocol RFC 3618 ASM only RPs knows about all sources in their domain Sources cause a PIM Register to the RP Tell RPs in other domains of it s sources Via MSDP SA (Source Active) messages RPs know about receivers in a domain Receivers cause a (*, G) Join to the RP RP can join the source tree in the peer domain Via normal PIM (S, G) joins 108

109 MSDP Overview MSDP Example Domain E MSDP Peers RP Join (*, ) Domain C Receiver Domain B RP RP RP ASN770 GLOP: /24 RP Domain D Domain A 109

110 MSDP Overview MSDP Example Domain E MSDP Peers Source Active Messages SA Domain C SA RP Receiver Domain B SA RP SA SA RP SA RP SA Message , SA RP ASN770 GLOP: /24 SA Message , Domain D Source Register , Domain A 110

111 MSDP Overview MSDP Example MSDP Peers Domain E RP Domain C Receiver Domain B RP RP RP ASN770 GLOP: /24 RP Domain D Source Domain A 111

112 MSDP Overview MSDP Example MSDP Peers Multicast Traffic Domain B Domain C RP Domain E RP Receiver RP RP RP ASN770 GLOP: /24 Domain D Source Domain A 112

113 MSDP Overview MSDP Example MSDP Peers Multicast Traffic Domain B Domain C RP Domain E RP Receiver RP RP RP ASN770 GLOP: /24 Domain D Source Domain A 113

114 MSDP Overview MSDP Example MSDP Peers Multicast Traffic Domain B Domain C RP Domain E RP Receiver RP RP RP ASN770 GLOP: /24 Domain D Source Domain A 114

115 Intra-domain uses for MSDP Anycast-RP Redundant RP technique for ASM which uses MSDP for RP synchronization Uses single defined RP address Two or more routers have same RP address RP address defined as a loopback interface Loopback address advertised as a host route Senders and receivers join/register with closest RP Closest RP determined from the unicast routing table Because RP is statically defined MSDP session(s) run between all RPs Informs RPs of sources in other parts of network RPs join SPT to active sources as necessary 115

116 Anycast RP Overview Src Src RP1 X A SA MSDP SA RP2 B Rec Rec Rec Rec 116

117 Anycast RP Overview Src Src RP1 X A RP2 B Rec Rec Rec Rec 117

118 Internet IP Multicast We can build multicast distribution trees. PIM SSM We can RPF on Inter-domain Sources MP-BGP We no longer need (or want) network-based source discovery SSM So Inter-domain IP Multicast is in every home, right? 118

119 Then What s Missing??!! Without an overlay mechanism, Multicast in the Internet is an all or nothing solution Each receiver must be on an IP multicast-enabled path Many core networks have IP multicast-enabled, but few edge networks accept multicast transit traffic Dr. Deering had tunneling in the original solution (DVMRP now outdated) Multicast-aware content owners Forced to provide unicast streams to gain audience size For live content, Unicast cannot scale dynamically Splitters/caches just distribute the problem Still has a cost per user We need a way to Tunnel through non-multicast domains! 119

120 Automatic Multicast Tunneling 120

121 AMT Automatic Multicast Tunneling Automatic IP multicast without explicit tunnels RFC Automatic Multicast Tunneling Allow multicast content distribution to extend to unicast-only connected receivers Bring the flat scaling properties of multicast to the Internet Provide the benefits of multicast wherever multicast is deployed Let the networks which have deployed multicast benefit from their deployment Work seamlessly with existing applications No OS kernel changes 121

122 Native Interdomain Multicast (SSM) As long as IP Multicast is enabled on every router from the source to the receivers, the benefits of IP Multicast are realized Mcast-Enabled ISP Content Owner Unicast-Only Network Mcast Traffic Mcast Join Mcast-Enabled Local Provider 122

123 Native Interdomain Multicast (SSM) The benefits being an unlimited number of Receivers can be served with a single stream of content at no additional costs Mcast-Enabled ISP Content Owner Unicast-Only Network Mcast Traffic Mcast Join Mcast-Enabled Local Provider 123

124 AMT Automatic Multicast Tunneling The AMT Anycast Address Allows for All AMT Gateways to Find the Closest AMT Relay the Nearest Edge of the Multicast Topology of the Source Mcast-Enabled ISP Content Owner Unicast-Only Network AMT Relay AMT Host Gateway Once the Multicast Join times-out, an AMT Relay Discovery is sent from Host Gateway toward the Global AMT Anycast Address Mcast-Enabled Local Provider Mcast Traffic Mcast Join AMT Messages 124

