idnits 2.17.1 draft-ietf-idmr-pim-dm-spec-04.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** Cannot find the required boilerplate sections (Copyright, IPR, etc.) in this document. Expected boilerplate is as follows today (2024-04-26) according to https://trustee.ietf.org/license-info : IETF Trust Legal Provisions of 28-dec-2009, Section 6.a: This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79. IETF Trust Legal Provisions of 28-dec-2009, Section 6.b(i), paragraph 2: Copyright (c) 2024 IETF Trust and the persons identified as the document authors. All rights reserved. IETF Trust Legal Provisions of 28-dec-2009, Section 6.b(i), paragraph 3: This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- ** Missing expiration date. The document expiration date should appear on the first and last page. ** The document seems to lack a 1id_guidelines paragraph about Internet-Drafts being working documents. ** The document seems to lack a 1id_guidelines paragraph about 6 months document validity. ** The document seems to lack a 1id_guidelines paragraph about the list of current Internet-Drafts. ** The document seems to lack a 1id_guidelines paragraph about the list of Shadow Directories. ** Bad filename characters: the document name given in the document, 'draft-ietf-idmr-PIM-DM-spec-04', contains other characters than digits, lowercase letters and dash. == Mismatching filename: the document gives the document name as 'draft-ietf-idmr-PIM-DM-spec-04', but the file name used is 'draft-ietf-idmr-pim-dm-spec-04' == No 'Intended status' indicated for this document; assuming Proposed Standard == The page length should not exceed 58 lines per page, but there was 8 longer pages, the longest (page 3) being 120 lines Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack an Abstract section. ** The document seems to lack a Security Considerations section. ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) ** The document seems to lack an Authors' Addresses Section. ** There are 194 instances of weird spacing in the document. Is it really formatted ragged-right, rather than justified? ** There are 10 instances of too long lines in the document, the longest one being 8 characters in excess of 72. ** There are 10 instances of lines with control characters in the document. == There are 3 instances of lines with multicast IPv4 addresses in the document. If these are generic example addresses, they should be changed to use the 233.252.0.x range defined in RFC 5771 Miscellaneous warnings: ---------------------------------------------------------------------------- == Line 14 has weird spacing: '... Drafts are ...' == Line 15 has weird spacing: '...cuments of t...' == Line 16 has weird spacing: '...ups may also ...' == Line 20 has weird spacing: '... Drafts may ...' == Line 21 has weird spacing: '...iate to use ...' == (189 more instances...) == Couldn't figure out when the document was first submitted -- there may comments or warnings related to the use of a disclaimer for pre-RFC5378 work that could not be issued because of this. Please check the Legal Provisions document at https://trustee.ietf.org/license-info to determine if you need the pre-RFC5378 disclaimer. -- The document date (Sept 12, 1996) is 10363 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) -- Missing reference section? 'Deering95' on line 497 looks like a reference -- Missing reference section? 'Deering91' on line 490 looks like a reference -- Missing reference section? 'DVMRP' on line 494 looks like a reference -- Missing reference section? 'Deering94b' on line 501 looks like a reference -- Missing reference section? 'RFC1112' on line 505 looks like a reference Summary: 14 errors (**), 0 flaws (~~), 11 warnings (==), 6 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 1 Network Working Group Deborah Estrin (USC) 2 Internet Draft Dino Farinacci (CISCO) 3 Ahmed Helmy (USC) 4 Van Jacobson (LBL) 5 Liming Wei (USC) 7 draft-ietf-idmr-PIM-DM-spec-04.txt Sept 12, 1996 9 Protocol Independent Multicast-Dense Mode (PIM-DM): Protocol 10 Specification 12 Status of This Memo 14 This document is an Internet Draft. Internet Drafts are working 15 documents of the Internet Engineering Task Force (IETF), its Areas, 16 and its Working Groups. (Note that other groups may also distribute 17 working documents as Internet Drafts). 