idnits 2.17.1 draft-ietf-lsr-isis-flood-reflection-05.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack an Introduction section. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (22 November 2021) is 886 days in the past. Is this intentional? Checking references for intended status: Experimental ---------------------------------------------------------------------------- No issues found here. Summary: 1 error (**), 0 flaws (~~), 1 warning (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group A. Przygienda, Ed. 3 Internet-Draft C. Bowers 4 Intended status: Experimental Juniper 5 Expires: 26 May 2022 Y. Lee 6 A. Sharma 7 Comcast 8 R. White 9 Juniper 10 22 November 2021 12 IS-IS Flood Reflection 13 draft-ietf-lsr-isis-flood-reflection-05 15 Abstract 17 This document describes a backwards compatible, optional ISIS 18 extension that allows the creation of IS-IS flood reflection 19 topologies. Flood reflection allows topologies in which L1 areas 20 provide transit forwarding for L2 using all available L1 nodes 21 internally. It accomplishes this by creating L2 flood reflection 22 adjacencies within each L1 area. Those adjacencies are used to flood 23 L2 LSPDUs, and they are used in the L2 SPF computation. However, 24 they are not used for forwarding within the flood reflection cluster. 25 This arrangement gives the L2 topology significantly better scaling 26 properties. As additional benefit, only those routers directly 27 participating in flood reflection have to support the feature. This 28 allows for the incremental deployment of scalable L1 transit areas in 29 an existing network, without the necessity of upgrading other routers 30 in the network. 32 Requirements Language 34 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 35 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 36 document are to be interpreted as described in RFC 2119 [RFC2119]. 38 Status of This Memo 40 This Internet-Draft is submitted in full conformance with the 41 provisions of BCP 78 and BCP 79. 43 Internet-Drafts are working documents of the Internet Engineering 44 Task Force (IETF). Note that other groups may also distribute 45 working documents as Internet-Drafts. The list of current Internet- 46 Drafts is at https://datatracker.ietf.org/drafts/current/. 48 Internet-Drafts are draft documents valid for a maximum of six months 49 and may be updated, replaced, or obsoleted by other documents at any 50 time. It is inappropriate to use Internet-Drafts as reference 51 material or to cite them other than as "work in progress." 53 This Internet-Draft will expire on 26 May 2022. 55 Copyright Notice 57 Copyright (c) 2021 IETF Trust and the persons identified as the 58 document authors. All rights reserved. 60 This document is subject to BCP 78 and the IETF Trust's Legal 61 Provisions Relating to IETF Documents (https://trustee.ietf.org/ 62 license-info) in effect on the date of publication of this document. 63 Please review these documents carefully, as they describe your rights 64 and restrictions with respect to this document. Code Components 65 extracted from this document must include Revised BSD License text as 66 described in Section 4.e of the Trust Legal Provisions and are 67 provided without warranty as described in the Revised BSD License. 69 Table of Contents 71 1. Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . 3 72 2. Description . . . . . . . . . . . . . . . . . . . . . . . . . 3 73 3. Further Details . . . . . . . . . . . . . . . . . . . . . . . 8 74 4. Flood Reflection TLV . . . . . . . . . . . . . . . . . . . . 9 75 5. Flood Reflection Discovery Sub-TLV . . . . . . . . . . . . . 10 76 6. Flood Reflection Discovery Tunnel Type Sub-Sub-TLV . . . . . 11 77 7. Flood Reflection Adjacency Sub-TLV . . . . . . . . . . . . . 12 78 8. Flood Reflection Discovery . . . . . . . . . . . . . . . . . 13 79 9. Flood Reflection Adjacency Formation . . . . . . . . . . . . 14 80 10. Route Computation . . . . . . . . . . . . . . . . . . . . . . 14 81 10.1. Tunnel Based Deployment . . . . . . . . . . . . . . . . 15 82 10.2. No Tunnel Deployment . . . . . . . . . . . . . . . . . . 15 83 11. Redistribution of Prefixes . . . . . . . . . . . . . . . . . 15 84 12. Special Considerations . . . . . . . . . . . . . . . . . . . 16 85 13. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 16 86 13.1. New IS-IS TLV Codepoint . . . . . . . . . . . . . . . . 16 87 13.2. Sub TLVs for TLV 242 . . . . . . . . . . . . . . . . . . 17 88 13.3. Sub-sub TLVs for Flood Reflection Discovery sub-TLV . . 17 89 13.4. Sub TLVs for TLV 22, 23, 25, 141, 222, and 223 . . . . . 17 90 14. Security Considerations . . . . . . . . . . . . . . . . . . . 17 91 15. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 18 92 16. References . . . . . . . . . . . . . . . . . . . . . . . . . 18 93 16.1. Informative References . . . . . . . . . . . . . . . . . 18 94 16.2. Normative References . . . . . . . . . . . . . . . . . . 