idnits 2.17.1 draft-ietf-sidr-rpki-rtr-21.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 abstract seems to contain references ([I-D.ietf-sidr-arch]), which it shouldn't. Please replace those with straight textual mentions of the documents in question. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (December 17, 2011) is 4514 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-10) exists of draft-ietf-sidr-pfx-validate-03 ** Obsolete normative reference: RFC 2385 (Obsoleted by RFC 5925) ** Obsolete normative reference: RFC 5226 (Obsoleted by RFC 8126) ** Obsolete normative reference: RFC 5246 (Obsoleted by RFC 8446) Summary: 4 errors (**), 0 flaws (~~), 2 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group R. Bush 3 Internet-Draft Internet Initiative Japan 4 Intended status: Standards Track R. Austein 5 Expires: June 19, 2012 Dragon Research Labs 6 December 17, 2011 8 The RPKI/Router Protocol 9 draft-ietf-sidr-rpki-rtr-21 11 Abstract 13 In order to formally validate the origin ASs of BGP announcements, 14 routers need a simple but reliable mechanism to receive RPKI 15 [I-D.ietf-sidr-arch] prefix origin data from a trusted cache. This 16 document describes a protocol to deliver validated prefix origin data 17 to routers. 19 Requirements Language 21 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 22 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 23 document are to be interpreted as described in [RFC2119]. 25 Status of this Memo 27 This Internet-Draft is submitted in full conformance with the 28 provisions of BCP 78 and BCP 79. 30 Internet-Drafts are working documents of the Internet Engineering 31 Task Force (IETF). Note that other groups may also distribute 32 working documents as Internet-Drafts. The list of current Internet- 33 Drafts is at http://datatracker.ietf.org/drafts/current/. 35 Internet-Drafts are draft documents valid for a maximum of six months 36 and may be updated, replaced, or obsoleted by other documents at any 37 time. It is inappropriate to use Internet-Drafts as reference 38 material or to cite them other than as "work in progress." 40 This Internet-Draft will expire on June 19, 2012. 42 Copyright Notice 44 Copyright (c) 2011 IETF Trust and the persons identified as the 45 document authors. All rights reserved. 47 This document is subject to BCP 78 and the IETF Trust's Legal 48 Provisions Relating to IETF Documents 49 (http://trustee.ietf.org/license-info) in effect on the date of 50 publication of this document. Please review these documents 51 carefully, as they describe your rights and restrictions with respect 52 to this document. Code Components extracted from this document must 53 include Simplified BSD License text as described in Section 4.e of 54 the Trust Legal Provisions and are provided without warranty as 55 described in the Simplified BSD License. 57 Table of Contents 59 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 60 2. Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 61 3. Deployment Structure . . . . . . . . . . . . . . . . . . . . . 4 62 4. Operational Overview . . . . . . . . . . . . . . . . . . . . . 4 63 5. Protocol Data Units (PDUs) . . . . . . . . . . . . . . . . . . 5 64 5.1. Serial Notify . . . . . . . . . . . . . . . . . . . . . . 5 65 5.2. Serial Query . . . . . . . . . . . . . . . . . . . . . . . 6 66 5.3. Reset Query . . . . . . . . . . . . . . . . . . . . . . . 7 67 5.4. Cache Response . . . . . . . . . . . . . . . . . . . . . . 7 68 5.5. IPv4 Prefix . . . . . . . . . . . . . . . . . . . . . . . 8 69 5.6. IPv6 Prefix . . . . . . . . . . . . . . . . . . . . . . . 9 70 5.7. End of Data . . . . . . . . . . . . . . . . . . . . . . . 9 71 5.8. Cache Reset . . . . . . . . . . . . . . . . . . . . . . . 10 72 5.9. Error Report . . . . . . . . . . . . . . . . . . . . . . . 10 73 5.10. Fields of a PDU . . . . . . . . . . . . . . . . . . . . . 11 74 6. Protocol Sequences . . . . . . . . . . . . . . . . . . . . . . 13 75 6.1. Start or Restart . . . . . . . . . . . . . . . . . . . . . 13 76 6.2. Typical Exchange . . . . . . . . . . . . . . . . . . . . . 14 77 6.3. No Incremental Update Available . . . . . . . . . . . . . 14 78 6.4. Cache has No Data Available . . . . . . . . . . . . . . . 15 79 7. Transport . . . . . . . . . . . . . . . . . . . . . . . . . . 15 80 7.1. SSH Transport . . . . . . . . . . . . . . . . . . . . . . 16 81 7.2. TLS Transport . . . . . . . . . . . . . . . . . . . . . . 17 82 7.3. TCP MD5 Transport . . . . . . . . . . . . . . . . . . . . 17 83 7.4. TCP-AO Transport . . . . . . . . . . . . . . . . . . . . . 18 84 8. Router-Cache Set-Up . . . . . . . . . . . . . . . . . . . . . 18 85 9. Deployment Scenarios . . . . . . . . . . . . . . . . . . . . . 19 86 10. Error Codes . . . . . . . . . . . . . . . . . . . . . . . . . 20 87 11. Security Considerations . . . . . . . . . . . . . . . . . . . 21 88 12. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 22 89 13. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 23 90 14. References . . . . . . . . . . . . . . . . . . . . . . . . . . 23 91 14.1. Normative References . . . . . . . . . . . . . . . . . . . 23 92 14.2. Informative References . . . . . . . . . . . . . . . . . . 24 93 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 24 95 1. Introduction 97 In order to formally validate the origin ASs of BGP announcements, 98 routers need a simple but reliable mechanism to receive RPKI 99 [I-D.ietf-sidr-arch] formally validated prefix origin data from a 100 trusted cache. This document describes a protocol to deliver 101 validated prefix origin data to routers. 