19 Internet Drafts are draft documents valid for a maximum of six 20 months. Internet Drafts may be updated, replaced, or obsoleted by 21 other documents at any time. It is not appropriate to use Internet 22 Drafts as reference material or to cite them other than as a 23 ``working'' draft'' or ``work in progress.'' 25 Please check the I-D abstract listing contained in each Internet 26 Draft directory to learn the current status of this or any other 27 Internet Draft. 29 1 Introduction 31 This specification defines a multicast routing algorithm for 32 multicast groups that are densely distributed across an internet. The 33 protocol is unicast routing protocol independent. It is based on the 34 PIM sparse-mode [Deering95] and employs the same packet formats. This 35 protocol is called dense-mode PIM. The design is based largely on 36 foundational work by Deering [Deering91]. 38 2 PIM-DM Protocol Overview 40 Dense-mode PIM uses Reverse Path Multicasting (RPM). RPM is a 41 technique in which a multicast datagram is forwarded if the receiving 42 interface is one used to forward unicast datagrams to the source of 43 the datagram. The multicast datagram is then forwarded out all other 44 interfaces. Dense-mode PIM builds source-based acyclic trees. 46 Dense-mode PIM is data driven, whereby it is assumed that all 47 downstream systems want to receive multicast datagrams. For densely 48 populated groups this is optimal. If some areas of the network do not 49 have group members, dense-mode PIM will prune branches of the 50 source-based tree. When group members leave the group, branches will 51 also be pruned. 53 Unlike DVMRP [DVMRP] packets are forwarded on all outgoing interfaces 54 (except the incoming) until pruning and truncation occurs. DVMRP 55 makes use of parent-child data to reduce the number of outgoing 56 interfaces used before pruning. In both protocols, once truncation 57 occurs pruning state is maintained and packets are only forwarded 58 onto outgoing interfaces that in fact reach downstream members. 60 We chose to accept additional overhead in favor of reduced dependency 61 on the unicast routing protocol, and reduced overall protocol 62 complexity. 64 Dense-mode PIM differs from sparse-mode PIM in two essential points: 65 1) there are no periodic joins transmitted, only explicit triggered 66 grafts/prunes, and 2) there is no Rendezvous Point (RP). 68 3 Background 70 Reverse Path Broadcasting (RPB) is different from RPF because 71 duplicate packets are avoided in the RPB that are sent in RPF. In 72 general, the number of duplicates sent on a link can be as high as 73 the number of routers directly connected to that link. 75 Estrin,Farinacci,Helmy,Jacobson,Wei [Page 2]^L 76 Reverse Path Multicasting (RPM) is different from RPF or RPB because 77 pruning information is propagated upstream. Leaf routers must know 78 that they are leaf routers so that in response to no IGMP reports for 79 a group, those leaf routers know to initiate the prune process. 81 In DVMRP there are routing protocol dependencies for a) building a 82 parent-child database so that duplicate packets can be eliminated, b) 83 eliminating duplicate packets on multi-access LANs, and c) sending 84 "split horizon with poison reverse" information to detect that a 85 router is not a leaf router (if a router does not receive any poison 86 reverse messages from other routers on a multi-access LAN then that 87 router acts as a leaf router for that LAN and knows to prune if there 88 are not IGMP reports on that LAN for a group G). 90 Dense-mode PIM will accept some duplicate packets in order to avoid 91 being routing protocol dependent and avoid building a child parent 92 database. 94 We introduce a simple prune mechanism for reducing duplicates on 95 multi-access LANs. We introduce a simple graft mechanism to reduce 96 join latency on previously pruned branches of a source-based 97 multicast tree. 99 We introduce an alternative leaf-router detection mechanism that does 100 not rely on a specific unicast routing protocol mechanism such as 101 split horizon with poison reverse. 103 These mechanisms are described below. 105 4 Protocol Description 107 4.1 Leaf network detection 109 In DVMRP poison reverse information tells a router that other routers 110 on the shared LAN use the LAN as their incoming interface. As a 111 result, even if the DR for that LAN does not hear any IGMP Reports 112 for a group, the DR will know to continue to forward multicast data 113 packets to that group, and NOT to send a prune message to its 114 upstream neighbor. 