idnits 2.17.1 draft-ymbk-rpki-rtr-protocol-05.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** You're using the IETF Trust Provisions' Section 6.b License Notice from 12 Sep 2009 rather than the newer Notice from 28 Dec 2009. (See https://trustee.ietf.org/license-info/) -- The document has an IETF Trust Provisions (28 Dec 2009) Section 6.c(ii) Publication Limitation clause. If this document is intended for submission to the IESG for publication, this constitutes an error. 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 == Using lowercase 'not' together with uppercase 'MUST', 'SHALL', 'SHOULD', or 'RECOMMENDED' is not an accepted usage according to RFC 2119. Please use uppercase 'NOT' together with RFC 2119 keywords (if that is what you mean). Found 'MUST not' in this paragraph: Serial Number: A 32-bit monotonically increasing, cardinal which wraps from 2^32-1 to 0. It denotes the logical version of a cache. A cache increments the value by one when it successfully updates its data from a parent cache or from primary RPKI data. As a cache is rcynicing, new incoming data, and implicit deletes, are marked with the new serial but MUST not be sent until the fetch is complete. A serial number is not commensurate between caches, nor need it be maintained across resets of the cache server. See [RFC1982] on DNS Serial Number Arithmetic for too much detail on serial number arithmetic. -- The document date (February 20, 2010) is 5178 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 (-13) exists of draft-ietf-sidr-arch-06 == Outdated reference: A later version (-09) exists of draft-ietf-sidr-repos-struct-01 Summary: 2 errors (**), 0 flaws (~~), 4 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group R. Bush 3 Internet-Draft IIJ 4 Intended status: Standards Track R. Austein 5 Expires: August 24, 2010 ISC 6 February 20, 2010 8 The RPKI/Router Protocol 9 draft-ymbk-rpki-rtr-protocol-05 11 Abstract 13 In order to formally validate the origin ASes of BGP announcements, 14 routers need a simple but reliable mechanism to receive RPKI 15 [I-D.ietf-sidr-arch] or analogous prefix origin data from a trusted 16 cache. This document describes a protocol to deliver validated 17 prefix origin data to routers over ssh. 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 RFC 2119 [RFC2119]. 25 Status of this Memo 27 This Internet-Draft is submitted to IETF in full conformance with the 28 provisions of BCP 78 and BCP 79. This document may not be modified, 29 and derivative works of it may not be created, and it may not be 30 published except as an Internet-Draft. 32 Internet-Drafts are working documents of the Internet Engineering 33 Task Force (IETF), its areas, and its working groups. Note that 34 other groups may also distribute working documents as Internet- 35 Drafts. 37 Internet-Drafts are draft documents valid for a maximum of six months 38 and may be updated, replaced, or obsoleted by other documents at any 39 time. It is inappropriate to use Internet-Drafts as reference 40 material or to cite them other than as "work in progress." 42 The list of current Internet-Drafts can be accessed at 43 http://www.ietf.org/ietf/1id-abstracts.txt. 45 The list of Internet-Draft Shadow Directories can be accessed at 46 http://www.ietf.org/shadow.html. 48 This Internet-Draft will expire on August 24, 2010. 50 Copyright Notice 52 Copyright (c) 2010 IETF Trust and the persons identified as the 53 document authors. All rights reserved. 55 This document is subject to BCP 78 and the IETF Trust's Legal 56 Provisions Relating to IETF Documents 57 (http://trustee.ietf.org/license-info) in effect on the date of 58 publication of this document. Please review these documents 59 carefully, as they describe your rights and restrictions with respect 60 to this document. Code Components extracted from this document must 61 include Simplified BSD License text as described in Section 4.e of 62 the Trust Legal Provisions and are provided without warranty as 63 described in the BSD License. 65 Table of Contents 67 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 68 2. Deployment Structure . . . . . . . . . . . . . . . . . . . . . 4 69 3. Operational Overview . . . . . . . . . . . . . . . . . . . . . 4 70 4. Protocol Data Units (PDUs) . . . . . . . . . . . . . . . . . . 5 71 4.1. Serial Notify . . . . . . . . . . . . . . . . . . . . . . 6 72 4.2. Serial Query . . . . . . . . . . . . . . . . . . . . . . . 6 73 4.3. Reset Query . . . . . . . . . . . . . . . . . . . . . . . 7 74 4.4. Cache Response . . . . . . . . . . . . . . . . . . . . . . 