idnits 2.17.00 (12 Aug 2021) /tmp/idnits522/draft-ietf-roll-useofrplinfo-03.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 : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (April 4, 2016) is 2238 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- ** Obsolete normative reference: RFC 2460 (Obsoleted by RFC 8200) == Outdated reference: draft-ietf-6tisch-architecture has been published as RFC 9030 == Outdated reference: draft-ietf-roll-routing-dispatch has been published as RFC 8138 Summary: 1 error (**), 0 flaws (~~), 3 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 ROLL Working Group M. Robles 3 Internet-Draft Ericsson 4 Intended status: Informational M. Richardson 5 Expires: October 6, 2016 SSW 6 P. Thubert 7 Cisco 8 April 4, 2016 10 When to use RFC 6553, 6554 and IPv6-in-IPv6 11 draft-ietf-roll-useofrplinfo-03 13 Abstract 15 This document looks at different data flows through LLN networks 16 where RPL is used to establish routing. The document enumerates the 17 cases where RFC 6553, RFC 6554 and IPv6-in-IPv6 encapsulation is 18 required. This analysis provides the basis on which to design 19 efficient compression of these headers. 21 Status of This Memo 23 This Internet-Draft is submitted in full conformance with the 24 provisions of BCP 78 and BCP 79. 26 Internet-Drafts are working documents of the Internet Engineering 27 Task Force (IETF). Note that other groups may also distribute 28 working documents as Internet-Drafts. The list of current Internet- 29 Drafts is at http://datatracker.ietf.org/drafts/current/. 31 Internet-Drafts are draft documents valid for a maximum of six months 32 and may be updated, replaced, or obsoleted by other documents at any 33 time. It is inappropriate to use Internet-Drafts as reference 34 material or to cite them other than as "work in progress." 36 This Internet-Draft will expire on October 6, 2016. 38 Copyright Notice 40 Copyright (c) 2016 IETF Trust and the persons identified as the 41 document authors. All rights reserved. 43 This document is subject to BCP 78 and the IETF Trust's Legal 44 Provisions Relating to IETF Documents 45 (http://trustee.ietf.org/license-info) in effect on the date of 46 publication of this document. Please review these documents 47 carefully, as they describe your rights and restrictions with respect 48 to this document. Code Components extracted from this document must 49 include Simplified BSD License text as described in Section 4.e of 50 the Trust Legal Provisions and are provided without warranty as 51 described in the Simplified BSD License. 53 Table of Contents 55 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 56 2. Terminology and Requirements Language . . . . . . . . . . . . 3 57 3. Sample/reference topology . . . . . . . . . . . . . . . . . . 3 58 4. Use cases . . . . . . . . . . . . . . . . . . . . . . . . . . 6 59 5. Storing mode . . . . . . . . . . . . . . . . . . . . . . . . 8 60 5.1. Example of Flow from RPL-aware-leaf to root . . . . . . . 9 61 5.2. Example of Flow from root to RPL-aware-leaf . . . . . . . 10 62 5.3. Example of Flow from root to not-RPL-aware-leaf . . . . . 10 63 5.4. Example of Flow from not-RPL-aware-leaf to root . . . . . 11 64 5.5. Example of Flow from RPL-aware-leaf to Internet . . . . . 12 65 5.6. Example of Flow from Internet to RPL-aware-leaf . . . . . 12 66 5.7. Example of Flow from not-RPL-aware-leaf to Internet . . . 13 67 5.8. Example of Flow from Internet to non-RPL-aware-leaf . . . 14 68 5.9. Example of Flow from RPL-aware-leaf to RPL-aware-leaf . . 14 69 5.10. Example of Flow from RPL-aware-leaf to non-RPL-aware-leaf 15 70 5.11. Example of Flow from not-RPL-aware-leaf to RPL-aware-leaf 17 71 5.12. Example of Flow from not-RPL-aware-leaf to not-RPL-aware- 72 leaf . . . . . . . . . . . . . . . . . . . . . . . . . . 18 73 6. Non Storing mode . . . . . . . . . . . . . . . . . . . . . . 19 74 6.1. Example of Flow from RPL-aware-leaf to root . . . . . . . 19 75 6.2. Example of Flow from root to RPL-aware-leaf . . . . . . . 20 76 6.3. Example of Flow from root to not-RPL-aware-leaf . . . . . 20 77 6.4. Example of Flow from not-RPL-aware-leaf to root . . . . . 21 78 6.5. Example of Flow from RPL-aware-leaf to Internet . . . . . 22 79 6.6. Example of Flow from Internet to RPL-aware-leaf . . . . . 22 80 6.7. Example of Flow from not-RPL-aware-leaf to Internet . . . 23 81 6.8. Example of Flow from Internet to non-RPL-aware-leaf . . . 23 82 6.9. Example of Flow from RPL-aware-leaf to RPL-aware-leaf . . 24 83 6.10. Example of Flow from RPL-aware-leaf to not-RPL-aware-leaf 25 84 6.11. Example of Flow from not-RPL-aware-leaf to RPL-aware-leaf 26 85 6.12. Example of Flow from not-RPL-aware-leaf to not-RPL-aware- 86 leaf . . . . . . . . . . . . . . . . . . . . . . . . . . 26 87 7. Observations about the problem . . . . . . . . . . . . . . . 27 88 7.1. Storing mode . . . . . . . . . . . . . . . . . . . . . . 27 89 7.2. Non-Storing mode . . . . . . . . . . . . . . . . . . . . 28 90 8. 6LoRH Compression cases . . . . . . . . . . . . . . . . . . . 28 91 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 29 92 10. Security Considerations . . . . . . . . . . . . . . . . . . . 29 93 11. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 29 94 12. References . . . . . . . . . . . . . . . . . . . . . . . . . 29 95 12.1. Normative References . . . . . . . . . . . . . . . . . . 29 96 12.2. Informative References . . . . . . . . . . . . . . . . . 30 98 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 30 100 1. Introduction 102 RPL [RFC6550] is a routing protocol for constrained networks. RFC 103 6553 [RFC6553] defines the "RPL option" (RPI), carried within the 104 IPv6 Hop-by-Hop header to quickly identify inconsistencies (loops) in 105 the routing topology. RFC 6554 [RFC6554] defines the "RPL Source 106 Route Header" (RH3), an IPv6 Extension Header to deliver datagrams 107 within a RPL routing domain, particularly in non-storing mode. 109 These various items are referred to as RPL artifacts, and they are 110 seen on all of the data-plane traffic that occurs in RPL routed 111 networks; they do not in general appear on the RPL control plane 112 traffic at all which is mostly hop-by-hop traffic (one exception 113 being DAO messages in non-storing mode). 115 It has become clear from attempts to do multi-vendor 116 interoperability, and from a desire to compress as many of the above 117 artifacts as possible that not all implementors agree when artifacts 118 are necessary, or when they can be safely omitted, or removed. 