116 Since dense-mode PIM does not rely on any unicast routing protocol 117 mechanisms, this problem is solved by using prune messages sent 118 upstream on a LAN. If a downstream router on a LAN determines that it 119 has no more downstream members for a group, then it can multicast a 120 prune message on the LAN. 122 A leaf router detects that there are no members downstream when it is 123 the only router on a network and there are no IGMP Host-Report 124 messages received from hosts. It determines there are no other 125 routers by not receiving PIM Router-Hello messages. 127 When a prune message is sent on an upstream LAN, it is data link 128 multicast and IP addressed to the all routers group address 129 224.0.0.13. The router to process the prune will be indicated by 130 inserting its address in the "Address" field of the message. The 131 address is obtained by an RPF lookup from the unicast routing table. 132 When the prune message is sent, the expected upstream router will 133 schedule a deletion request of the LAN from its outgoing interfaces 134 for the (S,G) entry from the prune list. The suggested delay time 135 before deletion should be greater than 3 seconds. 137 Note the special case for equal-cost paths. When an upstream router 138 is chosen by an RPF lookup there may be equal-cost paths to reach the 139 source. The higher IP addressed system is always chosen. If the 140 unicast routing protocol does not store all available equal-cost 141 paths in the routing table, the "Address" field may contain the 142 address of the wrong upstream router. To avoid this situation, the 143 "Address" field may optionally be set to 0.0.0.0 which means that all 144 upstream routers (the ones that have the LAN as an outgoing interface 145 for the (S,G) entry) may process the packet. 147 Other routers on the LAN will hear the prune message and respond with 148 a join if they still expect multicast datagrams from the expected 149 upstream router. The PIM-Join message is data link multicast and IP 150 addressed to the all routers group address 224.0.0.13. The router to 151 process the join will be indicated by inserting its address in the 152 "Address" field of the message. The address is determined by an RPF 153 lookup from the unicast routing table. When the expected router 154 receives the join message, it will cancel the deletion request. 156 Routers will randomly generate a join message delay timer. If a join 157 is heard from another router before a router sends its own, it will 158 cancel sending its own join. This will reduce traffic on the LAN. The 159 suggested join delay timer should be from 1 to 3 seconds. 161 If the expected upstream router does not receive any PIM-Join 162 messages before the schedule time for the deletion request expires, 163 it deletes the outgoing LAN interface from the (S,G) multicast 164 forwarding entry. 166 Note that if the join message is lost, the deletion will occur and there 167 will be a no data delivery for the amount of time the interface remains 168 in Prune state. To reduce the probability of this occurrence, a router that 169 overrides a prune may send multiple joins back-to-back or have a small 170 delay between successive joins. 172 If an (S,G) entry contains an empty outgoing interface list, a prune 173 is sent upstream. Prune information is flushed periodically. This (or 174 a loss of state) causes the packets to be sent in RPF mode again 175 which in turn triggers prune messages. 177 4.2 New members joining an existing group 179 If a router is directly connected to a host that wants to become a 180 member of a group, the router may optionally, send a PIM-Graft 181 message towards known sources. This allows join latency to be reduced 182 below that indicated by the relatively large timeout value suggested 183 for prune information. 185 If a receiving router has state for group G, it adds the interface on 186 which the IGMP Report or PIM-Graft was received for all known (S,G). 187 If the (S,G) entry was a negative cache entry, the router sends a 188 PIM-Graft message upstream towards S. 190 If routers have no group state, they do nothing since dense-mode PIM 191 will deliver a multicast datagram to all interfaces when creating 192 state for a group. 194 Any routers receiving the PIM-Graft message, uses the received 195 interface as an incoming interface for any (S,G) entry, will not add 196 the interface to the outgoing interface list. 198 The PIM-Graft message is the only PIM message that uses a positive 199 acknowledgment strategy. Senders of PIM-Graft messages unicast them 200 to their upstream RPF neighbors. The neighbor processes each (S,G) 201 and immediately acknowledges each (S,G) in a PIM-GraftAck message. 