7 75 4.5. IPv4 Prefix . . . . . . . . . . . . . . . . . . . . . . . 8 76 4.6. IPv6 Prefix . . . . . . . . . . . . . . . . . . . . . . . 9 77 4.7. End of Data . . . . . . . . . . . . . . . . . . . . . . . 9 78 4.8. Cache Reset . . . . . . . . . . . . . . . . . . . . . . . 10 79 4.9. Error Report . . . . . . . . . . . . . . . . . . . . . . . 10 80 4.10. Fields of a PDU . . . . . . . . . . . . . . . . . . . . . 11 81 5. Protocol Sequences . . . . . . . . . . . . . . . . . . . . . . 12 82 5.1. Start or Restart . . . . . . . . . . . . . . . . . . . . . 12 83 5.2. Typical Exchange . . . . . . . . . . . . . . . . . . . . . 13 84 5.3. No Incremental Update Available . . . . . . . . . . . . . 14 85 5.4. Cache has No Data Available . . . . . . . . . . . . . . . 14 86 6. SSH Transport . . . . . . . . . . . . . . . . . . . . . . . . 15 87 7. Router-Cache Set-Up . . . . . . . . . . . . . . . . . . . . . 15 88 8. Deployment Scenarios . . . . . . . . . . . . . . . . . . . . . 16 89 9. Error Codes . . . . . . . . . . . . . . . . . . . . . . . . . 17 90 10. Security Considerations . . . . . . . . . . . . . . . . . . . 17 91 11. Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 92 12. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 19 93 13. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 19 94 14. References . . . . . . . . . . . . . . . . . . . . . . . . . . 19 95 14.1. Normative References . . . . . . . . . . . . . . . . . . . 19 96 14.2. Informative References . . . . . . . . . . . . . . . . . . 20 98 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 20 100 1. Introduction 102 In order to formally validate the origin ASes of BGP announcements, 103 routers need a simple but reliable mechanism to receive RPKI 104 [I-D.ietf-sidr-arch] or analogous formally validated prefix origin 105 data from a trusted cache. This document describes a protocol to 106 deliver validated prefix origin data to routers over ssh. 108 Section 2 describes the deployment structure and Section 3 then 109 presents an operational overview. The binary payloads of the 110 protocol are formally described in Section 4, and the expected PDU 111 sequences are described in Section 5. And the transport protocol is 112 described in Section 6. Section 7 details how routers and caches are 113 configured to connect and authenticate. Section 8 describes likely 114 deployment scenarios. The traditional security and IANA 115 considerations end the document. 117 2. Deployment Structure 119 Deployment of the RPKI to reach routers has a three level structure 120 as follows: 122 Global RPKI: The authoritative data of the RPKI are published in a 123 distributed set of servers, RPKI publication repositories, e.g. 124 the IANA, RIRs, NIRs, and ISPs, see [I-D.ietf-sidr-repos-struct]. 126 Local Caches: A local set of one or more collected and verified non- 127 authoritative caches. A relying party, e.g. router or other 128 client, MUST have a formally authenticated trust relationship 129 with, and a secure transport channel to, any non-authoritative 130 cache(s) it uses. 132 Routers: A router fetches data from a local cache using the protocol 133 described in this document. It is said to be a client of the 134 cache. There are mechanisms for the router to assure itself of 135 the authenticity of the cache and to authenticate itself to the 136 cache. 138 3. Operational Overview 140 A router establishes and keeps open an authenticated connection to a 141 cache with which it has an client/server relationship. It is 142 configured with a semi-ordered list of caches, and establishes a 143 connection to the highest preference cache that accepts one. 145 Periodically, the router sends to the cache the serial number of the 146 highest numbered data record it has received from that cache, i.e. 147 the router's current serial number. When a router establishes a new 148 connection to a cache, or wishes to reset a current relationship, it 149 sends a Reset Query. 151 The Cache responds with all data records which have serial numbers 152 greater than that in the router's query. This may be the null set, 153 in which case the End of Data PDU is still sent. Note that 'greater' 154 must take wrap-around into account, see [RFC1982]. 