120 An interim meeting went through the 24 cases defined here to discover 121 if there were any shortcuts, and this document is the result of that 122 discussion. This document should not be defining anything new, but 123 it may clarify what is correct and incorrect behaviour. 125 The related document A Routing Header Dispatch for 6LoWPAN (6LoRH) 126 [I-D.ietf-roll-routing-dispatch] defines a method to compress RPL 127 Option information and Routing Header type 3 (RFC6554) and an 128 efficient IP-in-IP technique. Uses cases proposed for the 129 [Second6TischPlugtest] involving 6loRH. 131 2. Terminology and Requirements Language 133 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 134 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 135 document are to be interpreted as described in RFC 2119 [RFC2119]. 137 Terminology defined in [RFC7102] 139 3. Sample/reference topology 141 A RPL network is composed of a 6LBR (6LoWPAN Border Router), Backbone 142 Router (6BBR), 6LR (6LoWPAN Router) and 6LN (6LoWPAN Node) as leaf 143 logically organized in a DODAG structure (Destination Oriented 144 Directed Acyclic Graph). 146 RPL defines the RPL Control messages (control plane ), a new ICMPv6 147 message with Type 155. DIS, DIO and DAO messages are all RPL Control 148 messages but with different Code values. 150 RPL supports two modes of Downward traffic: in storing mode, it is 151 fully stateful or an in non-storing, it is fully source routed. A 152 RPL Instance is either fully storing or fully non-storing, i.e. a RPL 153 Instance with a combination of storing and non-storing nodes is not 154 supported with the current specifications. 156 +--------------+ 157 | Upper Layers | 158 | | 159 +--------------+ 160 | RPL | 161 | | 162 +--------------+ 163 | ICMPv6 | 164 | | 165 +--------------+ 166 | IPv6 | 167 | | 168 +--------------+ 169 | 6LoWPAN | 170 | | 171 +--------------+ 172 | PHY-MAC | 173 | | 174 +--------------+ 176 Figure 1: RPL Stack. 178 +---------+ 179 +---+Internet | 180 | +---------+ 181 | 182 +----+--+ 183 |DODAG | node:01 184 +---------+Root +----------+ 185 | |6LBR | | 186 | +----+--+ | 187 | | | 188 | | | 189 ... ... ... 190 | | | 191 +-----+-+ +--+---+ +--+---+ 192 |6LR | | | | | 193 +-----+ | | | | | 194 | | 11 | | 12 | | 13 +------+ 195 | +-----+-+ +-+----+ +-+----+ | 196 | | | | | 197 | | | | | 198 | 21 | 22 | 23 | 24 | 25 199 +-+---+ +-+---+ +--+--+ +- --+ +---+-+ 200 |Leaf | | | | | |Leaf| |Leaf | 201 | 6LR | | | | | | 6LN| | 6LR | 202 +-----+ +-----+ +-----+ +----+ +-----+ 204 Figure 2: A reference RPL Topology. 206 The numbers in or above the nodes are there so that they may be 207 referenced in subsequent sections. The leaf marked 6LN (24) is a 208 device which does not speak RPL at all, but uses Router- 209 Advertisements, 6LowPAN DAR/DAC and efficient-ND only to participate 210 in the network. 212 This document is in part motivated by the work that is ongoing at the 213 6TiSCH working group. The 6TiSCH architecture 214 [I-D.ietf-6tisch-architecture] draft explains the network 215 architecture of a 6TiSCH network. This architecture is used for the 216 remainder of this document. 218 The scope of the 6TiSCH Architecture is a Backbone Link that 219 federates multiple LLNs (mesh) as a single IPv6 Multi-Link Subnet. 220 Each LLN in the subnet is anchored at a Backbone Router (6BBR). The 221 Backbone Routers interconnect the LLNs over the Backbone Link and 222 emulate that the LLN nodes are present on the Backbone thus creating 223 a so-called: Multi-Link Subnet. An LLN node can move freely from an 224 LLN anchored at a Backbone Router to another LLN anchored at the same 225 or a different Backbone Router inside the Multi-Link Subnet and 226 conserve its addresses. 228 | 229 +-----+ 230 | | Border Router to the RPL domain 231 | | (may be a RPL virtual root) 232 +-----+ 233 | 234 | Backbone 235 +-------------------+-------------------+ 236 | | | 237 +-----+ +-----+ +-----+ 238 | | Backbone | | Backbone | | Backbone 239 | | router | | router | | router 240 +|---|+ +-|||-+ +-[_]-+ 241 | | PCI-exp / | \ USB | Ethernet 242 ( ) ( ) ( )( )( ) (6LBR == RPL DODAG root) 243 o o o o o o o o o o o o 244 o o o o o o o o o o o o o o o o 245 o o o o o o o o o o 6LR == RPL router) o o 246 o o o o o o o z 247 o o o o o o (6LoWPAN Host) 249 <----------------------- RPL Instance ------------------------> 251 Figure 3: RPL domain architecture 253 4. Use cases 255 In data plane context a combination of RFC6553, RFC6554 and IPv6-in- 256 IPv6 encapsulation is going to be analyzed for the following traffic 257 flows: 259 RPL-aware-leaf to root 261 root to RPL-aware-leaf 263 not-RPL-aware-leaf to root 265 root to not-RPL-aware-leaf 267 RPL-aware-leaf to Internet 269 Internet to RPL-aware-leaf 270 not-RPL-aware-leaf to Internet 272 Internet to not-RPL-aware-leaf 274 RPL-aware-leaf to RPL-aware-leaf 276 RPL-aware-leaf to not-RPL-aware-leaf 278 not-RPL-aware-leaf to RPL-aware-leaf 280 not-RPL-aware-leaf to not-RPL-aware-leaf 282 This document assumes a rule that a Header cannot be inserted or 283 removed on the fly inside an IPv6 packet that is being routed. This 284 is a fundamental precept of the IPv6 architecture as outlined in 285 [RFC2460] is that Extensions may not be added or removed except by 286 the sender or the receiver. 288 A second important thing is that packets with a Hop-by-Hop option 289 which are marked with option type 01 ([RFC2460] section 4.2) must be 290 discarded if received by a host or router which does not understand 291 that option. This means that in general, any packet that leaves the 292 RPL domain of an LLN (or leaves the LLN entirely) is likely to be 293 discarded if it still contains an [RFC6553] RPL Option Header known 294 as the RPI. 296 The combination of these two rules means that the arrangement of 297 headers must be done so that traffic intended to exit the RPL domain 298 can have the RPI option removed prior to leaving the RPL domain. 300 An intermediate router that needs to add a header must encapsulate 301 the packet in an (additional) outer IP header where the new header 302 can be placed. 304 This also means that a Header can only be removed by an intermediate 305 router if it is placed in an encapsulating IPv6 Header, and in that 306 case, the whole encapsulating header must be removed - a replacement 307 may be added. Further, an intermediate router can only remove such 308 an outer header if that outer header has the router as the 309 destination! 311 Both RPI and RH3 headers may be modified by routers on the path of 312 the packet without the need to add to remove an encapsulating header. 