202 This is relatively easy, since the receiver simply changes the IGMP 203 code from Graft to Graft-Ack and unicasts the original packet back to 204 the source. The sender periodically retransmits the PIM-Graft message 205 for any (S,G) that has not been acknowledged. Note that the sender 206 need not keep a retransmission list for each neighbor since PIM- 207 Grafts are only sent to the RPF neighbor. Only the (S,G) entry needs 208 to be tagged for retransmission. 210 4.3 Protocol Scenario 212 A multicast datagram is sent by a source host. If a receiving router 213 has no forwarding cache state for the source sending to group G, it 214 creates an (S,G) entry. The incoming interface for (S,G) is 215 determined by doing an RPF lookup in the unicast routing table. The 216 (S,G) outgoing interface list contains dense-mode configured 217 interfaces that have PIM routers present or host members for group G. 219 A PIM-Prune message is triggered when an (S,G) entry is built with an 220 empty outgoing interface list. This type of entry is called a 221 negative cache entry. This can occur when a leaf router has no local 222 members for group G or a prune message was received from a downstream 223 router which causes the outgoing interface list to become NULL. PIM- 224 Prune messages are never sent on LANs in response to a received 225 multicast packet that is associated with a negative cache entry. 227 PIM-Prune messages received on a point to point link are not delayed 228 before processing as they are in the LAN procedure. If the prune is 229 received on an interface that is in the outgoing interface list, it 230 is deleted immediately. Otherwise the prune is ignored. 232 When a multicast datagram is received on the incorrect LAN interface 233 (i.e. not the RPF interface) the packet is silently discarded. If it 234 is received on an incorrect point-to-point interface, Prunes may be 235 sent in a rate-limited fashion. Prunes may also be rate-limited on 236 point-to-point interfaces when a multicast datagram is received for a 237 negative cache entry. 239 4.4 Designated Router election 241 The dense-mode PIM designated router (DR) election uses the same 242 procedure as in sparse-mode PIM. A DR is necessary for each multi- 243 access LAN so a single router sends IGMP Host-Query messages to 244 solicit host group membership. 246 Each PIM router connected to a multi-access LAN should transmit PIM 247 Router-Hello messages every 30 seconds onto the LAN to support DR 248 election. The highest addressed router becomes the DR. The discovered 249 PIM routers should be timed out after 90 seconds. If the DR goes 250 down, a new DR is elected. 252 DR election is only necessary on multi-access networks. It is not 253 required that PIM Hello messages be sent on point-to-point links. 255 4.5 Parallel paths to a source 257 Two or more routers may receive the same multicast datagram that was 258 replicated upstream. In particular, if two routers have equal cost 259 paths to a source and are connected on a common multi-access network, 260 duplicate datagrams will travel downstream onto the LAN. Dense-mode 261 PIM will detect such a situation and will not let it persist. 263 If a router receives a multicast datagram on a multi-access LAN from 264 a source whose corresponding (S,G) outgoing interface list includes 265 the received interface, the packet must be a duplicate. In this case 266 a single forwarder must be elected. Using PIM Assert messages 267 addressed to 224.0.0.13 on the LAN, upstream routers can decide which 268 one becomes the forwarder. Downstream routers listen to the Asserts 269 so they know which one was elected (i.e. typically this is the same 270 as the downstream router's RPF neighbor but there are circumstances 271 when using different unicast protocols where this might not be the 272 case). 274 The upstream router elected is the one that has the shortest distance 275 to the source. Therefore, when a packet is received on an outgoing 276 interface a router will send an Assert packet on the LAN indicating 277 what metric it uses to reach the source of the data packet. The 278 router with the smallest numerical metric will become the forwarder. 279 All other upstream routers will delete the interface from their 280 outgoing interface list. The downstream routers also do the 281 comparison in case the forwarder is different than the RPF neighbor. 