125 AMT Automatic Multicast Tunneling The AMT Relay sends back a Relay Advertisement (i.e. use me ). Mcast-Enabled ISP Content Owner Unicast-Only Network AMT Relay AMT Host Gateway Mcast Traffic Mcast Join AMT Messages Mcast-Enabled Local Provider 125

126 AMT Automatic Multicast Tunneling (S,G) is learned from the AMT Join Message, then (S,G) PIM Join is sent toward the Source Mcast-Enabled ISP Content Owner Unicast-Only Network AMT Relay AMT Host Gateway Note: AMT Join/Leaves are really IGMP/MLD packets tunneled to/from Relay Mcast Traffic Mcast Join AMT Messags Mcast-Enabled Local Provider 126

127 AMT Automatic Multicast Tunneling AMT Relay Replicates Stream on Behalf of Downstream AMT Receiver, Adding a Unicast Header Destined to the Receiver Unicast-Only Network Mcast-Enabled ISP Content Owner AMT Relay AMT Host Gateway Mcast-Enabled Local Provider Mcast Traffic Mcast Join AMT Messages Ucast Stream 127

128 AMT Automatic Multicast Tunneling Additional receivers are served by the AMT Relays; the benefits of IP Multicast are retained by the content owner and all enabled networks in the path. Mcast-Enabled ISP Content Owner Unicast-Only Network AMT Relay AMT Host Gateway AMT Host Gateways Mcast-Enabled Local Provider Mcast Traffic Mcast Join AMT Messages Ucast Stream 128

129 AMT Automatic Multicast Tunneling Creates an Expanding Radius of Incentive to Deploy Multicast Mcast-Enabled ISP Content Owner Unicast-Only Network AMT Relay Enables Multicast Content to a Large (Global) Audience AMT Host Gateway AMT Host Gateways Mcast-Enabled Local Provider Mcast Traffic Mcast Join AMT Messages Ucast Stream 129

130 AMT Automatic Multicast Tunneling Creates an Expanding Radius of Incentive to Deploy Multicast Mcast-Enabled ISP Content Owner Unicast-Only Network AMT Relay Enables Multicast Content to a Large (Global) Audience AMT Host Gateway AMT Host Gateways Mcast-Enabled Local Provider Mcast Traffic Mcast Join AMT Messages Ucast Stream 130

131 AMT Automatic Multicast Tunneling Creates an Expanding Radius of Incentive to Deploy Multicast Mcast-Enabled ISP Content Owner AMT Relay Enables Multicast Content to a Large (Global) Audience AMT Host Gateway AMT Host Gateways Mcast-Enabled Local Provider Mcast Traffic Mcast Join AMT Messages Ucast Stream 131

132 IPv6 Multicast 132

133 IPv4 vs. IPv6 Multicast IP Service IPv4 Solution IPv6 Solution Address Range 32-Bit, Class D 128-Bit (112-Bit Group) Routing Forwarding Protocol-Independent All IGPs and GBP4+ PIM-DM, PIM-SM: ASM, SSM, BiDir Protocol-Independent All IGPs and BGP4+ with v6 Mcast SAFI PIM-SM: ASM, SSM, BiDir Group Management IGMPv1, v2, v3 MLDv1, v2 Domain Control Boundary/Border Scope Identifier Interdomain Source Discovery MSDP Across Independent PIM Domains Single RP Within Globally Shared Domains 133

134 IPv6 Multicast Addresses per RFC bits 8 FF 4 Flags 4 Scope Group-ID F F Flags R P T 8 bits 8 bits Scope Flags = Scope = T or Lifetime, 0 if Permanent, 1 if Temporary P for Unicast-based Assignments R for Embedded RP Others Are Undefined and Must Be Zero 1 = interface-local 2 = link 4 = admin-local 5 = site 8 = organization E = global 0, 3, F = reserved Note: Other scopes (6, 7, 9-D) are unassigned but can be used 134