313 Both headers were designed with this modification in mind, and both 314 the RPL RH and the RPL option are marked mutable but recoverable, so 315 an IPsec AH security header can be applied across these headers, but 316 it may not secure all the values in those headers. 318 RPI should be present in every single RPL data packet. There is one 319 exception in non-storing mode: when a packet is going down from the 320 route. In a downward non-storing mode, the entire route is written, 321 so there can be no loops by construction, nor any confusion about 322 which forwarding table to use. There may be cases (such as in 323 6tisch) where the instanceID may still be needed to pick an 324 appropriate priority or channel at each hop. 326 The applicability for storing (RPL-SN) and non-Storing (RPL-NSN) 327 modes for the previous cases is showed as follows: 329 In tables, the term "RPL aware leaf" is has been shortened to "Raf", 330 and "not-RPL aware leaf" has been shortened to "~Raf" to make the 331 table fit in available space. 333 The earlier examples are more complete to make sure that the process 334 is clear, while later examples are more consise. 336 5. Storing mode 338 This table summarizes what headers are needed in the following 339 scenarios, and indicates the IPIP header must be inserted on a hop- 340 by-hop basis, and when it can target the destination node directly. 341 There are three possible situations: hop-by-hop necessary (indicated 342 by "hop"), or destination address possible (indicated by "dst"). In 343 all cases hop by hop can be used. In cases where no IPIP header is 344 needed, the column is left blank. 346 +---------------+--------------+--------------+----------+----------+ 347 | Use Case | RPI (RFC | RH3 (RFC | IP-in-IP | IPIP dst | 348 | | 6553) | 6554) | | | 349 +---------------+--------------+--------------+----------+----------+ 350 | Raf to root | Yes | No | No | -- | 351 | root to Raf | Yes | No | No | -- | 352 | root to ~Raf | Yes | No | Yes | hop | 353 | ~Raf to root | Yes | No | Yes | root | 354 | Raf to | Yes | No | Yes | root | 355 | Internet | | | | | 356 | Internet to | Yes | No | Yes | raf | 357 | Raf | | | | | 358 | ~Raf to | Yes | No | Yes | root | 359 | Internet | | | | | 360 | Internet to | Yes | No | Yes | hop | 361 | ~Raf | | | | | 362 | Raf to Raf | Yes | No | No | -- | 363 | Raf to ~Raf | Yes | No | Yes | hop | 364 | ~Raf to Raf | Yes | No | Yes | dst | 365 | ~Raf to ~Raf | Yes | No | Yes | hop | 366 +---------------+--------------+--------------+----------+----------+ 368 Table 1: Headers needed in Storing mode: RPI, RH3, IP-in-IP 369 encapsulation 371 5.1. Example of Flow from RPL-aware-leaf to root 373 As states in Section 16.2 of [RFC6550] a RPL-aware-leaf node does 374 not generally issue DIO messages; a leaf node accepts DIO messages 375 from upstream. (When the inconsistency in routing occurs, a leaf 376 node will generate a DIO with an infinite rank, to fix it). It may 377 issue DAO and DIS messages though it generally ignores DAO and DIS 378 messages. 380 In storing mode, it is suitable to use RFC 6553 (RPI) to send RPL 381 Information instanceID and rank information. 383 In this case the flow comprises: 385 RPL-aware-leaf (6LN) --> 6LR --> 6LR,... --> root (6LBR) 387 Note: In this document 6LRs, 6LBR are always full-fledge RPL routers, 388 and are the RPL root node. 390 The 6LN inserts the RPI header, and send the packet to 6LR which 391 decrement the rank in RPI and send the packet up. When the packet 392 arrives to 6LBR, the RPI is removed and the packet is processed. 394 The RPI header can be removed by the 6LBR because the packet is 395 addressed to the 6LBR. The 6LN must know that it is communicating 396 with the 6LBR in order to be able to make use of this scenario. The 397 6LN can know the address of the 6LBR because it knows the address of 398 the root via the DODAGID in the DIO messages. 400 +-------------------+-----+------+------+ 401 | Header | 6LN | 6LR | 6LBR | 402 +-------------------+-----+------+------+ 403 | Inserted headers | RPI | -- | -- | 404 | Removed headers | -- | -- | RPI | 405 | Re-added headers | -- | -- | -- | 406 | Modified headers | -- | RPI | -- | 407 | Untouched headers | -- | -- | -- | 408 +-------------------+-----+------+------+ 410 Storing: Summary of the use of headers from RPL-aware-leaf to root 412 5.2. Example of Flow from root to RPL-aware-leaf 414 In this case the flow comprises: 416 root (6LBR)--> 6LR --> RPL-aware-leaf (6LN) 418 In this case the 6LBR insert RPI header and send the packet down, the 419 6LR is going to increment the rank in RPI (examines instanceID for 420 multiple tables), the packet is processed in 6LN and RPI removed. 422 No IPIP header is required. 424 +-------------------+------+-------+------+ 425 | Header | 6LBR | 6LR | 6LN | 426 +-------------------+------+-------+------+ 427 | Inserted headers | RPI | -- | -- | 428 | Removed headers | -- | -- | RPI | 429 | Re-added headers | -- | -- | -- | 430 | Modified headers | -- | RPI | -- | 431 | Untouched headers | -- | -- | -- | 432 +-------------------+------+-------+------+ 434 Storing: Summary of the use of headers from root to RPL-aware-leaf 436 5.3. Example of Flow from root to not-RPL-aware-leaf 438 In this case the flow comprises: 440 root (6LBR)--> 6LR --> not-RPL-aware-leaf (6LN) 441 It includes IPv6-in-IPv6 encapsulation to transmit information not 442 related with the RPL domain. In the 6LBR the RPI header is inserted 443 into an IPv6-in-IPv6 header addressed to the last 6LR, which removes 444 the header before pass the packet to the IPv6 node. 446 The question in this scenario is how the root knows how to address 447 the IPv6-in-IPv6 header. It can not know that the destination isn't 448 RPL aware, so it must insert an IPv6 that can be removed on the last 449 RPL aware node. Since the root can not know in a storing network 450 where the last RPL aware node is, the IPv6-in-IPv6 header must added 451 hop-by-hop along the path from root to leaf. 453 An alternative option is to add an attribute in the RPL Target Option 454 to indicate that the target is not RPL aware: future work may explore 455 this possibility. 