103 Section 3 describes the deployment structure and Section 4 then 104 presents an operational overview. The binary payloads of the 105 protocol are formally described in Section 5, and the expected PDU 106 sequences are described in Section 6. The transport protocol options 107 are described in Section 7. Section 8 details how routers and caches 108 are configured to connect and authenticate. Section 9 describes 109 likely deployment scenarios. The traditional security and IANA 110 considerations end the document. 112 The protocol is extensible to support new PDUs with new semantics 113 when and as needed, as indicated by deployment experience. PDUs are 114 versioned should deployment experience call for change. 116 2. Glossary 118 The following terms are used with special meaning: 120 Global RPKI: The authoritative data of the RPKI are published in a 121 distributed set of servers at the IANA, RIRs, NIRs, and ISPs, see 122 [I-D.ietf-sidr-repos-struct]. 124 Cache: A coalesced copy of the RPKI which is periodically fetched/ 125 refreshed directly or indirectly from the global RPKI using the 126 [RFC5781] protocol/tools. Relying party software is used to 127 gather and validate the distributed data of the RPKI into a cache. 128 Trusting this cache further is a matter between the provider of 129 the cache and a relying party. 131 Serial Number: A 32-bit monotonically increasing ordinal which wraps 132 from 2^32-1 to 0. It denotes the logical version of a cache. A 133 cache increments the value by one when it successfully updates its 134 data from a parent cache or from primary RPKI data. As a cache is 135 receiving, new incoming data and implicit deletes are marked with 136 the new serial but MUST NOT be sent until the fetch is complete. 137 A serial number is not commensurate between caches, nor need it be 138 maintained across resets of the cache server. See [RFC1982] on 139 DNS Serial Number Arithmetic for too much detail on serial number 140 arithmetic. 142 Session ID: When a cache server is started, it generates a session 143 identifier to uniquely identify the instance of the cache and to 144 bind it to the sequence of Serial Numbers that cache instance will 145 generate. This allows the router to restart a failed session 146 knowing that the Serial Number it is using is commensurate with 147 that of the cache. 149 3. Deployment Structure 151 Deployment of the RPKI to reach routers has a three level structure 152 as follows: 154 Global RPKI: The authoritative data of the RPKI are published in a 155 distributed set of servers, RPKI publication repositories, e.g. 156 the IANA, RIRs, NIRs, and ISPs, see [I-D.ietf-sidr-repos-struct]. 158 Local Caches: A local set of one or more collected and verified 159 caches. A relying party, e.g. router or other client, MUST have a 160 trust relationship with, and a trusted transport channel to, any 161 authoritative cache(s) it uses. 163 Routers: A router fetches data from a local cache using the protocol 164 described in this document. It is said to be a client of the 165 cache. There MAY be mechanisms for the router to assure itself of 166 the authenticity of the cache and to authenticate itself to the 167 cache. 169 4. Operational Overview 171 A router establishes and keeps open a connection to one or more 172 caches with which it has client/server relationships. It is 173 configured with a semi-ordered list of caches, and establishes a 174 connection to the most preferred cache, or set of caches, which 175 accept the connections. 177 The router MUST choose the most preferred, by configuration, cache or 178 set of caches so that the operator may control load on their caches 179 and the Global RPKI. 181 Periodically, the router sends to the cache the serial number of the 182 highest numbered data it has received from that cache, i.e. the 183 router's current serial number. When a router establishes a new 184 connection to a cache, or wishes to reset a current relationship, it 185 sends a Reset Query. 187 The Cache responds with all data records which have serial numbers 188 greater than that in the router's query. This may be the null set, 189 in which case the End of Data PDU is still sent. Note that 'greater' 190 must take wrap-around into account, see [RFC1982]. 192 When the router has received all data records from the cache, it sets 193 its current serial number to that of the serial number in the End of 194 Data PDU. 196 When the cache updates its database, it sends a Notify message to 197 every currently connected router. This is a hint that now would be a 198 good time for the router to poll for an update, but is only a hint. 199 The protocol requires the router to poll for updates periodically in 200 any case. 202 Strictly speaking, a router could track a cache simply by asking for 203 a complete data set every time it updates, but this would be very 204 inefficient. The serial number based incremental update mechanism 205 allows an efficient transfer of just the data records which have 206 changed since last update. As with any update protocol based on 207 incremental transfers, the router must be prepared to fall back to a 208 full transfer if for any reason the cache is unable to provide the 209 necessary incremental data. Unlike some incremental transfer 210 protocols, this protocol requires the router to make an explicit 211 request to start the fallback process; this is deliberate, as the 212 cache has no way of knowing whether the router has also established 213 sessions with other caches that may be able to provide better 214 service. 216 As a cache server must evaluate certificates and ROAs which are time 217 dependent, servers' clocks MUST be correct to a tolerance of 218 approximately an hour. 220 5. Protocol Data Units (PDUs) 222 The exchanges between the cache and the router are sequences of 223 exchanges of the following PDUs according to the rules described in 224 Section 6. 226 Fields with unspecified content MUST be zero on transmission and MAY 227 be ignored on receipt. 