idnits 2.17.1 draft-ietf-pim-anycast-rp-05.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** It looks like you're using RFC 3978 boilerplate. You should update this to the boilerplate described in the IETF Trust License Policy document (see https://trustee.ietf.org/license-info), which is required now. -- Found old boilerplate from RFC 3978, Section 5.1 on line 15. -- Found old boilerplate from RFC 3978, Section 5.5 on line 44. ** This document has an original RFC 3978 Section 5.4 Copyright Line, instead of the newer IETF Trust Copyright according to RFC 4748. ** This document has an original RFC 3978 Section 5.5 Disclaimer, instead of the newer disclaimer which includes the IETF Trust according to RFC 4748. ** The document seems to lack an RFC 3979 Section 5, para. 1 IPR Disclosure Acknowledgement. ** The document seems to lack an RFC 3979 Section 5, para. 2 IPR Disclosure Acknowledgement. ** The document seems to lack an RFC 3979 Section 5, para. 3 IPR Disclosure Invitation. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- == No 'Intended status' indicated for this document; assuming Proposed Standard == The page length should not exceed 58 lines per page, but there was 11 longer pages, the longest (page 2) being 60 lines == It seems as if not all pages are separated by form feeds - found 0 form feeds but 12 pages Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the RFC 3978 Section 5.4 Copyright Line does not match the current year -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (January 30, 2006) is 6661 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) == Outdated reference: A later version (-12) exists of draft-ietf-pim-sm-v2-new-11 Summary: 6 errors (**), 0 flaws (~~), 5 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group Dino Farinacci 3 INTERNET-DRAFT Yiqun Cai 4 Expiration Date: July 2006 cisco Systems 5 January 30, 2006 7 Anycast-RP using PIM 8 10 Status of this Memo 12 By submitting this Internet-Draft, each author represents that any 13 applicable patent or other IPR claims of which he or she is aware 14 have been or will be disclosed, and any of which he or she becomes 15 aware will be disclosed, in accordance with Section 6 of BCP 79. 17 Internet-Drafts are working documents of the Internet Engineering 18 Task Force (IETF), its areas, and its working groups. Note that other 19 groups may also distribute working documents as Internet-Drafts. 21 Internet-Drafts are draft documents valid for a maximum of six months 22 and may be updated, replaced, or obsoleted by other documents at any 23 time. It is inappropriate to use Internet-Drafts as reference 24 material or to cite them other than as "work in progress." 26 The list of current Internet-Drafts can be accessed at 27 http://www.ietf.org/1id-abstracts.html 29 The list of Internet-Draft Shadow Directories can be accessed at 30 http://www.ietf.org/shadow.html. 32 Copyright Notice 34 Copyright (C) The Internet Society (2006). This document is subject 35 to the rights, licenses and restrictions contained in BCP 78, and 36 except as set forth therein, the authors retain all their rights. 38 This document and the information contained herein are provided on an 39 "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS 40 OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY AND THE INTERNET 41 ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED, 42 INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE 43 INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED 44 WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. 46 Abstract 48 This specification allows Anycast-RP (Rendezvous Point) to be used 49 inside a domain that runs Protocol Independent Multicast (PIM) only. 50 There are no other multicast protocols required to support Anycast- 51 RP, such as MSDP, which has been used traditionally to solve this 52 problem. 54 1.0 Introduction 56 Anycast-RP as described in [I1] is a mechanism ISP-based backbones 57 have used to get fast convergence when a PIM Rendezvous Point (RP) 58 router fails. To allow receivers and sources to Rendezvous to the 59 closest RP, the packets from a source need to get to all RPs to find 60 joined receivers. 62 This notion of receivers finding sources is the fundamental problem 63 of source discovery which MSDP was intended to solve. However, if one 64 would like to retain the Anycast-RP benefits from [I1] with less 65 protocol machinery, removing MSDP from the solution space is an 66 option. 68 This memo extends the Register mechanism in PIM so Anycast-RP 69 functionality can be retained without using MSDP. 71 1.1 Terminology 73 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 74 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 75 document are to be interpreted as described in [N2]. 77 2.0 Requirements 79 o An IP address is chosen to use as the RP address. This address is 80 statically configured, or distributed using a dynamic protocol, to 81 all PIM routers throughout the domain. 83 o A set of routers in the domain are chosen to act as RPs for this 84 RP address. These routers are called the Anycast-RP set. 86 o Each router in the Anycast-RP set is configured with a loopback 87 interface using the RP address. 