282 This is important so downstream routers send subsequent Prunes or 283 Grafts to the correct neighbor. 285 Associated with the metric is a metric preference value. This is 286 provided to deal with the case where the upstream routers may run 287 different unicast routing protocols. The numerically smaller metric 288 preference is always preferred. The metric preference should be 289 treated as the high-order part of an Assert metric comparison. 290 Therefore, a metric value can be compared with another metric value 291 provided both metric preferences are the same. A metric preference 292 can be assigned per unicast routing protocol and needs to be 293 consistent for all routers on the LAN. 295 The following Assert rules are provided: 297 Multicast packet received on outgoing interface: 299 1 Do unicast routing table lookup on source IP address from 300 data packet. 302 2 Send Assert on interface for source IP address from data 303 packet, include metric preference of routing protocol and 304 metric from routing table lookup. 306 3 If route is not found, Use metric preference of 0x7fffffff 307 and metric 0xffffffff. 309 Asserts received on outgoing interface: 311 1 Compare metric received in Assert with the one you would 312 have advertised in an Assert. If the value in the Assert is 313 less than your value, prune the interface. If the value is 314 the same, compare IP addresses, if your address is less 315 than the Assert sender, prune the interface. 317 2 If you have won the election and there are directly 318 connected members on the LAN, keep the interface in your 319 outgoing interface list. You are the forwarder for the LAN. 321 3 If you have won the election but there are no directly 322 connected members on the LAN, schedule to prune the 323 interface. The LAN might be a stub LAN with no members (and 324 no downstream routers). If no subsequent Joins are 325 received, delete the interface from the outgoing interface 326 list. Otherwise keep the interface in your outgoing 327 interface. You are the forwarder for the LAN. 329 Asserts received on incoming interface: 331 1 Downstream routers will select the upstream router with the 332 smallest metric as their RPF neighbor. If two metrics are 333 the same, the highest IP address is chosen to break the 334 tie. 336 2 If the downstream routers have downstream members, they 337 must schedule a join to inform the upstream router packets 338 should be forwarded on the LAN. This will cause the 339 upstream forwarder to cancel its delayed pruning of the 340 interface. 342 4.6 Timing out multicast forwarding entries 344 Each (S,G) entry has timers associated with it. During this time 345 source-based tree state is kept in the network. 347 There should be multiple timers set. One for the multicast 348 routing entry itself and one for each interface in the outgoing 349 interface list. The outgoing interface stays active in the list 350 as long as there is multicast traffic for the entry or there is 351 an explicit Graft received on the interface. If neither occurs 352 the interface will be deleted from the list after 3 minutes, by 353 default. 355 Once all interfaces in the outgoing interface list are not 356 active, a timer should be set for the (S,G) entry. During this 357 time the entry is known as a negative state entry at which a 358 prune is triggered. Once the (S,G) entry times out, it can be 359 recreated when the next multicast packet or join arrives. 361 4.7 Source address aggregation and Pruning 363 An (S,G) entry in the multicast routing table will contain a 364 source address to be as specific as necessary depending where 365 the router is in relation to a source. Close to the source, this 366 will typically be a subnet number. Far from the source, it may 367 be a network number or a supernet route. Prunes sent may be 368 rather ineffective if the source being pruned is not specific 369 enough. 371 For example, initially a multicast datagram may be flooded 372 throughout an Autonomous System (AS). Within the AS, there is 373 complete subnet information in the unicast routing tables of all 374 routers. Once the datagram exits the AS, it is likely there are 375 routers that don't have subnet information. If these routers 376 send Prunes for aggregate sources, routers close to the source 377 will not know where to reach the source since they have more 378 specific information than what was provided in the Prune 379 message. This results in traffic being sent further on a branch 380 of the multicast tree than necessary. 