229 5.1. Serial Notify 231 The cache notifies the router that the cache has new data. 233 The Session ID reassures the router that the serial numbers are 234 commensurate, i.e. the cache session has not been changed. 236 Serial Notify is only message that the cache can send that is not in 237 response to a message from the router. 239 0 8 16 24 31 240 .-------------------------------------------. 241 | Protocol | PDU | | 242 | Version | Type | Session ID | 243 | 0 | 0 | | 244 +-------------------------------------------+ 245 | | 246 | Length=12 | 247 | | 248 +-------------------------------------------+ 249 | | 250 | Serial Number | 251 | | 252 `-------------------------------------------' 254 5.2. Serial Query 256 Serial Query: The router sends Serial Query to ask the cache for all 257 payload PDUs which have serial numbers higher than the serial number 258 in the Serial Query. 260 The cache replies to this query with a Cache Response PDU 261 (Section 5.4) if the cache has a, possibly null, record of the 262 changes since the serial number specified by the router. If there 263 have been no changes since the router last queried, the cache then 264 sends an End Of Data PDU. 266 If the cache does not have the data needed to update the router, 267 perhaps because its records do not go back to the Serial Number in 268 the Serial Query, then it responds with a Cache Reset PDU 269 (Section 5.8). 271 The Session ID tells the cache what instance the router expects to 272 ensure that the serial numbers are commensurate, i.e. the cache 273 session has not been changed. 275 0 8 16 24 31 276 .-------------------------------------------. 277 | Protocol | PDU | | 278 | Version | Type | Session ID | 279 | 0 | 1 | | 280 +-------------------------------------------+ 281 | | 282 | Length=12 | 283 | | 284 +-------------------------------------------+ 285 | | 286 | Serial Number | 287 | | 288 `-------------------------------------------' 290 5.3. Reset Query 292 Reset Query: The router tells the cache that it wants to receive the 293 total active, current, non-withdrawn, database. The cache responds 294 with a Cache Response PDU (Section 5.4). 296 0 8 16 24 31 297 .-------------------------------------------. 298 | Protocol | PDU | | 299 | Version | Type | reserved = zero | 300 | 0 | 2 | | 301 +-------------------------------------------+ 302 | | 303 | Length=8 | 304 | | 305 `-------------------------------------------' 307 5.4. Cache Response 309 Cache Response: The cache responds with zero or more payload PDUs. 310 When replying to a Serial Query request (Section 5.2), the cache 311 sends the set of all data records it has with serial numbers greater 312 than that sent by the client router. When replying to a Reset Query, 313 the cache sends the set of all data records it has; in this case the 314 withdraw/announce field in the payload PDUs MUST have the value 1 315 (announce). 317 In response to a Reset Query, the new value of the Session ID tells 318 the router the instance of the cache session for future confirmation. 319 In response to a Serial Query, the Session ID being the same 320 reassures the router that the serial numbers are commensurate, i.e. 321 the cache session has not changed. 323 0 8 16 24 31 324 .-------------------------------------------. 325 | Protocol | PDU | | 326 | Version | Type | Session ID | 327 | 0 | 3 | | 328 +-------------------------------------------+ 329 | | 330 | Length=8 | 331 | | 332 `-------------------------------------------' 334 5.5. IPv4 Prefix 336 0 8 16 24 31 337 .-------------------------------------------. 338 | Protocol | PDU | | 339 | Version | Type | reserved = zero | 340 | 0 | 4 | | 341 +-------------------------------------------+ 342 | | 343 | Length=20 | 344 | | 345 +-------------------------------------------+ 346 | | Prefix | Max | | 347 | Flags | Length | Length | zero | 348 | | 0..32 | 0..32 | | 349 +-------------------------------------------+ 350 | | 351 | IPv4 Prefix | 352 | | 353 +-------------------------------------------+ 354 | | 355 | Autonomous System Number | 356 | | 357 `-------------------------------------------' 359 The lowest order bit of the Flags field is 1 for an announcement and 360 0 for a withdrawal. 362 In the RPKI, nothing prevents a signing certificate from issuing two 363 identical ROAs, and nothing prohibits the existence of two identical 364 route: or route6: objects in the IRR. In this case there would be no 365 semantic difference between the objects, merely a process redundancy. 367 In the RPKI, there is also an actual need for what might appear to a 368 router as identical IPvX PDUs. This can occur when an upstream 369 certificate is being reissued or there is an address ownership 370 transfer up the validation chain. The ROA would be identical in the 371 router sense, i.e. have the same {prefix, len, max-len, asn}, but a 372 different validation path in the RPKI. This is important to the 373 RPKI, but not to the router. 375 The cache server MUST ensure that it has told the router client to 376 have one and only one IPvX PDU for a unique {prefix, len, max-len, 377 asn} at any one point in time. Should the router client receive an 378 IPvX PDU with a {prefix, len, max-len, asn} identical to one it 379 already has active, it SHOULD raise a Duplicate Announcement Received 380 error. 382 5.6. IPv6 Prefix 384 0 8 16 24 31 385 .-------------------------------------------. 386 | Protocol | PDU | | 387 | Version | Type | reserved = zero | 388 | 0 | 6 | | 389 +-------------------------------------------+ 390 | | 391 | Length=32 | 392 | | 393 +-------------------------------------------+ 394 | | Prefix | Max | | 395 | Flags | Length | Length | zero | 396 | | 0..128 | 0..128 | | 397 +-------------------------------------------+ 398 | | 399 +--- ---+ 400 | | 401 +--- IPv6 Prefix ---+ 402 | | 403 +--- ---+ 404 | | 405 +-------------------------------------------+ 406 | | 407 | Autonomous System Number | 408 | | 409 `-------------------------------------------' 411 5.7. End of Data 413 End of Data: Cache tells router it has no more data for the request. 