457 +-------------------+-----------+-----------+------+ 458 | Header | 6LBR | 6LR | IPv6 | 459 +-------------------+-----------+-----------+------+ 460 | Inserted headers | IPIP(RPI) | -- | -- | 461 | Removed headers | -- | IPIP(RPI) | -- | 462 | Re-added headers | -- | -- | -- | 463 | Modified headers | -- | -- | -- | 464 | Untouched headers | -- | -- | -- | 465 +-------------------+-----------+-----------+------+ 467 Storing: Summary of the use of headers from root to not-RPL-aware- 468 leaf 470 5.4. Example of Flow from not-RPL-aware-leaf to root 472 In this case the flow comprises: 474 not-RPL-aware-leaf (6LN) --> 6LR --> root (6LBR) 476 When the packet arrives from IPv6 node to 6LR, the 6LR will insert an 477 RPI header, encapsuladed in a IPv6-in-IPv6 header. The IPv6-in-IPv6 478 header can be addressed to the next hop, or to the root. The root 479 removes the header and process the packet. 481 +-------------------+------+------------+-----------+ 482 | Header | IPv6 | 6LR | 6LBR | 483 +-------------------+------+------------+-----------+ 484 | Inserted headers | -- | IPIP(RPI) | -- | 485 | Removed headers | -- | -- | IPIP(RPI) | 486 | Re-added headers | -- | -- | -- | 487 | Modified headers | -- | -- | -- | 488 | Untouched headers | -- | -- | -- | 489 +-------------------+------+------------+-----------+ 491 Storing: Summary of the use of headers from not-RPL-aware-leaf to 492 root 494 5.5. Example of Flow from RPL-aware-leaf to Internet 496 RPL information from RFC 6553 should not go out to Internet as it 497 will cause the packet to be discarded at the first non-RPI aware 498 router. The 6LBR must be able to take this information out before 499 sending the packet upwards to the Internet. This requires the RPI 500 header be placed in an IPIP header that the root can remove. 502 In this case the flow comprises: 504 RPL-aware-leaf (6LN) --> 6LR --> root (6LBR) --> Internet 506 The 6LN will insert the RPI in a IPv6-in-IPv6 in a outer header, 507 which may be addressed to the 6LBR (root), or alternatively, it could 508 be addressed hop-by-hop. 510 +-------------------+-----------+------+-----------+----------+ 511 | Header | 6LN | 6LR | 6LBR | Internet | 512 +-------------------+-----------+------+-----------+----------+ 513 | Inserted headers | IPIP(RPI) | -- | -- | -- | 514 | Removed headers | -- | -- | IPIP(RPI) | -- | 515 | Re-added headers | -- | -- | -- | -- | 516 | Modified headers | -- | RPI | -- | -- | 517 | Untouched headers | -- | -- | -- | -- | 518 +-------------------+-----------+------+-----------+----------+ 520 Storing: Summary of the use of headers from RPL-aware-leaf to 521 Internet 523 5.6. Example of Flow from Internet to RPL-aware-leaf 525 In this case the flow comprises: 527 Internet --> root (6LBR) --> 6LR --> RPL-aware-leaf (6LN) 528 When the packet arrives from Internet to 6LBR the RPI header is added 529 in a outer IPv6-in-IPv6 header and send to 6LR, which modifies the 530 rank in the RPI. When the packet arrives 6LN the RPI header is 531 removed and the packet processed. 533 +-------------------+----------+------------+------+------------+ 534 | Header | Internet | 6LBR | 6LR | 6LN | 535 +-------------------+----------+------------+------+------------+ 536 | Inserted headers | -- | IPIP(RPI) | -- | -- | 537 | Removed headers | -- | -- | -- | IPIP(RPI) | 538 | Re-added headers | -- | -- | -- | -- | 539 | Modified headers | -- | -- | RPI | -- | 540 | Untouched headers | -- | -- | -- | -- | 541 +-------------------+----------+------------+------+------------+ 543 Storing: Summary of the use of headers from Internet to RPL-aware- 544 leaf 546 5.7. Example of Flow from not-RPL-aware-leaf to Internet 548 In this case the flow comprises: 550 not-RPL-aware-leaf (6LN) --> 6LR --> root (6LBR) --> Internet 552 The 6LR node will add an IPIP(RPI) header addressed either to the 553 root, or hop-by-hop such that the root can remove the RPI header 554 before passing upwards. 556 The originating node will ideally leave the IPv6 flow label as zero 557 so that it can be better compressed through the LLN, and the 6LBR 558 will set the flow label to a non-zero value when sending to the 559 Internet. 561 +-------------------+------+------------+------------+----------+ 562 | Header | 6LN | 6LR | 6LBR | Internet | 563 +-------------------+------+------------+------------+----------+ 564 | Inserted headers | -- | IPIP(RPI) | -- | -- | 565 | Removed headers | -- | -- | IPIP(RPI) | -- | 566 | Re-added headers | -- | -- | -- | -- | 567 | Modified headers | -- | -- | -- | -- | 568 | Untouched headers | -- | -- | -- | -- | 569 +-------------------+------+------------+------------+----------+ 571 Storing: Summary of the use of headers from not-RPL-aware-leaf to 572 Internet 574 5.8. Example of Flow from Internet to non-RPL-aware-leaf 576 In this case the flow comprises: 578 Internet --> root (6LBR) --> 6LR --> not-RPL-aware-leaf (6LN) 580 The 6LBR will have to add an RPI header within an IPIP header. The 581 IPIP will need to be addressed hop-by-hop along the path as in 582 storing mode, the 6LBR has no idea if the 6LN is RPL aware or not, 583 nor what the closest attached 6LR node is. 585 The 6LBR MAY set the flow label on the inner IPIP header to zero in 586 order to aid in compression, as the packet will not emerge again from 587 the LLN. 589 +-------------------+----------+------------+------------+------+ 590 | Header | Internet | 6LBR | 6LR | IPv6 | 591 +-------------------+----------+------------+------------+------+ 592 | Inserted headers | -- | IPIP(RPI) | -- | -- | 593 | Removed headers | -- | -- | IPIP(RPI) | -- | 594 | Re-added headers | -- | -- | -- | -- | 595 | Modified headers | -- | -- | -- | -- | 596 | Untouched headers | -- | -- | -- | -- | 597 +-------------------+----------+------------+------------+------+ 599 Storing: Summary of the use of headers from Internet to non-RPL- 600 aware-leaf 602 5.9. Example of Flow from RPL-aware-leaf to RPL-aware-leaf 604 In [RFC6550] RPL allows a simple one-hop optimization for both 605 storing and non-storing networks. A node may send a packet destined 606 to a one-hop neighbor directly to that node. Section 9 in [RFC6550]. 608 In this case the flow comprises: 610 6LN --> 6LR --> common parent (6LR) --> 6LR --> 6LN 612 This case is assumed in the same RPL Domain. In the common parent, 613 the direction of RPI is changed (from increasing to decreasing the 614 rank). 616 While the 6LR nodes will update the RPI, no node needs to add or 617 remove the RPI, so no IPIP headers are necessary. The ability to do 618 this depends upon the sending know that the destination is: a) inside 619 the LLN, and b) RPL capable. 621 The sender can determine if the destination is inside the LLN by 622 looking if the destination address is matched by the DIO's PIO 623 option. This check may be modified by the use of backbone routers, 624 but in this case it is assumed that the backbone routers are RPL 625 capable and so can process the RPI header correctly. 