89 o Each router in the Anycast-RP set also needs a separate IP address, 90 to be used for communication between the RPs. 92 o The RP address, or a prefix that covers the RP address, is injected 93 into the unicast routing system inside of the domain. 95 o Each router in the Anycast-RP set is configured with the addresses 96 of all other routers in the Anycast-RP set. This must be 97 consistently configured in all RPs in the set. 99 3.0 Mechanism 101 The following diagram illustrates a domain using 3 RPs where 102 receivers are joining to the closest RP according to where unicast 103 routing metrics take them and 2 sources sending packets to their 104 respective RPs. 106 The rules described in this section do not override the rules in 107 [N1]. They are intended to blend with the rules in [N1]. If there is 108 any question on the interpretation, precedent is given to [N1]. 110 S1-----RP1 RP2 RP3------S3 111 / \ | 112 / \ | 113 R1 R1' R2 115 Assume the above scenario is completely connected where R1, R1', and 116 R2 are receivers for a group, and S1 and S3 send to that group. 117 Assume RP1, RP2 and RP3 are all assigned the same IP address which is 118 used as the Anycast-RP address (let's say the IP address is RPA). 120 Note, the address used for the RP address in the domain (the 121 Anycast-RP address) needs to be different than the addresses used by 122 the Ancyast-RP routers to communicate with each other. 124 The following procedure is used when S1 starts sourcing traffic: 126 o S1 sends a multicast packet. 128 o The DR directly attached to S1 will form a PIM Register 129 message to send to the Anycast-RP address (RPA). The unicast 130 routing system will deliver the PIM Register message to the 131 nearest RP, in this case RP1. 133 o RP1 will receive the PIM Register message, decapsulate it, send the 134 packet down the shared-tree to get the packet to receivers R1 and 135 R1'. 137 o RP1 is configured with RP2 and RP3's IP address. Since the 138 Register message did not come from one of the RPs in the 139 anycast-RP set, RP1 assumes the packet came from a DR. If the 140 Register is not addressed to the Anycast-RP address, an error 141 has occurred and it should be rate-limited logged. 143 o RP1 will then send a copy of the Register message from S1's 144 DR to both RP2 and RP3. RP1 will use its own IP address as 145 the source address for the PIM Register message. 147 o RP1 MAY join back to the source-tree by triggering a (S1,G) Join 148 message toward S1. However, RP1 MUST create (S1,G) state. 150 o RP1 sends a Register-Stop back to the DR. If, for some reason, 151 the Register messages to RP2 and RP3 are lost, then when the 152 Register suppression timer expires in the DR, it will resend 153 Registers to allow another chance for all RPs in the Anycast-RP 154 set to obtain the (S,G) state. 156 o RP2 receives the Register message from RP1, decapsulates it, and 157 also sends the packet down the shared-tree to get the packet to 158 receiver R2. 160 o RP2 sends a Register-Stop back to the RP1. RP2 MAY wait to send 161 the Register-Stop if it decides to join the source-tree. RP2 162 should wait until it has received data from the source on the 163 source-tree before sending the Register-Stop. If RP2 decides to 164 wait, the Register-Stop will be sent when the next Register is 165 received. If RP2 decides not to wait, the Register-Stop is sent 166 now. 168 o RP2 MAY join back to the source-tree by triggering a (S1,G) Join 169 message toward S1. However, RP2 MUST create (S1,G) state. 171 o RP3 receives the Register message from RP1, decapsulates it, but 172 since there are no receivers joined for the group, it can discard 173 the packet. 175 o RP3 sends a Register-Stop back to the RP1. 177 o RP3 creates (S1,G) state so when a receiver joins after S1 starts 178 sending, RP3 can join quickly to the source-tree for S1. 180 o RP1 processes the Register-Stop from each of RP2 and RP3. There 181 is no specific action taken when processing Register-Stop messages. 183 The procedure for S3 sending follows the same as above but it is RP3 184 which sends a copy of the Register originated by S3's DR to RP1 and 185 RP2. Therefore, this example shows how sources anywhere in the 186 domain, associated with different RPs, can reach all receivers, also 187 associated with different RPs, in the same domain. 189 4.0 Observations and Guidelines about this Proposal 191 o An RP will send a copy of a Register only if the Register is 192 received from an IP address not in the Anycast-RP list (i.e. the 193 Register came from a DR and not another RP). An implementation 194 SHOULD safeguard against inconsistently configured anycast-RP sets 195 in each RP by copying the TTL from a Register message to the 196 Register messages it copies and sends to other RPs. 198 o Each DR that PIM registers for a source will send the message to 199 the anycast-RP address (which results in the packet getting to the 200 closest physical RP). Therefore there are no changes to the DR 201 logic. 