415 The Session ID MUST be the same as that of the corresponding Cache 416 Response which began the, possibly null, sequence of data PDUs. 418 0 8 16 24 31 419 .-------------------------------------------. 420 | Protocol | PDU | | 421 | Version | Type | Session ID | 422 | 0 | 7 | | 423 +-------------------------------------------+ 424 | | 425 | Length=12 | 426 | | 427 +-------------------------------------------+ 428 | | 429 | Serial Number | 430 | | 431 `-------------------------------------------' 433 5.8. Cache Reset 435 The cache may respond to a Serial Query informing the router that the 436 cache cannot provide an incremental update starting from the serial 437 number specified by the router. The router must decide whether to 438 issue a Reset Query or switch to a different cache. 440 0 8 16 24 31 441 .-------------------------------------------. 442 | Protocol | PDU | | 443 | Version | Type | reserved = zero | 444 | 0 | 8 | | 445 +-------------------------------------------+ 446 | | 447 | Length=8 | 448 | | 449 `-------------------------------------------' 451 5.9. Error Report 453 This PDU is used by either party to report an error to the other. 455 Error reports are only sent as responses to other PDUs. 457 The Error Code is described in Section 10. 459 If the error is not associated with any particular PDU, the Erroneous 460 PDU field MUST be empty and the Length of Encapsulated PDU field MUST 461 be zero. 463 An Error Report PDU MUST NOT be sent for an Error Report PDU. If an 464 erroneous Error Report PDU is received, the session SHOULD be 465 dropped. 467 If the error is associated with a PDU of excessive, or possibly 468 corrupt, length, the Erroneous PDU field MAY be truncated. 470 The diagnostic text is optional, if not present the Length of Error 471 Text field SHOULD be zero. If error text is present, it SHOULD be a 472 string in US-ASCII, for maximum portability; if non-US-ASCII 473 characters are absolutely required, the error text MUST use UTF-8 474 encoding. 476 0 8 16 24 31 477 .-------------------------------------------. 478 | Protocol | PDU | | 479 | Version | Type | Error Code | 480 | 0 | 10 | | 481 +-------------------------------------------+ 482 | | 483 | Length | 484 | | 485 +-------------------------------------------+ 486 | | 487 | Length of Encapsulated PDU | 488 | | 489 +-------------------------------------------+ 490 | | 491 ~ Copy of Erroneous PDU ~ 492 | | 493 +-------------------------------------------+ 494 | | 495 | Length of Error Text | 496 | | 497 +-------------------------------------------+ 498 | | 499 | Arbitrary Text | 500 | of | 501 ~ Error Diagnostic Message ~ 502 | | 503 `-------------------------------------------' 505 5.10. Fields of a PDU 507 PDUs contain the following data elements: 509 Protocol Version: An ordinal, currently 0, denoting the version of 510 this protocol. 512 PDU Type: An ordinal, denoting the type of the PDU, e.g. IPv4 513 Prefix, etc. 515 Serial Number: The serial number of the RPKI Cache when this ROA was 516 received from the cache's up-stream cache server or gathered from 517 the global RPKI. A cache increments its serial number when 518 completing an rigorously validated update from a parent cache, for 519 example via rcynic. See [RFC1982] on DNS Serial Number Arithmetic 520 for too much detail on serial number arithmetic. 522 Session ID: When a cache server is started, it generates a session 523 identifier to identify the instance of the cache and to bind it to 524 the sequence of Serial Numbers that cache instance will generate. 525 This allows the router to restart a failed session knowing that 526 the Serial Number it is using is commensurate with that of the 527 cache. If, at any time, either the router or the cache finds the 528 value of the session identifiers they hold disagree, they MUST 529 completely drop the session and the router MUST flush all data 530 learned from that cache. 532 The Session ID might be a pseudo-random, a monotonically 533 increasing value if the cache has reliable storage, etc. An 534 implementation which uses a fine granularity of time for the 535 Serial Number might never change the Session ID. 537 Length: A 32 bit ordinal which has as its value the count of the 538 bytes in the entire PDU, including the eight bytes of header which 539 end with the length field. 541 Flags: The lowest order bit of the Flags field is 1 for an 542 announcement and 0 for a withdrawal, whether this PDU announces a 543 new right to announce the prefix or withdraws a previously 544 announced right. A withdraw effectively deletes one previously 545 announced IPvX Prefix PDU with the exact same Prefix, Length, Max- 546 Len, and ASN. 548 Prefix Length: An ordinal denoting the shortest prefix allowed for 549 the prefix. 551 Max Length: An ordinal denoting the longest prefix allowed by the 552 prefix. This MUST NOT be less than the Prefix Length element. 554 Prefix: The IPv4 or IPv6 prefix of the ROA. 556 Autonomous System Number: ASN allowed to announce this prefix, a 32 557 bit ordinal. 559 Zero: Fields shown as zero or reserved MUST be zero. The value of 560 such a field MUST be ignored on receipt. 562 6. Protocol Sequences 564 The sequences of PDU transmissions fall into three conversations as 565 follows: 567 6.1. Start or Restart 569 Cache Router 570 ~ ~ 571 | <----- Reset Query -------- | R requests data (or Serial Query) 572 | | 573 | ----- Cache Response -----> | C confirms request 574 | ------- IPvX Prefix ------> | C sends zero or more 575 | ------- IPvX Prefix ------> | IPv4 and IPv6 Prefix 576 | ------- IPvX Prefix ------> | Payload PDUs 577 | ------ End of Data ------> | C sends End of Data 578 | | and sends new serial 579 ~ ~ 581 When a transport session is first established, the router MAY send a 582 Reset Query and the cache responds with a data sequence of all data 583 it contains. 