627 The other check, that the destination is RPL capable is not currently 628 discernible by the sender. This information is necessary to 629 distinguish this test case from Section 5.10. 631 +-------------+-------+---------------+---------------+-----+-------+ 632 | Header | 6LN | 6LR | 6LR (common | 6LR | 6LN | 633 | | src | | parent) | | dst | 634 +-------------+-------+---------------+---------------+-----+-------+ 635 | Inserted | RPI | -- | -- | -- | -- | 636 | headers | | | | | | 637 | Removed | -- | -- | -- | -- | RPI | 638 | headers | | | | | | 639 | Re-added | -- | -- | -- | -- | -- | 640 | headers | | | | | | 641 | Modified | -- | RPI | RPI | -- | -- | 642 | headers | | (decreasing | (increasing | | | 643 | | | rank) | rank) | | | 644 | Untouched | -- | -- | -- | -- | -- | 645 | headers | | | | | | 646 +-------------+-------+---------------+---------------+-----+-------+ 648 Storing: Summary of the use of headers for RPL-aware-leaf to RPL- 649 aware-leaf 651 5.10. Example of Flow from RPL-aware-leaf to non-RPL-aware-leaf 653 In this case the flow comprises: 655 6LN --> 6LR --> common parent (6LR) --> 6LR --> not-RPL-aware 6LN 657 The sender, being aware out of band, that the receiver is not RPL 658 aware, sends adds an RPI header inside an IPIP header. The IPIP 659 header needs to be addressed on a hop-by-hop basis so that the last 660 6LR can remove the RPI header. 662 ,---. 663 / \ 664 ( 6LR2 ) IP3,RPI,IP,ULP 665 ,-" . 666 ,-" `---' `. 667 ,' `. 668 ,---. ,-" `,---. 669 / +" / \ 670 ( 6LR1 ) Remove the IP3,RPI( 6LR3 ) 671 \ / \ / 672 /---' `---'| 673 / IP2,RPI,IP,ULP \ 674 / | 675 / \ 676 ,---+-. | 677 / \ +--+----+ 678 ( 6LN ) | | 679 \ / | IPv6 | IP,ULP 680 `-----' | | 681 IP1,RPI,IP,ULP +-------+ 683 Figure 4: Solution IPv6-in-IPv6 in each hop 685 Alternatively, if the definition of the Option Type field of RPL 686 Option '01' were changed so that it isn't a "discard if not 687 recognized", then no IPIP header would be necessary. This change is 688 an incompatible on-the-wire change and would require some kind of 689 flag day, possibly a change that is done simultaenously with an 690 updated 6LoRH compress. 692 +-----------+-----------+-----------+------------+-----------+------+ 693 | Header | 6LN | 6LR | 6LR | 6LR | IPv6 | 694 | | | | (common | | | 695 | | | | parent) | | | 696 +-----------+-----------+-----------+------------+-----------+------+ 697 | Inserted | IPIP(RPI) | -- | -- | -- | -- | 698 | headers | | | | | | 699 | Removed | -- | -- | -- | IPIP(RPI) | -- | 700 | headers | | | | | | 701 | Re-added | -- | -- | -- | -- | -- | 702 | headers | | | | | | 703 | Modified | -- | IPIP(RPI) | IPIP(RPI) | -- | -- | 704 | headers | | | | | | 705 | Untouched | -- | -- | -- | -- | -- | 706 | headers | | | | | | 707 +-----------+-----------+-----------+------------+-----------+------+ 709 Storing: Summary of the use of headers from RPL-aware-leaf to not- 710 RPL-aware-leaf 712 5.11. Example of Flow from not-RPL-aware-leaf to RPL-aware-leaf 714 In this case the flow comprises: 716 not-RPL-aware 6LN --> 6LR --> common parent (6LR) --> 6LR --> 6LN 718 The 6LR receives the packet from the the IPv6 node and inserts and 719 the RPI header encapsulated in IPv6-in-IPv6 header. The IPIP header 720 could be addresses to the 6LN if the destination is known to the RPL 721 aware, otherwise must send the packet using a hop-by-hop IPIP header. 722 Similar considerations apply from section Section 5.10. 724 +-----------+------+-----------+------------+-----------+-----------+ 725 | Header | IPv6 | 6LR | common | 6LR | 6LN | 726 | | | | parent | | | 727 | | | | (6LR) | | | 728 +-----------+------+-----------+------------+-----------+-----------+ 729 | Inserted | -- | IPIP(RPI) | -- | -- | -- | 730 | headers | | | | | | 731 | Removed | -- | -- | -- | -- | IPIP(RPI) | 732 | headers | | | | | | 733 | Re-added | -- | -- | -- | -- | -- | 734 | headers | | | | | | 735 | Modified | -- | -- | IPIP(RPI) | IPIP(RPI) | -- | 736 | headers | | | | | | 737 | Untouched | -- | -- | -- | -- | -- | 738 | headers | | | | | | 739 +-----------+------+-----------+------------+-----------+-----------+ 741 Storing: Summary of the use of headers from not-RPL-aware-leaf to 742 RPL-aware-leaf 744 5.12. Example of Flow from not-RPL-aware-leaf to not-RPL-aware-leaf 746 In this case the flow comprises: 748 not-RPL-aware 6LN (IPv6 node)--> 6LR --> root (6LBR) --> 6LR --> not- 749 RPL-aware 6LN (IPv6 node) 751 This flow combines the problems of the two previous sections. There 752 is no choice at the first 6LR: it must insert an RPI, and to do that 753 it must add an IPIP header. That IPIP header must be addressed on a 754 hop-by-hop basis. 756 +-------------+--------+-----------+-----------+-----------+--------+ 757 | Header | IPv6 | 6LR | 6LR | 6LR | IPv6 | 758 | | src | | (common | | dst | 759 | | | | parent) | | | 760 +-------------+--------+-----------+-----------+-----------+--------+ 761 | Inserted | -- | IPIP(RPI) | -- | -- | -- | 762 | headers | | | | | | 763 | Removed | -- | -- | -- | IPIP(RPI) | -- | 764 | headers | | | | | | 765 | Re-added | -- | -- | -- | -- | -- | 766 | headers | | | | | | 767 | Modified | -- | -- | -- | -- | -- | 768 | headers | | | | | | 769 | Untouched | -- | -- | -- | -- | -- | 770 | headers | | | | | | 771 +-------------+--------+-----------+-----------+-----------+--------+ 773 Storing: Summary of the use of headers from not-RPL-aware-leaf to 774 not-RPL-aware-leaf 776 6. Non Storing mode 778 +------------------+------+------+-------+-----------+ 779 | Use Case | RPI | RH3 | IPIP | IPIP dst | 780 +------------------+------+------+-------+-----------+ 781 | Raf to root | Yes | No | No | -- | 782 | root to Raf | Yes | Yes | No | -- | 783 | root to ~Raf | No | Yes | Yes | -- | 784 | ~Raf to root | Yes | No | Yes | root | 785 | Raf to Internet | Yes | No | Yes | root | 786 | Internet to Raf | opt | Yes | Yes | dst | 787 | ~Raf to Internet | Yes | No | Yes | root | 788 | Internet to ~Raf | opt | Yes | Yes | 6LR | 789 | Raf to Raf | Yes | Yes | Yes | root/dst | 790 | Raf to ~Raf | Yes | Yes | Yes | root/6LN | 791 | ~Raf to Raf | Yes | Yes | Yes | root/6LN | 792 | ~Raf to ~Raf | Yes | Yes | Yes | root/6LN | 793 +------------------+------+------+-------+-----------+ 795 Table 2: Headers needed in Non-Storing mode: RPI, RH3, IP-in-IP 796 encapsulation 798 6.1. Example of Flow from RPL-aware-leaf to root 800 In non-storing mode the leaf node uses default routing to send 801 traffic to the root. The RPI header must be included to avoid/detect 802 loops. 