203 o Packets flow to all receivers no matter what RP they have joined 204 to. 206 o The source gets Registered to a single RP by the DR. It's the 207 responsibility of the RP that receives the PIM Register 208 messages from the DR (the closest RP to the DR based on routing 209 metrics) to get the packet to all other RPs in the Anycast-RP 210 set. 212 o Logic is changed only in the RPs. The logic change is for 213 sending copies of Register messages. Register-Stop processing is 214 unchanged. However, an implementation MAY suppress sending 215 Register-Stop messages in response to a Register received from 216 an RP. 218 o The rate-limiting of Register and Register-Stop messages are done 219 end-to-end. That is from DR -> RP1 -> {RP2 and RP3}. There is no 220 need for specific rate-limiting logic between the RPs. 222 o When topology changes occur, the existing source-tree adjusts 223 as it does today according to [N1]. The existing shared-trees, 224 as well, adjust as it does today according to [N1]. 226 o Physical RP changes are as fast as unicast route convergence. 227 Retaining the benefit of [I1]. 229 o An RP that doesn't support this specification can be mixed with 230 RPs that do support this specification. However, the non-supporter 231 RPs should not have sources registering to it but may have 232 receivers joined to it. 234 o If Null Registers are sent (Registers with an IP header and no IP 235 payload), they MUST be replicated to all of the RPs in the Anycast- 236 RP set so that source state remains alive for active sources. 238 o The number of RPs in the Anycast-RP set should remain small so the 239 amount of non-native replication is kept to a minimum. 241 o Since the RP, who receives a Register from the DR, will send copies 242 of the Register to the other RPs at the same time it sends a 243 Register-Stop to the DR, there could be packet loss and lost state 244 in the other RPs until the time the DR sends Register messages 245 again. 247 5.0 Interaction with MSDP running in an Anycast-PIM Router 249 The objective of this Anycast-PIM proposal is to remove the 250 dependence on using MSDP. This can be achieved by removing MSDP 251 peering between the Anycast RPs. However, to advertise internal 252 sources to routers outside of a PIM routing domain and to learn 253 external sources from other routing domains, MSDP may still be 254 required. 256 5.1 Anycast-PIM Stub Domain Functionality 258 In this capacity, when there are internal sources that need to be 259 advertised externally, an Anycast-RP which receives a Register 260 message, either from a DR or an Anycast-RP, should process it as 261 described in this specification as well as how to process a Register 262 message as described in [N1]. That means an SA for the same internal 263 source could be originated by multiple Anycast-RPs doing the MSDP 264 peering. There is nothing inherently wrong with this other than the 265 source is being advertised into the MSDP infrastructure from multiple 266 places from the source domain. However, if this is not desirable, 267 configuration of one or more (rather than all) Anycast-RP MSDP 268 routers would allow only those routers to originate SAs for the 269 internal source. And in some situations, there is a good possibility 270 not all Anycast-RPs in the set will have MSDP peering sessions so 271 this issue can be mitigated to a certain extent. 273 From an Anycast-RP perspective, a source should be considered 274 internal to a domain, when it is discovered by an Anycast-RP through 275 a received Register message. Regardless, if the Register message was 276 sent by a DR, another Anycast-RP member, or the router itself. 278 For learning sources external to a domain, the MSDP SA messages could 279 arrive at multiple MSDP-peering Anycast-RPs. The rules for processing 280 an SA, according to [I1], should be followed. That is, if G is joined 281 in the domain, an (S,G) join is sent towards the source. And if data 282 accompanies the SA, each Anycast-PIM RP doing MSDP peering will 283 forward the data down each of their respective shared-trees. 285 The above assumes each Anycast-RP has external MSDP peering 286 connections. If this is not the case, the Anycast-PIM routers with 287 the MSDP peering connections would follow the same procedure as if a 288 Data-Register or Null-Register was received from either a DR or 289 another Anycast-RP. That is, they would send Registers to the other 290 members of the Anycast-RP set. 292 If there is a mix of Anycast-RPs that do and do not have external 293 MSDP peering connections, then the ones that do must be configured 294 with the set that do not. So Register messages are sent only to the 295 members of the Anycast-RP set that do not have external MSDP peering 296 connections. 298 The amount of Register traffic generated by this MSDP-peering RP 299 would be equal to the number of active sources external to the 300 domain. The Source-Active state would have to be conveyed to all 301 other RPs in the Anycast-RP set since the MSDP-peering RP would not 302 know about the group membership associated with the other RPs. To 303 avoid this periodic control traffic, it is recommended that all 304 Anycast-RPs be configured with external MSDP peering sessions so no 305 RP in the Anycast-RP set will have to originate Register messages on 306 behalf of external sources. 