18 95 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 19 97 1. Glossary 99 This section is introduced first with the intention of allowing quick 100 reference later in the document to terms introduced 102 Flood Reflector: 103 Node configured to connect L2 only to flood reflector clients and 104 reflect (reflood) ISIS L2 LSPs amongst them. 106 Flood Reflector Client: 107 Node configured to build flood reflector adjacencies and normal L2 108 nodes. 110 Flood Reflector Adjacency: 111 ISIS L2 adjacency limited by one end being client and the other 112 reflector and agreeing on the same Flood Reflector Cluster ID. 114 Flood Reflector Cluster: 115 Collection of clients and flood reflectors configured with the 116 same cluster identifier. Cluster ID value of 0 SHOULD NOT be used 117 since it may be used in the future for special purposes. 119 Tunnel Deployment: 120 Deployment where flood reflector clients build a full mesh of 121 tunnels in L1 to "shortcut" forwarding of L2 traffic through the 122 cluster. 124 No Tunnel Deployment: 125 Deployment where flood reflector clients redistribute L2 126 reachability into L1 to allow forwarding through the cluster 127 without underlying tunnels. 129 2. Description 131 Due to the inherent properties of link-state protocols the number of 132 IS-IS routers within a flooding domain is limited by processing and 133 flooding overhead on each node. While that number can be maximized 134 by well written implementations and techniques such as exponential 135 back-offs, IS-IS will still reach a saturation point where no further 136 routers can be added to a single flooding domain. In some L2 137 backbone deployment scenarios, this limit presents a significant 138 challenge. 140 The traditional approach to increasing the scale of an IS-IS 141 deployement is to break it up into multiple L1 flooding domains and a 142 single L2 backbone. This works well for designs where an L2 backbone 143 connects L1 access topologies, but it is limiting where a large L2 is 144 supposed to span large number of routers. In such scenarios, an 145 alternative approach is to consider multiple L2 flooding domains 146 connected together via L1 flooding domains. In other words, L2 147 flooding domains are connected by "L1/L2 lanes" through the L1 areas 148 to form a single L2 backbone again. Unfortunately, in its simplest 149 implementation, this requires the inclusion of most, or all, of the 150 transit L1 routers as L1/L2 to allow traffic to flow along optimal 151 paths through such transit areas. Consequently, this approach fails 152 to reduce the number of L2 routers involved, so it fails to increase 153 the scalability of the L2 backbone. 155 +----+ +-------+ +-------+ +-------+ +----+ 156 | R1 | | R10 +------------+ R20 +---------------+ R30 | | R4 | 157 | L2 +--+ L1/L2 | | L1 | | L1/L2 +--+ L2 | 158 | | | +--------+ +-+ | +------------+ | | | 159 +----+ ++-+--+-+ | | +---+---+----------+ +-+--+-++ +----+ 160 | | | | | | | | | | | | | 161 | | | | | | | | | +-----------+ | | 162 | | +-------+ | | | | | | | | | | 163 | | | | | | | | | | | +------+ | 164 | +------+ +--------+ | +-------+ | | | 165 | | | | | | | | | | | | | 166 +----+ ++------+---+ | +---+---+---+--+ | +-------+------++ +----+ 167 | R2 | | R11 | | | | | R21 | | | | | R31 | | R5 | 168 | L2 +--+ L1/L2 +------------+ L1 +---------------+ L1/L2 +--+ L2 | 169 | | | | | | | | | | | | | | | | 170 +----+ ++------+---+ | | +---+--++ | +-------+------++ +----+ 171 | | | | | | | | | | | | | 172 | +---------------+ | | | | | | | | 173 | | | | | | | | | | | | | 174 | | +--------------+ | +-----------------+ | 175 | | | | | | | | | | | | | 176 +----+ ++-+--+-+ | | +------+---+---+-----+ | | | ++-----++ +----+ 177 | R3 | | R12 | +----------| R22 | | +----+ R32 | | R6 | 178 | L2 +--+ L1/L2 | +--------| L1 +-------+ | | L1/L2 +--+ L2 | 179 | | | +------------+ |---------------+ | | | 180 +----+ +-------+ +-------+-------------+ +-------+ +----+ 182 Figure 1: Example topology 184 Figure 1 is an example of a network where a topologically rich L1 185 area is used to provide transit between six different L2-only routers 186 (R1-R6). Note that the six L2-only routers do not have connectivity 187 to one another over L2 links. To take advantage of the abundance of 188 paths in the L1 transit area, all the intermediate systems could be 189 placed into both L1 and L2, but this essentially combines the 190 separate L2 flooding domains into a single one, triggering again 191 maximum L2 scale limitation we try to address in first place. 193 A more effective solution would allow to reduce the number of links 194 and routers exposed in L2, while still utilizing the full L1 topology 195 when forwarding through the network. 