804 RPL-aware-leaf (6LN) --> 6LR --> root (6LBR) 806 This situation is the same case as storing mode. 808 +-------------------+-----+-----+------+ 809 | Header | 6LN | 6LR | 6LBR | 810 +-------------------+-----+-----+------+ 811 | Inserted headers | RPI | -- | -- | 812 | Removed headers | -- | -- | RPI | 813 | Re-added headers | -- | -- | RPI | 814 | Modified headers | -- | -- | -- | 815 | Untouched headers | -- | -- | -- | 816 +-------------------+-----+-----+------+ 818 Non Storing: Summary of the use of headers from RPL-aware-leaf to 819 root 821 6.2. Example of Flow from root to RPL-aware-leaf 823 In this case the flow comprises: 825 root (6LBR)--> 6LR --> RPL-aware-leaf (6LN) 827 The 6LBR will insert an RH3, and may optionally insert an RPI header. 828 No IPIP header is necessary as the traffic originates with an RPL 829 aware node. 831 +-------------------+-----------------+------+----------+ 832 | Header | 6LBR | 6LR | 6LN | 833 +-------------------+-----------------+------+----------+ 834 | Inserted headers | (opt: RPI), RH3 | -- | -- | 835 | Removed headers | -- | -- | RH3,RPI | 836 | Re-added headers | -- | -- | -- | 837 | Modified headers | -- | RH3 | -- | 838 | Untouched headers | -- | -- | -- | 839 +-------------------+-----------------+------+----------+ 841 Non Storing: Summary of the use of headers from root to RPL-aware- 842 leaf 844 6.3. Example of Flow from root to not-RPL-aware-leaf 846 In this case the flow comprises: 848 root (6LBR)--> 6LR --> not-RPL-aware-leaf (IPv6 node) 850 In 6LBR the RH3 is added, and modified in 6LR where it is fully 851 consumed, but left there. If the RPI is left present, the IPv6 node 852 which does not understand it will drop it, therefore the RPI should 853 be removed before reaching the IPv6-only node. To permit removal, an 854 IPIP header (hop-by-hop) or addressed to the last 6LR is necessary. 855 Due the complete knowledge of the topology at the root, the 6LBR is 856 able to address the IPIP header to the last 6LR. 858 Omitting the RPI entirely is therefore a better solution, as no IPIP 859 header is necessary. 861 +-------------------+------+-----+------+ 862 | Header | 6LBR | 6LR | IPv6 | 863 +-------------------+------+-----+------+ 864 | Inserted headers | RH3 | -- | -- | 865 | Removed headers | -- | -- | -- | 866 | Re-added headers | -- | -- | -- | 867 | Modified headers | -- | RH3 | -- | 868 | Untouched headers | -- | -- | -- | 869 +-------------------+------+-----+------+ 871 Non Storing: Summary of the use of headers from root to not-RPL- 872 aware-leaf 874 6.4. Example of Flow from not-RPL-aware-leaf to root 876 In this case the flow comprises: 878 IPv6-node --> 6LR1 --> 6LR2 --> root (6LBR) 880 In this case the RPI is added by the first 6LR, encapsulated in an 881 IPIP header, and is not modified in the followings 6LRs. The RPI and 882 entire packet is consumed by the root. 884 +-------------------+------+------------+------+------------+ 885 | Header | IPv6 | 6LR1 | 6LR2 | 6LBR | 886 +-------------------+------+------------+------+------------+ 887 | Inserted headers | -- | IPIP(RPI) | -- | -- | 888 | Removed headers | -- | -- | -- | IPIP(RPI) | 889 | Re-added headers | -- | -- | -- | -- | 890 | Modified headers | -- | -- | -- | -- | 891 | Untouched headers | -- | IPIP(RPI) | -- | -- | 892 +-------------------+------+------------+------+------------+ 894 Non Storing: Summary of the use of headers from not-RPL-aware-leaf to 895 root 897 6.5. Example of Flow from RPL-aware-leaf to Internet 899 In this case the flow comprises: 901 RPL-aware-leaf (6LN) --> 6LR --> root (6LBR) --> Internet 903 This case requires that the RPI be added, but remoted by the 6LBR. 904 The 6LN must therefore add the RPI inside an IPIP header, addressed 905 to the root. This case is identical to storing-mode case. 907 The IPv6 flow label should be set to zero to aid in compression, and 908 the 6LBR will set it to a non-zero value when sending towards the 909 Internet. 911 +-------------------+-----------+------+------------+----------+ 912 | Header | 6LN | 6LR | 6LBR | Internet | 913 +-------------------+-----------+------+------------+----------+ 914 | Inserted headers | IPIP(RPI) | -- | -- | -- | 915 | Removed headers | -- | -- | IPIP(RPI) | -- | 916 | Re-added headers | -- | -- | -- | -- | 917 | Modified headers | -- | -- | -- | -- | 918 | Untouched headers | -- | RPI | -- | -- | 919 +-------------------+-----------+------+------------+----------+ 921 Non Storing: Summary of the use of headers from RPL-aware-leaf to 922 Internet 924 6.6. Example of Flow from Internet to RPL-aware-leaf 926 In this case the flow comprises: 928 Internet --> root (6LBR) --> 6LR --> RPL-aware-leaf (6LN) 930 The 6LBR must add an RH3 header. As the 6LBR will know the path and 931 address of the target not, it can address the IPIP header to that 932 node. The 6LBR will zero the flow label upon entry in order to aid 933 compression. 935 The RPI may be added or not. 937 +----------------+----------+--------------------+------------+-----+ 938 | Header | Internet | 6LBR | 6LR | 6LN | 939 +----------------+----------+--------------------+------------+-----+ 940 | Inserted | -- | IPIP(RH3,opt:RPI) | -- | -- | 941 | headers | | | | | 942 | Removed | -- | -- | IPIP(RH3) | -- | 943 | headers | | | | | 944 | Re-added | -- | -- | -- | -- | 945 | headers | | | | | 946 | Modified | -- | -- | IPIP(RH3) | -- | 947 | headers | | | | | 948 | Untouched | -- | -- | -- | -- | 949 | headers | | | | | 950 +----------------+----------+--------------------+------------+-----+ 952 Non Storing: Summary of the use of headers from Internet to RPL- 953 aware-leaf 955 6.7. Example of Flow from not-RPL-aware-leaf to Internet 957 In this case the flow comprises: 959 not-RPL-aware-leaf (6LN) --> 6LR --> root (6LBR) --> Internet 961 In this case the flow label is recommended to be zero in the IPv6 962 node. As RPL headers are added in the IPv6 node, the first 6LN will 963 add an RPI header inside a new IPIP header. The IPIP header will be 964 addressed to the root. This case is identical to the storing-mode 965 case. 967 +-------------------+------+-----------+------------+----------+ 968 | Header | IPv6 | 6LR | 6LBR | Internet | 969 +-------------------+------+-----------+------------+----------+ 970 | Inserted headers | -- | IPIP(RPI) | -- | -- | 971 | Removed headers | -- | -- | IPIP(RPI) | -- | 972 | Re-added headers | -- | -- | -- | -- | 973 | Modified headers | -- | -- | -- | -- | 974 | Untouched headers | -- | -- | -- | -- | 975 +-------------------+------+-----------+------------+----------+ 977 Non Storing: Summary of the use of headers from not-RPL-aware-leaf to 978 Internet 980 6.8. Example of Flow from Internet to non-RPL-aware-leaf 982 In this case the flow comprises: 984 Internet --> root (6LBR) --> 6LR --> not-RPL-aware-leaf (6LN) 985 The 6LBR must add an RH3 header inside an IPIP header. The 6LBR will 986 know the path, and will recognize that the final node is not an RPL 987 capable node as it will have received the connectivity DAO from the 988 nearest 6LR. The 6LBR can therefore make the IPIP header destination 989 be the last 6LR. The 6LBR will zero the flow label upon entry in 990 order to aid compression. 