585 Alternatively, if the router has significant unexpired data from a 586 broken session with the same cache, it MAY start with a Serial Query 587 containing the Session ID from the previous session to ensure the 588 serial numbers are commensurate. 590 This Reset Query sequence is also used when the router receives a 591 Cache Reset, chooses a new cache, or fears that it has otherwise lost 592 its way. 594 To limit the length of time a cache must keep the data necessary to 595 generate incremental updates, a router MUST send either a Serial 596 Query or a Reset Query no less frequently than once an hour. This 597 also acts as a keep alive at the application layer. 599 As the cache MAY not keep updates for little more than one hour, the 600 router MUST have a polling interval of no greater than once an hour. 602 6.2. Typical Exchange 604 Cache Router 605 ~ ~ 606 | -------- Notify ----------> | (optional) 607 | | 608 | <----- Serial Query ------- | R requests data 609 | | 610 | ----- Cache Response -----> | C confirms request 611 | ------- IPvX Prefix ------> | C sends zero or more 612 | ------- IPvX Prefix ------> | IPv4 and IPv6 Prefix 613 | ------- IPvX Prefix ------> | Payload PDUs 614 | ------ End of Data ------> | C sends End of Data 615 | | and sends new serial 616 ~ ~ 618 The cache server SHOULD send a notify PDU with its current serial 619 number when the cache's serial changes, with the expectation that the 620 router MAY then issue a serial query earlier than it otherwise might. 621 This is analogous to DNS NOTIFY in [RFC1996]. The cache MUST rate 622 limit Serial Notifies to no more frequently than one per minute. 624 When the transport layer is up and either a timer has gone off in the 625 router, or the cache has sent a Notify, the router queries for new 626 data by sending a Serial Query, and the cache sends all data newer 627 than the serial in the Serial Query. 629 To limit the length of time a cache must keep old withdraws, a router 630 MUST send either a Serial Query or a Reset Query no less frequently 631 than once an hour. 633 6.3. No Incremental Update Available 635 Cache Router 636 ~ ~ 637 | <----- Serial Query ------ | R requests data 638 | ------- Cache Reset ------> | C cannot supply update 639 | | from specified serial 640 | <------ Reset Query ------- | R requests new data 641 | ----- Cache Response -----> | C confirms request 642 | ------- IPvX Prefix ------> | C sends zero or more 643 | ------- IPvX Prefix ------> | IPv4 and IPv6 Prefix 644 | ------- IPvX Prefix ------> | Payload PDUs 645 | ------ End of Data ------> | C sends End of Data 646 | | and sends new serial 647 ~ ~ 649 The cache may respond to a Serial Query with a Cache Reset, informing 650 the router that the cache cannot supply an incremental update from 651 the serial number specified by the router. This might be because the 652 cache has lost state, or because the router has waited too long 653 between polls and the cache has cleaned up old data that it no longer 654 believes it needs, or because the cache has run out of storage space 655 and had to expire some old data early. Regardless of how this state 656 arose, the cache replies with a Cache Reset to tell the router that 657 it cannot honor the request. When a router receives this, the router 658 SHOULD attempt to connect to any more preferred caches in its cache 659 list. If there are no more preferred caches it MUST issue a Reset 660 Query and get an entire new load from the cache. 662 6.4. Cache has No Data Available 664 Cache Router 665 ~ ~ 666 | <----- Serial Query ------ | R requests data 667 | ---- Error Report PDU ----> | C No Data Available 668 ~ ~ 670 Cache Router 671 ~ ~ 672 | <----- Reset Query ------- | R requests data 673 | ---- Error Report PDU ----> | C No Data Available 674 ~ ~ 676 The cache may respond to either a Serial Query or a Reset Query 677 informing the router that the cache cannot supply any update at all. 678 The most likely cause is that the cache has lost state, perhaps due 679 to a restart, and has not yet recovered. While it is possible that a 680 cache might go into such a state without dropping any of its active 681 sessions, a router is more likely to see this behavior when it 682 initially connects and issues a Reset Query while the cache is still 683 rebuilding its database. 685 When a router receives this kind of error, the router SHOULD attempt 686 to connect to any other caches in its cache list, in preference 687 order. If no other caches are available, the router MUST issue 688 periodic Reset Queries until it gets a new usable load from the 689 cache. 691 7. Transport 693 The transport layer session between a router and a cache carries the 694 binary Protocol Data Units (PDUs) in a persistent session. 696 To prevent cache spoofing and DoS attacks by illegitimate routers, it 697 is highly desirable that the router and the cache are authenticated 698 to each other. Integrity protection for payloads is also desirable 699 to protect against monkey in the middle (MITM) attacks. 700 Unfortunately, there is no protocol to do so on all currently used 701 platforms. Therefore, as of this document, there is no mandatory to 702 implement transport which provides authentication and integrity 703 protection. 705 To reduce exposure to dropped but non-terminated sessions, both 706 caches and routers SHOULD enable keep alives when available in the 707 chosen transport protocol. 709 It is expected that, when TCP-AO [RFC5925] is available on all 710 platforms deployed by operators, it will become the mandatory to 711 implement transport. 