382 The problem is fixed by sending source host specific prunes. 383 However, to maintain proper scaling of routing information, each 384 router along the path performs longest match lookups for the 385 source specified in the Prune message. Therefore, they keep the 386 level of aggregation that best suits thier position to the 387 source in the topology. 389 A source host specific Prune is encoded by copying the 390 instigating source IP address from the multicast datagram into 391 the PIM message using a mask length of 32. 393 5 Packet Formats 395 This section describes the details of the packet formats for PIM control 396 messages. 398 All PIM control messages have protocol number 103. 400 Basically, PIM messages are either unicast (e.g. Registers and 401 Register-Stop), or multicast hop-by-hop to `ALL-PIM-ROUTERS' group 402 `224.0.0.13' (e.g. Join/Prune, Asserts, etc.). 404 0 1 2 3 405 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 406 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 407 |PIM Ver| Type | Addr length | Checksum | 408 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 410 PIM Ver 411 PIM Version number is 2. 413 Type 414 Types for specific PIM messages. 415 PIM Types are: 417 0 = Hello 418 1 = Register 419 2 = Register-Stop 420 3 = Join/Prune 421 4 = Bootstrap 422 5 = Assert 423 6 = Graft 424 7 = Graft-Ack 425 8 = Candidate-RP-Advertisement 426 Addr length 427 Address length in bytes. Throughout this section this would 428 indicate the number of bytes in the Address field of an address, 429 including unicast and group addresses. 431 Checksum 432 The checksum is the 16-bit one's complement of the one's 433 complement sum of the entire PIM message, 434 (excluding the data portion in the Register message). 435 For computing the checksum, the checksum field is zeroed. 437 { For all the packet format details, refer to the PIM sparse- 438 mode specification.} 439 5.1 PIM-Hello Message 441 It is sent periodically by PIM routers on all interfaces. 443 5.2 PIM-SM-Register Message 445 Used in sparse-mode. Refer to PIM sparse-mode specification. 447 5.3 PIM-SM-Register-Stop Message 449 Used in sparse-mode. Refer to PIM sparse-mode specification. 451 5.4 Join/Prune Message 453 It is sent by routers towards upstream sources. A join creates 454 forwarding state and a prune destroys forwarding state. Joins 455 are sent to build source specific trees. Prunes are sent to 456 prune source trees when members leave groups as well as sources 457 that do not use the shared tree. 459 5.5 PIM-SM-Bootstrap Message 461 Used in sparse-mode. Refer to PIM sparse-mode specification. 463 5.6 PIM-Assert Message 465 The PIM-Assert message is sent when a multicast data packet is 466 received on an outgoing interface corresponding to the (S,G) or 467 (*,G) associated with the source. 469 5.7 PIM-Graft Message 471 This message is sent by a downstream router to a neighboring 472 upstream router to reinstate a previously pruned branch of a 473 source tree. This is done for dense-mode groups only. The format 474 is the same as a Join/Prune message. 476 5.8 PIM-Graft-Ack Message 478 Sent in response to a received Graft message. The Graft-Ack is 479 only sent if the interface in which the Graft was received is 480 not the incoming interface for the respective (S,G). This is 481 done for dense-mode groups only. The format is the same as 482 Join/Prune message. 484 5.9 Candidate-RP-Advertisement 486 Used in sparse-mode. Refer to PIM sparse-mode specification. 488 6 References 490 [Deering91] S.E. Deering. Multicast Routing in a Datagram 491 Internetwork. PhD thesis, Electrical Engineering Dept., Stanford 492 University, December 1991. 494 [DVMRP] RFC 1075, Distance Vector Multicast Routing Protocol. 495 Waitzman, D., Partridge, C., Deering, S.E, November 1988 497 [Deering95] Protocol Independent Multicast Sparse-Mode (PIM-SM): 498 Protocol Specification. S. Deering, D. Estrin, D. Farinacci, V. 499 Jacobson, G. Liu, L. Wei, P. Sharma, A. Helmy, September 1995 501 [Deering94b] An Architecture for Wide-Area Multicast Routing, S. 502 Deering, D. Estrin, D. Farinacci, V. Jacobson, G. Liu,L. Wei, 503 USC Technical Report, available from authors, Feburary 1994. 505 [RFC1112] Host Extensions for IP Multicasting, Network Working 506 Group, RFC 1112, S. Deering, August 1989 508 References