197 [RFC8099] describes Topology Transparent Zones (TTZ) for OSPF. The 198 TTZ mechanism represents a group of OSPF routers as a full mesh of 199 adjacencies between the routers at the edge of the group. A similar 200 mechanism could be applied to ISIS as well. However, a full mesh of 201 adjacencies between edge routers (or L1/L2 nodes) significantly 202 limits the scale of the topology. The topology in Figure 1 has 6 L1/ 203 L2 nodes. Figure 2 illustrates a full mesh of L2 adjacencies between 204 the 6 L1/L2 nodes, resulting in (5 * 6)/2 = 15 L2 adjacencies. In a 205 somewhat larger topology containing 20 L1/L2 nodes, the number of L2 206 adjacencies in a full mesh rises to 190. 208 +----+ +-------+ +-------------------------------+-------+ +----+ 209 | R1 | | R10 | | | R30 | | R4 | 210 | L2 +--+ L1/L2 +------------------------------------+ L1/L2 +--+ L2 | 211 | | | | | | | | | 212 +----+ ++-+-+--+-+ | +-+--+---++ +----+ 213 | | | | | | | | 214 | +----------------------------------------------+ | 215 | | | | | | | | 216 | +-----------------------------------+ | | | | 217 | | | | | | | | 218 | +----------------------------------------+ | | 219 | | | | | | | | 220 +----+ ++-----+- | | | | -----+-++ +----+ 221 | R2 | | R11 | | | | | | R31 | | R5 | 222 | L2 +--+ L1/L2 +------------------------------------+ L1/L2 +--+ L2 | 223 | | | | | | | | | | | | 224 +----+ ++------+------------------------------+ | | +----+-++ +----+ 225 | | | | | | | | 226 | | | | | | | | 227 | +-------------------------------------------+ | 228 | | | | | | | | 229 | | | | +----------+ | 230 | | | | | | | | 231 | | | | +-----+ | | 232 | | | | | | | | 233 +----+ ++----+-+-+ | +-+-+--+-++ +----+ 234 | R3 | | R12 | | L2 adjacency | R32 | | R6 | 235 | L2 +--+ L1/L2 +------------------------------------+ L1/L2 +--+ L2 | 236 | | | | | | | | | 237 +----+ +-------+----+ +-------+ +----+ 238 Figure 2: Example topology represented in L2 with a full mesh of 239 L2 adjacencies between L1/L2 nodes 241 BGP, as specified in [RFC4271], faced a similar scaling problem, 242 which has been solved in many networks by deploying BGP route 243 reflectors [RFC4456]. We note that BGP route reflectors do not 244 necessarily have to be in the forwarding path of the traffic. This 245 incongruity of forwarding and control path for BGP route reflectors 246 allows the control plane to scale independently of the forwarding 247 plane. 249 We propose here a similar solution for IS-IS. A simple example of 250 what a flood reflector control plane approach would look like is 251 shown in Figure 3, where router R21 plays the role of a flood 252 reflector. Each L1/L2 ingress/egress router builds a tunnel to the 253 flood reflector, and an L2 adjacency is built over each tunnel. In 254 this solution, we need only 6 L2 adjacencies, instead of the 15 255 needed for a full mesh. In a somewhat larger topology containing 20 256 L1/L2 nodes, this solution requires only 20 L2 adjacencies, instead 257 of the 190 need for a full mesh. Multiple flood reflectors can be 258 used, allowing the network operator to balance between resilience, 259 path utilization, and state in the control plane. The resulting L2 260 adjacency scale is R*n, where R is the number of flood reflectors 261 used and n is the number of L1/L2 nodes. This compares quite 262 favorably with n*(n-1)/2 L2 adjacencies required in a fully meshed L2 263 solution. 265 +----+ +-------+ +-------+ +----+ 266 | R1 | | R10 | | R30 | | R4 | 267 | L2 +--+ L1/L2 +--------------+ +-----------------+ L1/L2 +--+ L2 | 268 | | | | L2 adj | | L2 adj | | | | 269 +----+ +-------+ over | | over +-------+ +----+ 270 tunnel | | tunnel 271 +----+ +-------+ +--+---+--+ +-------+ +----+ 272 | R2 | | R11 | | R21 | | R31 | | R5 | 273 | L2 +--+ L1/L2 +-----------+ L1/L2 +--------------+ L1/L2 +--+ L2 | 274 | | | | L2 adj | flood | L2 adj | | | | 275 +----+ +-------+ over |reflector| over +-------+ +----+ 276 tunnel +--+---+--+ tunnel 277 +----+ +-------+ | | +-------+ +----+ 278 | R3 | | R12 +--------------+ +-----------------+ R32 | | R6 | 279 | L2 +--+ L1/L2 | L2 adj L2 adj | L1/L2 +--+ L2 | 280 | | | | over over | | | | 281 +----+ +-------+ tunnel tunnel +-------+ +----+ 283 Figure 3: Example topology represented in L2 with L2 adjacencies 284 from each L1/ L2 node to a single flood reflector 286 As illustrated in Figure 3, when R21 plays the role of flood 287 reflector, it provides L2 connectivity among all of the previously 288 disconnected L2 islands by reflooding all L2 LSPDUs. At the same 289 time, R20 and R22 remain L1-only routers. L1-only routers and 290 L1-only links are not visible in L2. In this manner, the flood 291 reflector allows us provide L2 control plane connectivity in a 292 scalable manner. 294 As described so far, the solution illustrated in Figure 3 relies only 295 on currently standardized ISIS functionality. Without new 296 functionality, however, the data traffic will traverse only R21. 297 This will unnecessarily create a bottleneck at R21 since there is 298 still available capacity in the paths crossing the L1-only routers 299 R20 and R22. 