713 Caches and routers MUST implement unprotected transport over TCP 714 using a port, rpki-rtr, to be assigned, see Section 12. Operators 715 SHOULD use procedural means, ACLs, ... to reduce the exposure to 716 authentication issues. 718 If available to the operator, caches and routers SHOULD use one of 719 the following more protected protocols. 721 Caches and routers SHOULD use TCP AO transport [RFC2385] over the 722 rpki-rtr port. 724 Caches and routers MAY use SSH transport [RFC4252] using using a the 725 normal SSH port. For an example, see Section 7.1. 727 Caches and routers MAY use TCP MD5 transport [RFC2385] using the 728 rpki-rtr port. 730 Caches and routers MAY use IPsec transport [RFC4301] using the rpki- 731 rtr port. 733 Caches and routers MAY use TLS transport [RFC5246] using using a 734 port, rpki-rtr-tls, to be assigned, see Section 12. 736 7.1. SSH Transport 738 To run over SSH, the client router first establishes an SSH transport 739 connection using the SSH transport protocol, and the client and 740 server exchange keys for message integrity and encryption. The 741 client then invokes the "ssh-userauth" service to authenticate the 742 application, as described in the SSH authentication protocol RFC 4252 743 [RFC4252]. Once the application has been successfully authenticated, 744 the client invokes the "ssh-connection" service, also known as the 745 SSH connection protocol. 747 After the ssh-connection service is established, the client opens a 748 channel of type "session", which results in an SSH session. 750 Once the SSH session has been established, the application invokes 751 the application transport as an SSH subsystem called "rpki-rtr". 752 Subsystem support is a feature of SSH version 2 (SSHv2) and is not 753 included in SSHv1. Running this protocol as an SSH subsystem avoids 754 the need for the application to recognize shell prompts or skip over 755 extraneous information, such as a system message that is sent at 756 shell start-up. 758 It is assumed that the router and cache have exchanged keys out of 759 band by some reasonably secured means. 761 Cache servers supporting SSH transport MUST accept RSA and DSA 762 authentication, and SHOULD accept ECDSA authentication. User 763 authentication MUST be supported; host authentication MAY be 764 supported. Implementations MAY support password authentication. 765 Client routers SHOULD verify the public key of the cache, to avoid 766 monkey in the middle attacks. 768 7.2. TLS Transport 770 Client routers using TLS transport MUST use client-side certificates 771 for authentication. While in principle any type of certificate and 772 certificate authority may be used, in general cache operators will 773 generally wish to create their own small-scale CA and issue 774 certificates to each authorized router. This simplifies credential 775 roll-over; any unrevoked, unexpired certificate from the proper CA 776 may be used. If such certificates are used, the CN field [RFC5280] 777 MUST be used to denote the router's identity. 779 Clients SHOULD verify the cache's certificate as well, to avoid 780 monkey in the middle attacks. 782 7.3. TCP MD5 Transport 784 If TCP-MD5 is used, implementations MUST support key lengths of at 785 least 80 printable ASCII bytes, per section 4.5 of [RFC2385]. 786 Implementations MUST also support hexadecimal sequences of at least 787 32 characters, i.e., 128 bits. 789 Key rollover with TCP-MD5 is problematic. Cache servers SHOULD 790 support [RFC4808]. 792 7.4. TCP-AO Transport 794 Implementations MUST support key lengths of at least 80 printable 795 ASCII bytes. Implementations MUST also support hexadecimal sequences 796 of at least 32 characters, i.e., 128 bits. MAC lengths of at least 797 96 bits MUST be supported, per section 5.3 of [RFC2385]. 799 The cryptographic algorithms and associcated parameters described in 800 [RFC5926] MUST be supported. 802 8. Router-Cache Set-Up 804 A cache has the public authentication data for each router it is 805 configured to support. 807 A router may be configured to peer with a selection of caches, and a 808 cache may be configured to support a selection of routers. Each must 809 have the name of, and authentication data for, each peer. In 810 addition, in a router, this list has a non-unique preference value 811 for each server in order of preference. This preference merely 812 denotes proximity, not trust, preferred belief, etc. The client 813 router attempts to establish a session with each potential serving 814 cache in preference order, and then starts to load data from the most 815 preferred cache to which it can connect and authenticate. The 816 router's list of caches has the following elements: 818 Preference: An ordinal denoting the router's preference to connect 819 to that cache, the lower the value the more preferred. 821 Name: The IP Address or fully qualified domain name of the cache. 823 Key: Any needed public key of the cache. 825 MyKey: Any needed private key or certificate of this client. 827 Due to the distributed nature of the RPKI, caches simply can not be 828 rigorously synchronous. A client may hold data from multiple caches, 829 but MUST keep the data marked as to source, as later updates MUST 830 affect the correct data. 832 Just as there may be more than one covering ROA from a single cache, 833 there may be multiple covering ROAs from multiple caches. The 834 results are as described in [I-D.ietf-sidr-pfx-validate]. 836 If data from multiple caches are held, implementations MUST NOT 837 distinguish between data sources when performing validation. 839 When a more preferred cache becomes available, if resources allow, it 840 would be prudent for the client to start fetching from that cache. 