156 When the router has received all data records from the cache, it sets 157 its current serial number to that of the serial number in the End of 158 Data PDU. 160 When the cache updates its database, it sends a Notify message to 161 every currently connected router. This is a hint that now would be a 162 good time for the router to poll for an update, but is only a hint. 163 The protocol requires the router to poll for updates periodically in 164 any case. 166 Strictly speaking, a router could track a cache simply by asking for 167 a complete data set every time it updates, but this would be very 168 inefficient. The serial number based incremental update mechanism 169 allows an efficient transfer of just the data records which have 170 changed since last update. As with any update protocol based on 171 incremental transfers, the router must be prepared to fall back to a 172 full transfer if for any reason the cache is unable to provide the 173 necessary incremental data. Unlike some incremental transfer 174 protocols, this protocol requires the router to make an explicit 175 request to start the fallback process; this is deliberate, as the 176 cache has no way of knowing whether the router has also established 177 sessions with other caches that may be able to provide better 178 service. 180 4. Protocol Data Units (PDUs) 182 The exchanges between the cache and the router are sequences of 183 exchanges of the following PDUs according to the rules described in 184 Section 5. 186 4.1. Serial Notify 188 The cache notifies the router that the cache has new data. 190 0 8 16 24 31 191 .-------------------------------------------. 192 | Protocol | PDU | | 193 | Version | Type | reserved = zero | 194 | 0 | 0 | | 195 +-------------------------------------------+ 196 | | 197 | Length | 198 | | 199 +-------------------------------------------+ 200 | | 201 | Serial Number | 202 | | 203 `-------------------------------------------' 205 4.2. Serial Query 207 Serial Query: The router sends Serial Query to ask the cache for all 208 payload PDUs which have serial numbers higher than the serial number 209 in the Serial Query. 211 The cache replys to this query with a Cache Response PDU 212 (Section 4.4) if the cache has a record of the changes since the 213 serial number specified by the router. If there have been no changes 214 since the router last queried, the cache responds with an End Of Data 215 PDU. If the cache does not have the data needed to update the 216 router, perhaps because its records do not go back to the Serial 217 Number in the Serial Query, then cache responds with a Cache Reset 218 PDU (Section 4.8). 220 0 8 16 24 31 221 .-------------------------------------------. 222 | Protocol | PDU | | 223 | Version | Type | reserved = zero | 224 | 0 | 1 | | 225 +-------------------------------------------+ 226 | | 227 | Length=12 | 228 | | 229 +-------------------------------------------+ 230 | | 231 | Serial Number | 232 | | 233 `-------------------------------------------' 235 4.3. Reset Query 237 Reset Query: The router tells the cache that it wants to receive the 238 total active, current, non-withdrawn, database. The cache responds 239 with a Cache Response PDU (Section 4.4). 241 0 8 16 24 31 242 .-------------------------------------------. 243 | Protocol | PDU | | 244 | Version | Type | reserved = zero | 245 | 0 | 2 | | 246 +-------------------------------------------+ 247 | | 248 | Length=8 | 249 | | 250 `-------------------------------------------' 252 4.4. Cache Response 254 Cache Response: The cache responds with zero or more payload PDUs. 255 When replying to a Serial Query request (Section 4.2), the cache 256 sends the set of all data records it has with serial numbers greater 257 than that sent by the client router. When replying to a Reset Query, 258 the cache sends the set of all data records it has; in this case the 259 withdraw/announce field in the payload PDUs MUST have the value 1 260 (announce). 262 0 8 16 24 31 263 .-------------------------------------------. 264 | Protocol | PDU | | 265 | Version | Type | reserved = zero | 266 | 0 | 3 | | 267 +-------------------------------------------+ 268 | | 269 | Length=8 | 270 | | 271 `-------------------------------------------' 273 4.5. IPv4 Prefix 275 0 8 16 24 31 276 .-------------------------------------------. 277 | Protocol | PDU | | 278 | Version | Type | Color | 279 | 0 | 4 | | 280 +-------------------------------------------+ 281 | | 282 | Length=20 | 283 | | 284 +-------------------------------------------+ 285 | | Prefix | Max | Data | 286 | Flags | Length | Length | Source | 287 | | 0..32 | 0..32 | RPKI/IRR | 288 +-------------------------------------------+ 289 | | 290 | IPv4 prefix | 291 | | 292 +-------------------------------------------+ 293 | | 294 | Autonomous System Number | 295 | | 296 `-------------------------------------------' 298 Due to the nature of the RPKI and the IRR, there can be multiple 299 identical IPvX PDUs. Hence the router will likely keep an internal 300 ref-count on each IPvX PDU. 302 In the RPKI, nothing prevents a signing certificate from issuing two 303 identical ROAs, and nothing prohibits the existence of two identical 304 route: or route6: objects in the IRR. In this case there would be no 305 semantic difference between the objects, merely a process redundancy. 