301 Hence, some new functionality is necessary to allow the L1/L2 edge 302 nodes (R10-12 and R30-32 in Figure 3) to recognize that the L2 303 adjacency to R21 should not be used for forwarding. The L1/L2 edge 304 nodes should forward traffic that would normally be forwarded over 305 the L2 adjacency to R21 over L1 links instead. This would allow the 306 forwarding within the L1 area to use the L1-only nodes and links 307 shown in Figure 1 as well. It allows networks to be built that use 308 the entire forwarding capacity of the L1 areas, while at the same 309 time introducing control plane scaling benefits provided by L2 flood 310 reflectors. 312 This document defines all extensions necessary to support flood 313 reflector deployment: 315 * A 'flood reflector adjacency' for all the adjacencies built for 316 the purpose of reflecting flooding information. This allows these 317 'flood reflectors' to participate in the IS-IS control plane 318 without being used in the forwarding plane. This is a purely 319 local operation on the L1/L2 ingress; it does not require 320 replacing or modifying any routers not involved in the reflection 321 process. Deployment-wise, it is far less tricky to just upgrade 322 the routers involved in flood reflection rather than have a flag 323 day on the whole ISIS domain. 325 * An (optional) full mesh of tunnels between the L1/L2 routers, 326 ideally load-balancing across all available L1 links. This 327 harnesses all forwarding paths between the L1/L2 edge nodes 328 without injecting unneeded state into the L2 flooding domain or 329 creating 'choke points' at the 'flood reflectors' themselves. The 330 draft is agnostic as to the tunneling technology used but provides 331 enough information for automatic establishment of such tunnels. 332 The discussion of ISIS adjacency formation and/or liveness 333 discovery on such tunnels is outside the scope of this draft and 334 is largely choice of the underlying implementation. A solution 335 without tunnels is also possible by applying judicious scoping of 336 reachability information between the levels as described in more 337 details later. 339 * Some way to support reflector redundancy, and potentially some way 340 to auto-discover and advertise such adjacencies as flood reflector 341 adjacencies. Such advertisements may allow L2 nodes outside the 342 L1 to perform optimizations in the future based on this 343 information. 345 3. Further Details 347 Several considerations should be noted in relation to such a flood 348 reflection mechanism. 350 First, this allows multi-area IS-IS deployments to scale without any 351 major modifications in the IS-IS implementation on most of the nodes 352 deployed in the network. Unmodified (traditional) L2 routers will 353 compute reachability across the transit L1 area using the flood 354 reflector adjacencies. 356 Second, the flood reflectors are not required to participate in 357 forwarding traffic through the L1 transit area. These flood 358 reflectors can be hosted on virtual devices outside the forwarding 359 topology. 361 Third, astute readers will realize that flooding reflection may cause 362 the use of suboptimal paths. This is similar to the BGP route 363 reflection suboptimal routing problem described in 364 [ID.draft-ietf-idr-bgp-optimal-route-reflection-28]. The L2 365 computation determines the egress L1/L2 and with that can create 366 illusions of ECMP where there is none. And in certain scenarios lead 367 to an L1/L2 egress which is not globally optimal. This represents a 368 straightforward instance of the trade-off between the amount of 369 control plane state and the optimal use of paths through the network 370 often encountered when aggregating routing information. 372 One possible solution to this problem is to expose additional 373 topology information into the L2 flooding domains. In the example 374 network given, links from router 01 to router 02 can be exposed into 375 L2 even when 01 and 02 are participating in flood reflection. This 376 information would allow the L2 nodes to build 'shortcuts' when the L2 377 flood reflected part of the topology looks more expensive to cross 378 distance wise. 380 Another possible variation is for an implementation to approximate 381 with the tunnel cost the cost of the underlying topology. 383 Redundancy can be achieved by building multiple flood reflectors in a 384 L1 area. Multiple flood reflectors do not need any synchronization 385 mechanisms amongst themselves, except standard ISIS flooding and 386 database maintenance procedures. 388 4. Flood Reflection TLV 390 The Flood Reflection TLV is a new top-level TLV that MAY appear in L2 391 IIHs. The Flood Reflection TLV indicates the flood reflector cluster 392 (based on Flood Reflection Cluster ID) that a given router is 393 configured to participate in. It also indicates whether the router 394 is configured to play the role of either flood reflector or flood 395 reflector client. The Flood Reflection Cluster ID and flood 396 reflector roles advertised in the IIHs are used to ensure that flood 397 reflector adjacencies are only formed between a flood reflector and 398 flood reflector client, and that the Flood Reflection Cluster IDs 399 match. The Flood Reflection TLV has the following format: 401 0 1 2 3 402 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 1 403 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 404 | Type | Length |C| RESERVED | 405 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 406 | Flood Reflection Cluster ID | 407 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 408 | Sub-TLVs ... 