135 IPv6 Layer 2 Multicast Addressing Mapping RFC FF 4 Flags 4 Scope IPv6 Multicast Address 80 High-Order 112 Bits 32 Low-Order Similar to IPv4: 5 bits are lost 80 Bits Lost (28 significant L3 multicast bits are mapped into 23 L2 MAC bits) More than 1 multicast address (in fact 2^80 ) will map to the same MAC address. For example: FF02:: FF3E:: Pick multicast group addresses that give distinct multicast MAC addresses xx-xx-xx-xx 48 Bits Ethernet MAC Address 135

136 Unicast-based Multicast addresses RFC FF 4 Flags 4 Scope 8 Rsvd 8 Plen 64 Network-Prefix 32 Group-ID RFC 3306 Unicast-based Multicast Addresses Similar to IPv4 GLOP Addressing (233/8 + ASN = 256 group addresses) Solves IPv6 global address allocation problem. Flags = 00PT P = 1, T = 1 => Unicast-based Multicast address Example Content provider s unicast prefix 1234:5678:9abc::/48 Multicast address FF3E:0030:1234:5678:9abc::1 (hex 30 is 48 bits) 136

137 IPv6 Routing for Multicast RPF-based on reachability to v6 source same as with v4 multicast RPF still protocol-independent Static routes, mroutes Unicast RIB: BGP, ISIS, OSPF, EIGRP, RIP, etc. Multiprotocol BGP (MP-BGP) Support for v6 mcast subaddress family 137

138 IPv6 Multicast Tree Building & Forwarding PIM-Sparse Mode (PIM-SM) RFC4601 PIM Source Specific Mode (SSM) RFC3569 SSM overview (v6 SSM needs MLDv2) Unicast, prefix-based multicast addresses ff30::/12 SSM range is ff3x::/96 PIM Bi-Directional Mode (BiDir) draft-ietf-pim-bidir-09.txt 138

139 RP Mapping Mechanisms for IPv6 Static RP assignment BSR Auto-RP no current plans Embedded RP 139

140 Embedded RP Addressing Multicast Address with Embedded RP address RFC FF 4 Flags 4 Scope 4 Rsvd 4 RPadr Proposed new multicast address type 8 Plen Uses unicast-based multicast addresses (RFC 3306) 64 Network-Prefix 32 Group-ID RP address is embedded in multicast address Flag bits = 0RPT R = 1, P = 1, T = 1 Embedded RP address Network-Prefix::RPadr = RP address For each unicast prefix you own, you now also own: 16 RPs for each of the 16 multicast scopes (256 total) with 2^32 multicast groups assigned to each RP (2^40 total) 140

141 Embedded RP Addressing Example Multicast Address with Embedded RP address 8 FF 4 Flags 4 Scope 4 Rsvd 4 RPadr 8 Plen 64 Network-Prefix 32 Group-ID FF76:0130:1234:5678:9abc:: :5678:9abc::1 Resulting RP address 141

142 Multicast Listener Discover MLD MLD is equivalent to IGMP in IPv4 MLD messages are transported over ICMPv6 Version number confusion MLDv1 corresponds to IGMPv2 RFC 2710 MLDv2 corresponds to IGMPv3, needed for SSM RFC 3810 MLD snooping draft-ietf-magma-snoop-12.txt 142

143 Conclusion 143

144 Now You Know Why multicast? Multicast fundamentals PIM protocols RP choices Multicast at Layer 2 Inter-domain IP multicast IPv6 Multicast bits 144

145 Participate in the My Favorite Speaker Contest Promote Your Favorite Speaker and You Could be a Winner Promote your favorite speaker through Twitter and you could win $200 of Cisco Press products (@CiscoPress) Send a tweet and include Your favorite speaker s Twitter handle <@Mr_Multicast> Two hashtags: #CLUS #MyFavoriteSpeaker You can submit an entry for more than one of your favorite speakers Don t forget to View the official rules at 145

146 Complete Your Online Session Evaluation Give us your feedback to be entered into a Daily Survey Drawing. A daily winner will receive a $750 Amazon gift card. Complete your session surveys though the Cisco Live mobile app or your computer on Cisco Live Connect. Don t forget: Cisco Live sessions will be available for viewing on-demand after the event at CiscoLive.com/Online 146

147 Continue Your Education Demos in the Cisco Campus Walk-in Self-Paced Labs Table Topics Meet the Engineer 1:1 meetings 147

148 Thank you 148

149 149