842 The client SHOULD attempt to maintain at least one set of data, 843 regardless of whether it has chosen a different cache or established 844 a new connection to the previous cache. 846 A client MAY drop the data from a particular cache when it is fully 847 in synch with one or more other caches. 849 A client SHOULD delete the data from a cache when it has been unable 850 to refresh from that cache for a configurable timer value. The 851 default for that value is twice the polling period for that cache. 853 If a client loses connectivity to a cache it is using, or otherwise 854 decides to switch to a new cache, it SHOULD retain the data from the 855 previous cache until it has a full set of data from one or more other 856 caches. Note that this may already be true at the point of 857 connection loss if the client has connections to more than one cache. 859 9. Deployment Scenarios 861 For illustration, we present three likely deployment scenarios. 863 Small End Site: The small multi-homed end site may wish to outsource 864 the RPKI cache to one or more of their upstream ISPs. They would 865 exchange authentication material with the ISP using some out of 866 band mechanism, and their router(s) would connect to one or more 867 up-streams' caches. The ISPs would likely deploy caches intended 868 for customer use separately from the caches with which their own 869 BGP speakers peer. 871 Large End Site: A larger multi-homed end site might run one or more 872 caches, arranging them in a hierarchy of client caches, each 873 fetching from a serving cache which is closer to the global RPKI. 874 They might configure fall-back peerings to up-stream ISP caches. 876 ISP Backbone: A large ISP would likely have one or more redundant 877 caches in each major PoP, and these caches would fetch from each 878 other in an ISP-dependent topology so as not to place undue load 879 on the global RPKI publication infrastructure. 881 Experience with large DNS cache deployments has shown that complex 882 topologies are ill-advised as it is easy to make errors in the graph, 883 e.g. not maintaining a loop-free condition. 885 Of course, these are illustrations and there are other possible 886 deployment strategies. It is expected that minimizing load on the 887 global RPKI servers will be a major consideration. 889 To keep load on global RPKI services from unnecessary peaks, it is 890 recommended that primary caches which load from the distributed 891 global RPKI not do so all at the same times, e.g. on the hour. 892 Choose a random time, perhaps the ISP's AS number modulo 60 and 893 jitter the inter-fetch timing. 895 10. Error Codes 897 This section contains a preliminary list of error codes. The authors 898 expect additions to this section during development of the initial 899 implementations. Errors which are considered fatal SHOULD cause the 900 session to be dropped. 902 0: Corrupt Data (fatal): The receiver believes the received PDU to 903 be corrupt in a manner not specified by other error codes. 905 1: Internal Error (fatal): The party reporting the error experienced 906 some kind of internal error unrelated to protocol operation (ran 907 out of memory, a coding assertion failed, et cetera). 909 2: No Data Available: The cache believes itself to be in good 910 working order, but is unable to answer either a Serial Query or a 911 Reset Query because it has no useful data available at this time. 912 This is likely to be a temporary error, and most likely indicates 913 that the cache has not yet completed pulling down an initial 914 current data set from the global RPKI system after some kind of 915 event that invalidated whatever data it might have previously held 916 (reboot, network partition, et cetera). 918 3: Invalid Request (fatal): The cache server believes the client's 919 request to be invalid. 921 4: Unsupported Protocol Version (fatal): The Protocol Version is not 922 known by the receiver of the PDU. 924 5: Unsupported PDU Type (fatal): The PDU Type is not known by the 925 receiver of the PDU. 927 6: Withdrawal of Unknown Record (fatal): The received PDU has Flag=0 928 but a record for the Prefix/PrefixLength/MaxLength triple does not 929 exist in the receiver's database. 931 7: Duplicate Announcement Received (fatal): The received PDU has an 932 identical {prefix, len, max-len, asn} tuple as a PDU which is 933 still active in the router. 935 11. Security Considerations 937 As this document describes a security protocol, many aspects of 938 security interest are described in the relevant sections. This 939 section points out issues which may not be obvious in other sections. 941 Cache Validation: In order for a collection of caches as described 942 in Section 9 to guarantee a consistent view, they need to be given 943 consistent trust anchors to use in their internal validation 944 process. Distribution of a consistent trust anchor is assumed to 945 be out of band. 947 Cache Peer Identification: The router initiates a transport session 948 to a cache, which it identifies by either IP address or fully 949 qualified domain name. Be aware that a DNS or address spoofing 950 attack could make the correct cache unreachable. No session would 951 be established, as the authorization keys would not match. 953 Transport Security: The RPKI relies on object, not server or 954 transport, trust. I.e. the IANA root trust anchor is distributed 955 to all caches through some out of band means, and can then be used 956 by each cache to validate certificates and ROAs all the way down 957 the tree. The inter-cache relationships are based on this object 958 security model, hence the inter-cache transport can be lightly 959 protected. 961 But this protocol document assumes that the routers can not do the 962 validation cryptography. Hence the last link, from cache to 963 router, is secured by server authentication and transport level 964 security. This is dangerous, as server authentication and 965 transport have very different threat models than object security. 