409 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 411 Type: TBD 413 Length: The length, in octets, of the following fields. 415 C (Client): This bit is set to indicate that the router acts as a 416 flood reflector client. When this bit is NOT set, the router acts 417 as a flood reflector. On a given router, the same value of the 418 C-bit MUST be advertised across all interfaces advertising the 419 Flood Reflection TLV in IIHs. 421 RESERVED: This field is reserved for future use. It MUST be set to 422 0 when sent and MUST be ignored when received. 424 Flood Reflection Cluster ID: Flood Reflection Cluster Identifier. 425 These same 32-bit value MUST be assigned to all of the flood 426 reflectors and flood reflector clients in the same L1 area. The 427 value MUST be unique across different L1 areas within the IGP 428 domain. On a given router, the same value of the Flood Reflection 429 Cluster ID MUST be advertised across all interfaces advertising 430 the Flood Reflection TLV in IIHs. This implies that a flood 431 reflector can participate in a single L1 area only. In case of 432 Cluster ID value of 0, the TLV MUST be ignored. 434 Sub-TLVs: Optional sub-TLVs. For future extensibility, the format 435 of the Flood Reflection TLV allows for the possibility of 436 including optional sub-TLVs. No sub-TLVs of the Flood Reflection 437 TLV are defined in this document. 439 The Flood Reflection TLV SHOULD NOT appear more than once in an IIH. 440 A router receiving multiple Flood Reflection TLVs in the same IIH 441 MUST use the values in the first TLV and it SHOULD adequately log 442 such violations subject to rate limiting. 444 5. Flood Reflection Discovery Sub-TLV 446 Flood Reflection Discovery sub-TLV is advertised as a sub-TLV of the 447 IS-IS Router Capability TLV-242, defined in [RFC7981]. The Flood 448 Reflection Discovery sub-TLV is advertised in L1 and L2 LSPs with 449 area flooding scope in order to enable the auto-discovery of flood 450 reflection capabilities. The Flood Reflection Discovery sub-TLV has 451 the following format: 453 0 1 2 3 454 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 1 455 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 456 | Type | Length |C| Reserved | 457 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 458 | Flood Reflection Cluster ID | 459 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 461 Type: TBD 463 Length: The length, in octets, of the following fields. 465 C (Client): This bit is set to indicate that the router acts as a 466 flood reflector client. When this bit is NOT set, the router acts 467 as a flood reflector. 469 RESERVED: This field is reserved for future use. It MUST be set to 470 0 when sent and MUST be ignored when received. 472 Flood Reflection Cluster ID: The Flood Reflection Cluster Identifier 473 is the same as that defined in the Flood Reflection TLV and obeys 474 the same rules. 476 The Flood Reflection Discovery sub-TLV SHOULD NOT appear more than 477 once in TLV 242. A router receiving multiple Flood Reflection 478 Discovery sub-TLVs in TLV 242 MUST use the values in the first sub- 479 TLV and it SHOULD adequately log such violations subject to rate 480 limiting. 482 6. Flood Reflection Discovery Tunnel Type Sub-Sub-TLV 484 Flood Reflection Discovery Tunnel Type sub-sub-TLV is advertised 485 optionally as a sub-sub-TLV of the Flood Reflection Discovery Sub- 486 TLV, defined in Section 5. It allows the automatic creation of L2 487 tunnels to be used as flood reflector adjacencies and L1 shortcut 488 tunnels. The Flood Reflection Tunnel Type sub-sub-TLV has the 489 following format: 491 0 1 2 3 492 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 1 493 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+---------------+ 494 | Type | Length | Reserved |F| 495 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 496 | Tunnel Encapsulation Attribute | 497 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 499 Type: TBD 501 Length: The length, in octets, of zero or more of the following 502 fields. 504 Reserved: SHOULD be 0 on transmission and ignored on reception. 506 F Flag: When set indicates flood reflection tunnel endpoint, when 507 clear, indicates possible L1 shortcut tunnel endpoint. 509 Tunnel Encapsulation Attribute: Carries encapsulation type and 510 further attributes necessary for tunnel establishment as defined 511 in [RFC9012]. Protocol type sub-TLV as defined in [RFC9012] MAY 512 be included but MUST when F flag is set include according type 513 that allows carrying of encapsulated ISIS frames. Such tunnel 514 type MUST provide according mechanisms to carry up to 515 `originatingL2LSPBufferSize` sized ISIS frames across. 