307 In the RPKI, there is also an actual need for what will appear to the 308 router as identical IPvX PDUs. This occurs when an upstream 309 certificate is being reissued or a site is changing providers, often 310 a 'make and break' situation. The ROA is identical in the router 311 sense, i.e. has the same {prefix, len, max-len, asn}, but has a 312 different validation path in the RPKI. This is important to the 313 RPKI, but not to the router. 315 The lowest order bit of the Flags field is 1 for an announcement and 316 0 for a withdrawal. 318 4.6. IPv6 Prefix 320 0 8 16 24 31 321 .-------------------------------------------. 322 | Protocol | PDU | | 323 | Version | Type | Color | 324 | 0 | 6 | | 325 +-------------------------------------------+ 326 | | 327 | Length=40 | 328 | | 329 +-------------------------------------------+ 330 | | Prefix | Max | Data | 331 | Flags | Length | Length | Source | 332 | | 0..128 | 0..128 | RPKI/IRR | 333 +-------------------------------------------+ 334 | | 335 +--- ---+ 336 | | 337 +--- IPv6 prefix ---+ 338 | | 339 +--- ---+ 340 | | 341 +-------------------------------------------+ 342 | | 343 | Autonomous System Number | 344 | | 345 `-------------------------------------------' 347 4.7. End of Data 349 End of Data: Cache tells router it has no more data for the request. 351 0 8 16 24 31 352 .-------------------------------------------. 353 | Protocol | PDU | | 354 | Version | Type | reserved = zero | 355 | 0 | 7 | | 356 +-------------------------------------------+ 357 | | 358 | Length=12 | 359 | | 360 +-------------------------------------------+ 361 | | 362 | Serial Number | 363 | | 364 `-------------------------------------------' 366 4.8. Cache Reset 368 The cache may respond to a Serial Query informing the router that the 369 cache cannot provide an incremental update starting from the serial 370 number specified by the router. The router must decide whether to 371 issue a Reset Query or switch to a different cache. 373 0 8 16 24 31 374 .-------------------------------------------. 375 | Protocol | PDU | | 376 | Version | Type | reserved = zero | 377 | 0 | 8 | | 378 +-------------------------------------------+ 379 | | 380 | Length=8 | 381 | | 382 `-------------------------------------------' 384 4.9. Error Report 386 This PDU is used by either party to report an error to the other. 388 If the error is not associated with any particular PDU, the Erroneous 389 PDU field should be empty and the Length of Encapsulated PDU field 390 should be zero. 392 The diagnostic text is optional, if not present the Length of Error 393 Text field should be zero. If error text is present, it SHOULD be a 394 string in US-ASCII, for maximum portability; if non-US-ASCII 395 characters are absolutely required, the error text MUST use UTF-8 396 encoding. 398 0 8 16 24 31 399 .-------------------------------------------. 400 | Protocol | PDU | | 401 | Version | Type | Error Number | 402 | 0 | 10 | | 403 +-------------------------------------------+ 404 | | 405 | Length | 406 | | 407 +-------------------------------------------+ 408 | | 409 | Length of Encapsulated PDU | 410 | | 411 +-------------------------------------------+ 412 | | 413 ~ Copy of Erroneous PDU ~ 414 | | 415 +-------------------------------------------+ 416 | | 417 | Length of Error Text | 418 | | 419 +-------------------------------------------+ 420 | | 421 | Arbitrary Text | 422 | of | 423 ~ Error Diagnostic Message ~ 424 | | 425 `-------------------------------------------' 427 4.10. Fields of a PDU 429 PDUs contain the following data elements: 431 Protocol Version: A cardinal, currently 0, denoting the version of 432 this protocol. 434 Serial Number: The serial number of the RPKI Cache when this ROA was 435 received from the cache's up-stream cache server or gathered from 436 the global RPKI. A cache increments its serial number when 437 completing an rcynic from a parent cache. See [RFC1982] on DNS 438 Serial Number Arithmetic for too much detail on serial number 439 arithmetic. 441 Length: A 32 bit cardinal which has as its value the count of the 442 bytes in the entire PDU, including the eight bytes of header which 443 end with the length field. 