967 So the strength of the trust relationship and the transport 968 between the router(s) and the cache(s) are critical. You're 969 betting your routing on this. 971 While we can not say the cache must be on the same LAN, if only 972 due to the issue of an enterprise wanting to off-load the cache 973 task to their upstream ISP(s), locality, trust, and control are 974 very critical issues here. The cache(s) really SHOULD be as 975 close, in the sense of controlled and protected (against DDoS, 976 MITM) transport, to the router(s) as possible. It also SHOULD be 977 topologically close so that a minimum of validated routing data 978 are needed to bootstrap a router's access to a cache. 980 The identity of the cache server SHOULD be verified and 981 authenticated by the router client, and vice versa, before any 982 data are exchanged. 984 Transports which can not provide the necessary authentication and 985 integrity (see Section 7) must rely on network design and 986 operational controls to provide protection against spoofing/ 987 corruption attacks. 989 12. IANA Considerations 991 This document requests the IANA to assign 'well known' TCP Port 992 Numbers to the RPKI-Router Protocol for the following, see Section 7: 994 rpki-rtr 995 rpki-rtr-tls 997 This document requests the IANA to create a registry for tuples of 998 Protocol Version / PDU Type, each of which may range from 0 to 255. 999 The name of the registry should be rpki-rtr-pdu. The policy for 1000 adding to the registry is RFC Required per [RFC5226], either 1001 standards track or experimental. The initial entries should be as 1002 follows: 1004 Protocol 1005 Version PDU Type 1006 -------- ------------------- 1007 0 0 - Serial Notify 1008 0 1 - Serial Query 1009 0 2 - Reset Query 1010 0 3 - Cache Response 1011 0 4 - IPv4 Prefix 1012 0 6 - IPv6 Prefix 1013 0 7 - End of Data 1014 0 8 - Cache Reset 1015 0 10 - Error Report 1016 0 255 - Reserved 1018 This document requests the IANA to create a registry for Error Codes 1019 0 to 255. The name of the registry should be rpki-rtr-error. The 1020 policy for adding to the registry is Expert Review per [RFC5226], 1021 where the responsible IESG area director should appoint the Expert 1022 Reviewer. The initial entries should be as follows: 1024 0 - Corrupt Data 1025 1 - Internal Error 1026 2 - No Data Available 1027 3 - Invalid Request 1028 4 - Unsupported Protocol Version 1029 5 - Unsupported PDU Type 1030 6 - Withdrawal of Unknown Record 1031 7 - Duplicate Announcement Received 1032 255 - Reserved 1034 This document requests the IANA to add an SSH Connection Protocol 1035 Subsystem Name, as defined in [RFC4250], of 'rpki-rtr'. 1037 13. Acknowledgments 1039 The authors wish to thank Steve Bellovin, Rex Fernando, Paul Hoffman, 1040 Russ Housley, Pradosh Mohapatra, Keyur Patel, Sandy Murphy, Robert 1041 Raszuk, John Scudder, Ruediger Volk, and David Ward. Particular 1042 thanks go to Hannes Gredler for showing us the dangers of unnecessary 1043 fields. 1045 14. References 1047 14.1. Normative References 1049 [I-D.ietf-sidr-pfx-validate] 1050 Mohapatra, P., Scudder, J., Ward, D., Bush, R., and R. 1051 Austein, "BGP Prefix Origin Validation", 1052 draft-ietf-sidr-pfx-validate-03 (work in progress), 1053 October 2011. 1055 [RFC1982] Elz, R. and R. Bush, "Serial Number Arithmetic", RFC 1982, 1056 August 1996. 1058 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1059 Requirement Levels", BCP 14, RFC 2119, March 1997. 1061 [RFC2385] Heffernan, A., "Protection of BGP Sessions via the TCP MD5 1062 Signature Option", RFC 2385, August 1998. 1064 [RFC4250] Lehtinen, S. and C. Lonvick, "The Secure Shell (SSH) 1065 Protocol Assigned Numbers", RFC 4250, January 2006. 1067 [RFC4252] Ylonen, T. and C. Lonvick, "The Secure Shell (SSH) 1068 Authentication Protocol", RFC 4252, January 2006. 1070 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 1071 Internet Protocol", RFC 4301, December 2005. 1073 [RFC5226] Narten, T. and H. Alvestrand, "Guidelines for Writing an 1074 IANA Considerations Section in RFCs", BCP 26, RFC 5226, 1075 May 2008. 1077 [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security 1078 (TLS) Protocol Version 1.2", RFC 5246, August 2008. 1080 [RFC5280] Cooper, D., Santesson, S., Farrell, S., Boeyen, S., 1081 Housley, R., and W. Polk, "Internet X.509 Public Key 1082 Infrastructure Certificate and Certificate Revocation List 1083 (CRL) Profile", RFC 5280, May 2008. 1085 [RFC5925] Touch, J., Mankin, A., and R. Bonica, "The TCP 1086 Authentication Option", RFC 5925, June 2010. 1088 [RFC5926] Lebovitz, G. and E. Rescorla, "Cryptographic Algorithms 1089 for the TCP Authentication Option (TCP-AO)", RFC 5926, 1090 June 2010. 1092 14.2. Informative References 1094 [I-D.ietf-sidr-arch] 1095 Lepinski, M. and S. Kent, "An Infrastructure to Support 1096 Secure Internet Routing", draft-ietf-sidr-arch-13 (work in 1097 progress), May 2011. 1099 [I-D.ietf-sidr-repos-struct] 1100 Huston, G., Loomans, R., and G. Michaelson, "A Profile for 1101 Resource Certificate Repository Structure", 1102 draft-ietf-sidr-repos-struct-09 (work in progress), 1103 July 2011. 1105 [RFC1996] Vixie, P., "A Mechanism for Prompt Notification of Zone 1106 Changes (DNS NOTIFY)", RFC 1996, August 1996. 1108 [RFC4808] Bellovin, S., "Key Change Strategies for TCP-MD5", 1109 RFC 4808, March 2007. 1111 [RFC5781] Weiler, S., Ward, D., and R. Housley, "The rsync URI 1112 Scheme", RFC 5781, February 2010. 1114 Authors' Addresses 1116 Randy Bush 1117 Internet Initiative Japan 1118 5147 Crystal Springs 1119 Bainbridge Island, Washington 98110 1120 US 1122 Phone: +1 206 780 0431 x1 1123 Email: randy@psg.com 1125 Rob Austein 1126 Dragon Research Labs 1128 Email: sra@hactrn.net