517 A flood reflector receiving multiple Flood Reflection Discovery 518 Tunnel Type sub-sub-TLVs in Flood Reflection Discovery sub-TLV with F 519 flag set SHOULD use one or more of the specified tunnel endpoints to 520 automatically establish one or more tunnels that will serve as flood 521 reflection adjacency(-ies). 523 A flood reflection client receiving multiple Flood Reflection 524 Discovery Tunnel Type sub-sub-TLVs in Flood Reflection Discovery sub- 525 TLV with F flag clear from other leaves MAY use one or more of the 526 specified tunnel endpoints to automatically establish one or more 527 tunnels that will serve as L1 tunnel shortcuts. 529 Optional address validation procedures as defined in [RFC9012] MUST 530 be disregarded. 532 7. Flood Reflection Adjacency Sub-TLV 534 The Flood Reflection Adjacency sub-TLV is advertised as a sub-TLV of 535 TLVs 22, 23, 25, 141, 222, and 223. Its presence indicates that a 536 given adjacency is a flood reflector adjacency. It is included in L2 537 area scope flooded LSPs. Flood Reflection Adjacency sub-TLV has the 538 following format: 540 0 1 2 3 541 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 1 542 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 543 | Type | Length |C| Reserved | 544 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 545 | Flood Reflection Cluster ID | 546 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 548 Type: TBD 550 Length: The length, in octets, of the following fields. 552 C (Client): This bit is set to indicate that the router advertising 553 this adjacency is a flood reflector client. When this bit is NOT 554 set, the router advertising this adjacency is a flood reflector. 556 RESERVED: This field is reserved for future use. It MUST be set to 557 0 when sent and MUST be ignored when received. 559 Flood Reflection Cluster ID: The Flood Reflection Cluster Identifier 560 is the same as that defined in the Flood Reflection TLV and obeys 561 the same rules. 563 The Flood Reflection Adjacency sub-TLV SHOULD NOT appear more than 564 once in a given TLV. A router receiving multiple Flood Reflection 565 Adjacency sub-TLVs in a TLV MUST use the values in the first sub-TLV 566 and it SHOULD adequately log such violations subject to rate 567 limiting. 569 8. Flood Reflection Discovery 571 A router participating in flood reflection MUST be configured as an 572 L1/L2 router. It SHOULD originate the Flood Reflection Discovery 573 sub-TLV with area flooding scope in L1 and L2. Normally, all routers 574 on the edge of the L1 area (those having traditional L2 adjacencies) 575 will advertise themselves as route reflector clients. Therefore, a 576 flood reflector client will have both traditional L2 adjacencies and 577 flood reflector L2 adjacencies. 579 A router acting as a flood reflector MUST NOT have any traditional L2 580 adjacencies. It will be an L1/L2 router only by virtue of having 581 flood reflector L2 adjacencies. A router desiring to act as a flood 582 reflector will advertise itself as such using the Flood Reflection 583 Discovery sub-TLV in L1 and L2. 585 A given flood reflector or flood reflector client can only 586 participate in a single cluster, as determined by the value of its 587 Flood Reflection Cluster ID and should disregard other routers' TLVs 588 for flood reflection purposes if the cluster ID is not matching. 590 Upon reception of Flood Reflection Discovery sub-TLVs, a router 591 acting as flood reflector client SHOULD initiate a tunnel towards 592 each flood reflector with which it shares an Flood Reflection Cluster 593 ID using one or more of the tunnel encapsulations provided with F 594 flag being set. The L2 adjacencies formed over such tunnels MUST be 595 marked as flood reflector adjacencies. If the client or reflector 596 has a direct L2 adjacency with the according remote side it SHOULD 597 use it instead of instantiating a new tunnel. 599 In absence of auto-discovery an implementation MAY use statically 600 configured tunnels to create flood reflection adjacencies. 602 The ISIS metrics for all flood reflection adjacencies in a cluster 603 SHOULD be uniform. 605 Upon reception of Flood Reflection Discover TLVs, a router acting as 606 a flood reflector client MAY initiate tunnels with L1-only 607 adjacencies towards any of the other flood reflector clients with 608 lower router IDs in its cluster using encapsulations with F flag 609 clear. These tunnels MAY be used for forwarding to improve the load- 610 balancing characteristics of the L1 area. If the clients have a 611 direct L2 adjacency they SHOULD use it instead of instantiating a new 612 tunnel. 