992 +--------------+----------+-------------------+--------------+------+ 993 | Header | Internet | 6LBR | 6LR | IPv6 | 994 +--------------+----------+-------------------+--------------+------+ 995 | Inserted | -- | IPIP(RH3,opt:RPI) | -- | -- | 996 | headers | | | | | 997 | Removed | -- | -- | IPIP(RH3, | -- | 998 | headers | | | RPI) | | 999 | Re-added | -- | -- | -- | -- | 1000 | headers | | | | | 1001 | Modified | -- | -- | -- | -- | 1002 | headers | | | | | 1003 | Untouched | -- | -- | -- | -- | 1004 | headers | | | | | 1005 +--------------+----------+-------------------+--------------+------+ 1007 NonStoring: Summary of the use of headers from Internet to non-RPL- 1008 aware-leaf 1010 6.9. Example of Flow from RPL-aware-leaf to RPL-aware-leaf 1012 In this case the flow comprises: 1014 6LN --> 6LR --> root (6LBR) --> 6LR --> 6LN 1016 This case involves only nodes in same RPL Domain. The originating 1017 node will add an RPI header to the original packet, and send the 1018 packet upwards. 1020 The originating node could put the RPI into an IPIP header addressed 1021 to the root, so that the 6LBR can remove that header. 1023 The 6LBR will need to insert an RH3 header, which requires that it 1024 add an IPIP header. It may be able to remove the RPI if it was 1025 contained in an IPIP header addressed to it. Otherwise, there may be 1026 an RPI header buried inside the inner IP header, which should get 1027 ignored. 1029 Networks that use the RPL P2P extension [RFC6997] are essentially 1030 non-storing DODAGs and fall into this scenario. 1032 +----------------+-----------+----------------+-----+---------------+ 1033 | Header | 6LN src | 6LBR | 6LR | 6LN dst | 1034 +----------------+-----------+----------------+-----+---------------+ 1035 | Inserted | IPIP(RPI) | IPIP(RH3 to | -- | -- | 1036 | headers | | 6LN,RPI) | | | 1037 | Removed | -- | -- | -- | IPIP(RH3,RPI) | 1038 | headers | | | | | 1039 | Re-added | -- | -- | -- | -- | 1040 | headers | | | | | 1041 | Modified | -- | -- | -- | -- | 1042 | headers | | | | | 1043 | Untouched | -- | -- | -- | -- | 1044 | headers | | | | | 1045 +----------------+-----------+----------------+-----+---------------+ 1047 Non Storing: Summary of the use of headers for RPL-aware-leaf to RPL- 1048 aware-leaf 1050 6.10. Example of Flow from RPL-aware-leaf to not-RPL-aware-leaf 1052 In this case the flow comprises: 1054 6LN --> 6LR --> root (6LBR) --> 6LR --> not-RPL-aware 6LN 1056 As in the previous case, the 6LN will insert an RPI header which MUST 1057 be in an IPIP header addressed to the root so that the 6LBR can 1058 remove this RPI. The 6LBR will then insert an RH3 inside a new IPIP 1059 header addressed to the 6LN above the destination node. 1061 +---------------+-----------+---------------+----------------+------+ 1062 | Header | 6LN | 6LBR | 6LR | IPv6 | 1063 +---------------+-----------+---------------+----------------+------+ 1064 | Inserted | IPIP(RPI) | IPIP(RH3, opt | -- | -- | 1065 | headers | | RPI) | | | 1066 | Removed | -- | IPIP(RPI) | IPIP(RH3, opt | -- | 1067 | headers | | | RPI) | | 1068 | Re-added | -- | -- | -- | -- | 1069 | headers | | | | | 1070 | Modified | -- | -- | -- | -- | 1071 | headers | | | | | 1072 | Untouched | -- | -- | -- | -- | 1073 | headers | | | | | 1074 +---------------+-----------+---------------+----------------+------+ 1076 Non Storing: Summary of the use of headers from RPL-aware-leaf to 1077 not-RPL-aware-leaf 1079 6.11. Example of Flow from not-RPL-aware-leaf to RPL-aware-leaf 1081 In this case the flow comprises: 1083 not-RPL-aware 6LN --> 6LR --> root (6LBR) --> 6LR --> 6LN 1085 This scenario is mostly identical to the previous one. The RPI is 1086 added by the first 6LR inside an IPIP header addressed to the root. 1087 The 6LBR will remove this RPI, and add it's own IPIP header 1088 containing an RH3 header. 1090 +-------------------+------+------------+-----------+------------+ 1091 | Header | IPv6 | 6LR | 6LBR | 6LN | 1092 +-------------------+------+------------+-----------+------------+ 1093 | Inserted headers | -- | IPIP(RPI) | IPIP(RH3) | -- | 1094 | Removed headers | -- | IPIP(RPI) | -- | IPIP(RH3) | 1095 | Re-added headers | -- | -- | -- | -- | 1096 | Modified headers | -- | -- | -- | -- | 1097 | Untouched headers | -- | -- | -- | -- | 1098 +-------------------+------+------------+-----------+------------+ 1100 Non Storing: Summary of the use of headers from not-RPL-aware-leaf to 1101 RPL-aware-leaf 1103 6.12. Example of Flow from not-RPL-aware-leaf to not-RPL-aware-leaf 1105 In this case the flow comprises: 1107 not-RPL-aware 6LN --> 6LR --> root (6LBR) --> 6LR --> not-RPL-aware 1108 6LN 1110 This scenario is the combination of the previous two cases. 1112 +--------------+------+-----------+-----------+--------------+------+ 1113 | Header | IPv6 | 6LR | 6LBR | 6LR | IPv6 | 1114 +--------------+------+-----------+-----------+--------------+------+ 1115 | Inserted | -- | IPIP(RPI) | IPIP(RH3) | -- | -- | 1116 | headers | | | | | | 1117 | Removed | -- | -- | IPIP(RPI) | IPIP(RH3, | -- | 1118 | headers | | | | opt RPI) | | 1119 | Re-added | -- | -- | -- | -- | -- | 1120 | headers | | | | | | 1121 | Modified | -- | -- | -- | -- | -- | 1122 | headers | | | | | | 1123 | Untouched | -- | -- | -- | -- | -- | 1124 | headers | | | | | | 1125 +--------------+------+-----------+-----------+--------------+------+ 1127 Non Storing: Summary of the use of headers from not-RPL-aware-leaf to 1128 not-RPL-aware-leaf 1130 7. Observations about the problem 1132 7.1. Storing mode 1134 In the completely general storing case, which includes not-RPL aware 1135 leaf nodes, it is not possible for a sending node to know if the 1136 destination is RPL aware, and therefore it must always use hop-by-hop 1137 IPIP encapsulation, and it can never omit the IPIP encapsulation. 