308 5.2 Anycast-PIM Transit Domain Functionality 310 Within a routing domain, it is recommended that an Anycast-RP set 311 defined in this specification should not be mixed with MSDP peering 312 among the members. In some cases, the source discovery will work but 313 it may not be obvious to the implementations what sources are local 314 to the domain and which are not. This may affect external MSDP 315 advertisement of internal sources. 317 Having said that, this draft makes no attempt to connect MSDP peering 318 domains together by using Anycast-PIM inside a transit domain. 320 6.0 IANA Considerations 322 This document makes no request to IANA. 324 Note to RFC Editor: this section may be removed on publication as an 325 RFC. 327 7.0 Security Consideration 328 This section describes the security consideration for Register and 329 Register-Stop messages between Anycast-RPs. For PIM messages between 330 DR and RP, please see [N1]. 332 7.1 Attack Based On Forged Messages 334 An attacker may forge a Register message using one of the addresses 335 in the Anycast-RP list in order to achieve one or more of the 336 following effects: 338 1. Overwhelm the target RP in a denial-of-service attack 339 2. Inject unauthorized data to receivers served by the RP 340 3. Inject unauthorized data and create bogus SA entries in other 341 PIM domains if the target RP has external MSDP peerings 343 An attacker may also forge a Register-Stop message using one of the 344 addresses in the Anycast-RP list. However, besides denial-of- 345 service, the effect of such an attack is limited because an RP 346 usually ignores Register-Stop messages. 348 7.2 Protect Register and Register-Stop Messages 350 The DOS attack using forged Register or Register-Stop messages can 351 not be prevented. But the RP can still be protected. For example, 352 the RP can rate-limit incoming messages. It can also choose to 353 refuse to process any Register-Stop messages. The actual protection 354 mechansim is implementation specific. 356 The distribution of unauthorized data and bogus Register messages can 357 be prevented using the method described in section 6.3.2 of [N1]. 358 When RP1 sends a copy of a register to RP2, RP1 acts as [N1] 359 describes the DR and RP2 acts as [N1] describes the RP. 361 As described in [N1], an RP can be configured using a unique SA and 362 SPI for traffic (Registers or Register-Stops) to each member of 363 Anycast-RPs in the list, but this results in a key management 364 problem; therefore, it may be preferable in PIM domains where all 365 Rendezvous Points are under a single administrative control, to use 366 the same authentication algorithm parameters (including the key) for 367 all Registered packets in a domain. 369 8.0 Acknowledgments 371 The authors would like to thank Yiqun Cai and Dino Farinacci for 372 prototyping this draft in the cisco IOS and Procket implementations, 373 respectively. 375 The authors would like to thank John Zwiebel for doing 376 interoperability testing of the two prototype implementations. 378 The authors would like to thank Thomas Morin from France Telecom for 379 having an extensive discussion on Multicast the Registers to an SSM- 380 based full mesh among the anycast-RP set. This idea may come in a 381 subsequent Internet Draft. 383 And finally, the authors would like to thank the following for their 384 comments on earlier drafts: 386 Greg Shepherd (Procket Networks (now cisco Systems)) 387 Lenny Giuliano (Juniper Networks) 388 Prashant Jhingran (Huawei Technologies) 389 Pekka Savola (CSC/FUNET) 390 Bill Fenner (AT&T) 391 James Lingard (Data Connection) 392 Amit Shukla (Juniper Networks) 393 Tom Pusateri (Juniper Networks) 395 9.0 Author Information 397 Dino Farinacci 398 cisco Systems 399 dino@cisco.com 401 Yiqun Cai 402 cisco Systems 403 ycai@cisco.com 405 10.0 References 407 10.1 Normative References 409 [N1] Fenner, Handley, Holbrook, Kouvelas, "Protocol Independent 410 Multicast - Sparse Mode (PIM-SM):Protocol Specification 411 (Revised)", Internet Draft draft-ietf-pim-sm-v2-new-11.txt, 412 October 2004. 414 [N2] Bradner, S., "Key words for use in RFCs to Indicate Requirement 415 Levels", BCP 14, RFC 2119, March 1997. 417 10.2 Informative References 419 [I1] Kim, Meyer, Kilmer, Farinacci, "Anycast RP mechanism using PIM 420 and MSDP", RFC 3446, January 2003. 422 Appendix A - Possible Configuration Language 424 A possible set of commands to be used could be: 426 ip pim anycast-rp 428 Where: 430 describes the Anycast-RP set for the RP which 431 is assigned to the group range. This IP address is the address 432 that first-hop and last-hop PIM routers use to register and join 433 to. 435 describes the IP address where Register messages copies 436 are sent to. This IP address is any address assigned to the RP 437 router not including the . 439 Example: 441 From the illustration above, the configuration commands would be: 443 ip pim anycast-rp RPA RP1 444 ip pim anycast-rp RPA RP2 445 ip pim anycast-rp RPA RP3 447 Comment: 449 It may be useful to include the local router IP address in the 450 command set so the above lines can be cut-and-pasted or scripted 451 into all the RPs in the Anycast-RP set. 453 But the implementation would have to be aware of its own address 454 and not inadvertently send a Register to itself.