614 9. Flood Reflection Adjacency Formation 616 In order to simplify both implementations and network deployments, 617 this draft does not allow the formation of complex hierarchies of 618 flood reflectors and clients or allow multiple clusters in a single 619 L1 area. Consequently, all flood reflectors and flood reflector 620 clients in the same L1 area MUST share the same Flood Reflector 621 Cluster ID. 623 A flood reflector MUST only form flood reflection adjacencies with 624 flood reflector clients with matching Cluster ID. A flood reflector 625 MUST NOT form any traditional L2 adjacencies. 627 Flood reflector clients MUST only form flood reflection adjacencies 628 with flood reflectors with matching Cluster ID. 630 Flood reflector clients MAY form traditional L2 adjacencies with 631 flood reflector clients or nodes not participating in flood 632 reflection. When two clients form traditional L2 adjacency Cluster 633 ID is disregarded. 635 The Flood Reflector Cluster ID and flood reflector roles advertised 636 in the Flood Reflection TLVs in IIHs are used to ensure that flood 637 reflection adjacencies that are established meet the above criteria. 639 On change in either flood reflection role or cluster ID on IIH on the 640 local or remote side the adjacency has to be reset and re-established 641 if possible. 643 Once a flood reflection adjacency is established, the flood reflector 644 and the flood reflector client MUST advertise the adjacency by 645 including the Flood Reflection Adjacency Sub-TLV in the Extended IS 646 reachability TLV or MT-ISN TLV. 648 10. Route Computation 650 To ensure loop-free routing, the route reflection client MUST follow 651 the normal L2 computation to determine L2 routes. This is because 652 nodes outside the L1 area will generally not be aware that flood 653 reflection is being performed. The flood reflection clients need to 654 produce the same result for the L2 route computation as a router not 655 participating in flood reflection. 657 10.1. Tunnel Based Deployment 659 In tunnel based option the reflection client, after L2 and L1 660 computation, MUST examine all L2 routes and replace all flood 661 reflector adjacencies with the correct underlying tunnel next-hop to 662 the egress. 664 10.2. No Tunnel Deployment 666 In case of deployment without underlying tunnels, the necessary L2 667 routes are distributed into the area, normally as L2->L1 routes. Due 668 to the rules in Section 9 the computation in the resulting topology 669 is relatively simple, the L2 SPF from a flood reflector client is 670 guaranteed to reach within a hop the Flood Reflector and in the 671 following hop the L2 egress to which it has a forwarding tunnel 672 again. All the flood reflector tunnel nexthops in the according L2 673 route can hence be removed and if the L2 route has no other ECMP L2 674 nexthops, the L2 route MUST be suppressed in the RIB by some means to 675 allow the less preferred L2->L1 route to be used to forward traffic 676 towards the advertising egress. 678 In the particular case the client has L2 routes which are not route 679 reflected, those will be naturally preferred (such routes normally 680 "hot-potato" route of the L1 area). However in the case the L2 route 681 through the flood reflector egress is "shorter" than such present non 682 flood reflected L2 routes, the node SHOULD ensure that such routes 683 are suppressed so the L2->L1 towards the egress still takes 684 preference. Observe that operationally this can be resolved in a 685 relatively simple way by configuring flood reflector adjacencies to 686 have a high metric, i.e. the flood reflector topology becomes "last 687 resort" and the leaves will try to "hot-potato" out the area as fast 688 as possible which is normally the desirable behavior. 690 In deployment scenarios where tunnels are not used, all L1/L2 edge 691 nodes MUST be ultimately flood reflector clients except during during 692 transition phase. 694 11. Redistribution of Prefixes 696 When L2 prefixes need to be redistributed into L1 by the route 697 reflector clients a client that does not have any L2 flood reflector 698 adjacencies MUST NOT redistribute those routes into the area in case 699 of application of Section 10.2. The L2 prefixes advertisements 700 redistributed into L1 with flood reflectors SHOULD be normally 701 limited to L2 intra-area routes (as defined in [RFC7775]), if the 702 information exists to distinguish them from other other L2 prefix 703 advertisements. 705 On the other hand, in topologies that make use of flood reflection to 706 hide the structure of L1 areas while still providing transit 707 forwarding across them using tunnels, we generally do not need to 708 redistribute L1 prefixes advertisements into L2. 710 12. Special Considerations 712 In pathological cases setting the overload bit in L1 (but not in L2) 713 can partition L1 forwarding, while allowing L2 reachability through 714 flood reflector adjacencies to exist. In such a case a node cannot 715 replace a route through a flood reflector adjacency with a L1 716 shortcut and the client can use the L2 tunnel to the flood reflector 717 for forwarding while it MUST initiate an alarm and declare 718 misconfiguration. 