445 Color: An arbitrary 16 bit field that might be used in some way. 447 Flags: The lowest order bit of the Flags field is 1 for an 448 announcement and 0 for a withdrawal, whether this PDU announces a 449 new right to announce the prefix or withdraws a previously 450 announced right. A withdraw effectively deletes one previously 451 announced IPvX Prefix PDU with the exact same Prefix, Length, Max- 452 Len, ASN, Data Source, and Color. 454 Prefix Length: A cardinal denoting the shortest prefix allowed for 455 the prefix. 457 Max Length: A cardinal denoting the longest prefix allowed by the 458 prefix. This MUST NOT be less than the Prefix Length element. 460 Data Source: A cardinal denoting the source of the data, e.g. for 461 RPKI data, it is 0, for IRR data it is 1. 463 Prefix: The IPv4 or IPv6 prefix of the ROA. 465 Autonomous System Number: ASN allowed to announce this prefix, a 32 466 bit cardinal. 468 5. Protocol Sequences 470 The sequences of PDU transmissions fall into three conversations as 471 follows: 473 5.1. Start or Restart 475 Cache Router 476 ~ ~ 477 | <----- Reset Query -------- | R requests data 478 | | 479 | ----- Cache Response -----> | C confirms request 480 | ------- IPvX Prefix ------> | C sends zero or more 481 | ------- IPvX Prefix ------> | IPv4 and IPv6 Prefix 482 | ------- IPvX Prefix ------> | Payload PDUs 483 | ------ End of Data ------> | C sends End of Data 484 | | and sends new serial 485 ~ ~ 487 When a transport session is first established, the router sends a 488 Reset Query and the cache responds with a data sequence of all data 489 it contains. 491 This Reset Query sequence is also used when the router receives a 492 Cache Reset, chooses a new cache, or fears that it has otherwise lost 493 its way. 495 To limit the length of time a cache must keep the data necessary to 496 generate incremental updates, a router MUST send either a Serial 497 Query or a Reset Query no less frequently than once an hour. This 498 also acts as a keep alive at the application layer. 500 5.2. Typical Exchange 502 Cache Router 503 ~ ~ 504 | -------- Notify ----------> | (optional) 505 | | 506 | <----- Serial Query ------- | R requests data 507 | | 508 | ----- Cache Response -----> | C confirms request 509 | ------- IPvX Prefix ------> | C sends zero or more 510 | ------- IPvX Prefix ------> | IPv4 and IPv6 Prefix 511 | ------- IPvX Prefix ------> | Payload PDUs 512 | ------ End of Data ------> | C sends End of Data 513 | | and sends new serial 514 ~ ~ 516 The cache server SHOULD send a notify PDU with its current serial 517 number when the cache's serial changes, with the expectation that the 518 router MAY then issue a serial query earlier than it otherwise might. 519 This is analogous to DNS NOTIFY in [RFC1996]. The cache SHOULD rate 520 limit Serial Notifies to no more frequently than one per minute. 522 When the transport layer is up and either a timer has gone off in the 523 router, or the cache has sent a Notify, the router queries for new 524 data by sending a Serial Query, and the cache sends all data newer 525 than the serial in the Serial Query. 527 To limit the length of time a cache must keep old withdraws, a router 528 MUST send either a Serial Query or a Reset Query no less frequently 529 than once an hour. 531 5.3. No Incremental Update Available 533 Cache Router 534 ~ ~ 535 | <----- Serial Query ------ | R requests data 536 | ------- Cache Reset ------> | C cannot supply update 537 | | from specified serial 538 | <------ Reset Query ------- | R requests new data 539 | ----- Cache Response -----> | C confirms request 540 | ------- IPvX Prefix ------> | C sends zero or more 541 | ------- IPvX Prefix ------> | IPv4 and IPv6 Prefix 542 | ------- IPvX Prefix ------> | Payload PDUs 543 | ------ End of Data ------> | C sends End of Data 544 | | and sends new serial 545 ~ ~ 547 The cache may respond to a Serial Query with a Cache Reset, informing 548 the router that the cache cannot supply an incremental update from 549 the serial number specified by the router. This might be because the 550 cache has lost state, or because the router has waited too long 551 between polls and the cache has cleaned up old data that it no longer 552 believes it needs, or because the cache has run out of storage space 553 and had to expire some old data early. Regardless of how this state 554 arose, the cache replies with a Cache Reset to tell the router that 555 it cannot honor the request. When a router receives this, the router 556 SHOULD attempt to connect to any more preferred