1138 See table Table 1 1140 The simplest fully general stiaution for storing mode is to always 1141 put in hop-by-hop IPIP headers. [I-D.ietf-roll-routing-dispatch] 1142 shows that this hop-by-hop IPIP header can be compressed down to 1143 {TBD} bytes. 1145 There are potential significant advantages to having a single code 1146 path that always processes IPIP headers with no options. 1148 If all RPL aware nodes can be told/configured that there are no non- 1149 RPL aware leaf nodes, then the only case where an IPIP header is 1150 needed is when communicating outside the LLN. The 6LBR knows well 1151 when the communication is from the outside, and the 6LN can tell by 1152 comparing the destination address to the prefix provided in the PIO. 1153 If it is known that there are no communications outside the RPL 1154 domain (noting that the RPL domain may well extend to outside the 1155 LLN), then RPI headers can be included in all packets, and IPIP 1156 headers are *never* needed. This may be significantly advantageous 1157 in relatively closed systems such as in building or industrial 1158 automation. Again, there are advantages to having a single code 1159 path. 1161 In order to support the above two cases with full generality, the 1162 different situations (always do IPIP vs never use IPIP) should be 1163 signaled in the RPL protocol itself. 1165 7.2. Non-Storing mode 1167 This the non-storing case, dealing with non-RPL aware leaf nodes is 1168 much easier as the 6LBR (DODAG root) has complete knowledge about the 1169 connectivity of all nodes, and all traffic flows through the root 1170 node. 1172 The 6LBR can recognize non-RPL aware leaf nodes because it will 1173 receive a DAO about that node from the 6LN immediately above that 1174 node. This means that the non-storing mode case can avoid ever using 1175 hop-by-hop IPIP headers. 1177 It is unclear what it would mean for an RH3 header to be present in a 1178 hop-by-hop IPIP header. The receiving node ought to consume the IPIP 1179 header, and therefore consume the RH3 as well, and then attempt to 1180 send the packet again. But intermediate 6LN nodes would not know how 1181 to forward the packet, so the RH3 would need to be retained. This is 1182 a new kind of IPv6 packet processing. Therefore it may be that on 1183 the outbound leg of non-storing RPL networks, that hop-by-hop IPIP 1184 header can NOT be used. 1186 [I-D.ietf-roll-routing-dispatch] shows how the destination=root, and 1187 destination=6LN IPIP header can be compressed down to {TBD} bytes. 1189 Unlike in the storing mode case, there are no need for all nodes to 1190 know about the existence of non-RPL aware nodes. Only the 6LBR needs 1191 to change when there are non-RPL aware nodes. Further, in the non- 1192 storing case, the 6LBR is informed by the DAOs when there are non-RPL 1193 aware nodes. 1195 8. 6LoRH Compression cases 1197 The [I-D.ietf-roll-routing-dispatch] proposes a compression method 1198 for RPI, RH3 and IPv6-in-IPv6. 1200 In Storing Mode, for the examples of Flow from RPL-aware-leaf to non- 1201 RPL-aware-leaf and non-RPL-aware-leaf to non-RPL-aware-leaf comprise 1202 an IP-in-IP and RPI compression headers. The type of this case is 1203 critical since IP-in-IP is encapsulating a RPI header. 1205 +--+-----+---+--------------+-----------+-------------+-------------+ 1206 |1 | 0|0 |TSE| 6LoRH Type 6 | Hop Limit | RPI - 6LoRH | LOWPAN IPHC | 1207 +--+-----+---+--------------+-----------+-------------+-------------+ 1209 Figure 5: Critical IP-in-IP (RPI). 1211 9. IANA Considerations 1213 There are no IANA considerations related to this document. 1215 10. Security Considerations 1217 The security considerations covering of [RFC6553] and [RFC6554] apply 1218 when the packets get into RPL Domain. 1220 11. Acknowledgments 1222 This work is partially funded by the FP7 Marie Curie Initial Training 1223 Network (ITN) METRICS project (grant agreement No. 607728). 1225 The authors would like to acknowledge the review, feedback, and 1226 comments of Thomas Watteyne, Xavier Vilajosana, Robert Cragie and 1227 Simon Duquennoy. 1229 12. References 1231 12.1. Normative References 1233 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1234 Requirement Levels", BCP 14, RFC 2119, 1235 DOI 10.17487/RFC2119, March 1997, 1236 . 1238 [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 1239 (IPv6) Specification", RFC 2460, December 1998. 1241 [RFC6550] Winter, T., Ed., Thubert, P., Ed., Brandt, A., Hui, J., 1242 Kelsey, R., Levis, P., Pister, K., Struik, R., Vasseur, 1243 JP., and R. Alexander, "RPL: IPv6 Routing Protocol for 1244 Low-Power and Lossy Networks", RFC 6550, 1245 DOI 10.17487/RFC6550, March 2012, 1246 . 1248 [RFC6553] Hui, J. and JP. Vasseur, "The Routing Protocol for Low- 1249 Power and Lossy Networks (RPL) Option for Carrying RPL 1250 Information in Data-Plane Datagrams", RFC 6553, 1251 DOI 10.17487/RFC6553, March 2012, 1252 . 1254 [RFC6554] Hui, J., Vasseur, JP., Culler, D., and V. Manral, "An IPv6 1255 Routing Header for Source Routes with the Routing Protocol 1256 for Low-Power and Lossy Networks (RPL)", RFC 6554, 1257 DOI 10.17487/RFC6554, March 2012, 1258 . 1260 12.2. Informative References 1262 [I-D.ietf-6tisch-architecture] 1263 Thubert, P., "An Architecture for IPv6 over the TSCH mode 1264 of IEEE 802.15.4", draft-ietf-6tisch-architecture-09 (work 1265 in progress), November 2015. 1267 [I-D.ietf-roll-routing-dispatch] 1268 Thubert, P., Bormann, C., Toutain, L., and R. Cragie, 1269 "6LoWPAN Routing Header", draft-ietf-roll-routing- 1270 dispatch-00 (work in progress), March 2016. 1272 [RFC6997] Goyal, M., Ed., Baccelli, E., Philipp, M., Brandt, A., and 1273 J. Martocci, "Reactive Discovery of Point-to-Point Routes 1274 in Low-Power and Lossy Networks", RFC 6997, 1275 DOI 10.17487/RFC6997, August 2013, 1276 . 1278 [RFC7102] Vasseur, JP., "Terms Used in Routing for Low-Power and 1279 Lossy Networks", RFC 7102, DOI 10.17487/RFC7102, January 1280 2014, . 1282 [Second6TischPlugtest] 1283 "2nd 6Tisch Plugtest", . 1286 Authors' Addresses 1288 Maria Ines Robles 1289 Ericsson 1290 Hirsalantie 11 1291 Jorvas 02420 1292 Finland 1294 Email: maria.ines.robles@ericsson.com 1295 Michael C. Richardson 1296 Sandelman Software Works 1297 470 Dawson Avenue 1298 Ottawa, ON K1Z 5V7 1299 CA 1301 Email: mcr+ietf@sandelman.ca 1302 URI: http://www.sandelman.ca/ 1304 Pascal Thubert 1305 Cisco Systems, Inc 1306 Village d'Entreprises Green Side 400, Avenue de Roumanille 1307 Batiment T3, Biot - Sophia Antipolis 06410 1308 France 1310 Email: pthubert@cisco.com