720 A flood reflector with directly L2 attached prefixes should advertise 721 those in L1 as well since based on preference of L1 routes the 722 clients will not try to use the L2 flood reflector adjacency to route 723 the packet towards them. A very, very corner case is when the flood 724 reflector is reachable via L2 flood reflector adjacency (due to 725 underlying L1 partition) only in which case the client can use the L2 726 tunnel to the flood reflector for forwarding towards those prefixes 727 while it MUST initiate an alarm and declare misconfiguration. 729 A flood reflector SHOULD NOT set the attached bit on its LSPs. 731 Instead of modifying the computation procedures one could imagine a 732 flood reflector solution where the Flood Reflector would re-advertise 733 the L2 prefixes with a 'third-party' next-hop but that would have 734 less desirable convergence properties than the solution proposed and 735 force a fork-lift of all L2 routers to make sure they disregard such 736 prefixes unless in the same L1 domain as the Flood Reflector. 738 Depending on pseudo-node choice in case of a broadcast domain with 739 multiple flood reflectors attached this can lead to a partitioned LAN 740 and hence a router discovering such a condition MUST initiate an 741 alarm and declare misconfiguration. 743 13. IANA Considerations 745 This document requests allocation for the following IS-IS TLVs and 746 Sub-TLVs. 748 13.1. New IS-IS TLV Codepoint 750 This document requests the following IS-IS TLV: 752 Value Name IIH LSP SNP Purge 753 ----- --------------------------------- --- --- --- ----- 754 TBD1 Flood Reflection y n n n 756 Suggested value for TBD1 is 161. 758 13.2. Sub TLVs for TLV 242 760 This document request the following registration in the "sub-TLVs for 761 TLV 242" registry. 763 Type Description 764 ---- ----------- 765 TBD2 Flood Reflection Discovery 767 Suggested value for TBD2 is 161. 769 13.3. Sub-sub TLVs for Flood Reflection Discovery sub-TLV 771 This document request the following registration in the "sub-sub-TLVs 772 for Flood Reflection Discovery sub-TLV" registry. 774 Type Description 775 ---- ----------- 776 TBD3 Flood Reflection Discovery Tunnel Encapsulation Attribute 778 Suggested value for TBD3 is 161. 780 13.4. Sub TLVs for TLV 22, 23, 25, 141, 222, and 223 782 This document requests the following registration in the "sub-TLVs 783 for TLV 22, 23, 25, 141, 222, and 223" registry. 785 Type Description 22 23 25 141 222 223 786 ---- -------------------------------- --- --- --- --- --- --- 787 TBD4 Flood Reflector Adjacency y y n y y y 789 Suggested value for TBD4 is 161. 791 14. Security Considerations 793 This document introduces no new security concerns to ISIS or other 794 specifications referenced in this document. 796 15. Acknowledgements 798 The authors thank Shraddha Hegde, Peter Psenak, Acee Lindem and Les 799 Ginsberg for their thorough review and detailed discussions. 801 16. References 803 16.1. Informative References 805 [ID.draft-ietf-idr-bgp-optimal-route-reflection-28] 806 Raszuk et al., R., "BGP Optimal Route Reflection", July 807 2019, . 810 [RFC4271] Rekhter, Y., Ed., Li, T., Ed., and S. Hares, Ed., "A 811 Border Gateway Protocol 4 (BGP-4)", RFC 4271, 812 DOI 10.17487/RFC4271, January 2006, 813 . 815 [RFC4456] Bates, T., Chen, E., and R. Chandra, "BGP Route 816 Reflection: An Alternative to Full Mesh Internal BGP 817 (IBGP)", RFC 4456, DOI 10.17487/RFC4456, April 2006, 818 . 820 [RFC8099] Chen, H., Li, R., Retana, A., Yang, Y., and Z. Liu, "OSPF 821 Topology-Transparent Zone", RFC 8099, 822 DOI 10.17487/RFC8099, February 2017, 823 . 825 16.2. Normative References 827 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 828 Requirement Levels", BCP 14, RFC 2119, 829 DOI 10.17487/RFC2119, March 1997, 830 . 832 [RFC7775] Ginsberg, L., Litkowski, S., and S. Previdi, "IS-IS Route 833 Preference for Extended IP and IPv6 Reachability", 834 RFC 7775, DOI 10.17487/RFC7775, February 2016, 835 . 837 [RFC7981] Ginsberg, L., Previdi, S., and M. Chen, "IS-IS Extensions 838 for Advertising Router Information", RFC 7981, 839 DOI 10.17487/RFC7981, October 2016, 840 . 842 [RFC9012] Patel, K., Van de Velde, G., Sangli, S., and J. Scudder, 843 "The BGP Tunnel Encapsulation Attribute", RFC 9012, 844 DOI 10.17487/RFC9012, April 2021, 845 . 847 Authors' Addresses 849 Tony Przygienda (editor) 850 Juniper 851 1137 Innovation Way 852 Sunnyvale, CA 853 United States of America 855 Email: prz@juniper.net 857 Chris Bowers 858 Juniper 859 1137 Innovation Way 860 Sunnyvale, CA 861 United States of America 863 Email: cbowers@juniper.net 865 Yiu Lee 866 Comcast 867 1800 Bishops Gate Blvd 868 Mount Laurel, NJ 08054 869 United States of America 871 Email: Yiu_Lee@comcast.com 873 Alankar Sharma 874 Comcast 875 1800 Bishops Gate Blvd 876 Mount Laurel, NJ 08054 877 United States of America 879 Email: Alankar_Sharma@comcast.com 880 Russ White 881 Juniper 882 1137 Innovation Way 883 Sunnyvale, CA 884 United States of America 886 Email: russw@juniper.net