caches in its cache 557 list. If there are no more preferred caches it MUST issue a Reset 558 Query and get an entire new load from the cache 560 5.4. Cache has No Data Available 562 Cache Router 563 ~ ~ 564 | <----- Serial Query ------ | R requests data 565 | ---- Error Report PDU ----> | C cannot supply update 566 ~ ~ 568 Cache Router 569 ~ ~ 570 | <----- Reset Query ------- | R requests data 571 | ---- Error Report PDU ----> | C cannot supply update 572 ~ ~ 574 The cache may respond to either a Serial Query or a Reset Query 575 informing the router that the cache cannot supply any update at all. 576 The most likely cause is that the cache has lost state, perhaps due 577 to a restart, and has not yet recovered. While it is possible that a 578 cache might go into such a state without dropping any of its active 579 sessions, a router is more likely to see this behavior when it 580 initially connects and issues a Reset Query while the cache is still 581 rebuilding its database. 583 When a router receives this kind of error, the router SHOULD attempt 584 to connect to any other caches in its cache list, in preference 585 order. If no other caches are available, the router MUST issue 586 periodic Reset Queries until it gets a new usable load from the cache 588 6. SSH Transport 590 The transport layer session between a router and a cache carries the 591 binary Protocol Data Units (PDUs) in a persistent SSH session. 593 To run over SSH, the client router first establishes an SSH transport 594 connection using the SSH transport protocol, and the client and 595 server exchange keys for message integrity and encryption. The 596 client then invokes the "ssh-userauth" service to authenticate the 597 application, as described in the SSH authentication protocol RFC 4252 598 [RFC4252]. Once the application has been successfully authenticated, 599 the client invokes the "ssh-connection" service, also known as the 600 SSH connection protocol. 602 After the ssh-connection service is established, the client opens a 603 channel of type "session", which results in an SSH session. 605 Once the SSH session has been established, the application invokes 606 the application transport as an SSH subsystem called "rpki-rtr". 607 Subsystem support is a feature of SSH version 2 (SSHv2) and is not 608 included in SSHv1. Running this protocol as an SSH subsystem avoids 609 the need for the application to recognize shell prompts or skip over 610 extraneous information, such as a system message that is sent at 611 shell start-up. 613 It is assumed that the router and cache have exchanged keys out of 614 band by some reasonably secured means. 616 7. Router-Cache Set-Up 618 A cache has the public authentication data for each router it is 619 configured to support. 621 A router may be configured to peer with a selection of caches, and a 622 cache may be configured to support a selection of routers. So each 623 must have the name of each peer and authentication data for each. In 624 addition, in a router, this list has a non-unique preference value 625 for each server in order of preference. The client router attempts 626 to establish a session with each potential serving cache in 627 preference order, and then starts to load data from the highest 628 preference cache to which it can connect and authenticate. The 629 router's list of caches has the following elements: 631 Preference: A cardinal denoting the router's preference to use that 632 cache, the lower the value the more preferred. 634 Name: The IP Address or fully qualified domain name of the cache. 636 Key: The public ssh key of the cache. 638 MyKey: The private ssh key of this client. 640 As caches can not be rigorously synchronous, a client which changes 641 servers can not combine data from different parent caches. 642 Therefore, when a lower preference cache becomes available, if 643 resources allow, it would be prudent for the client to start a new 644 buffer for that cache's data, and only switch to those data when that 645 buffer is fully up to date. 647 8. Deployment Scenarios 649 For illustration, we present three likely deployment scenarios. 651 Small End Site: The small multi-homed end site may wish to outsource 652 the RPKI cache to one or more of their upstream ISPs. They would 653 exchange authentication material with the ISP using some out of 654 band mechanism, and their router(s) would connect to one or more 655 up-streams' caches. The ISPs would likely deploy caches intended 656 for customer use separately from the caches with which their own 657 BGP speakers peer. 659 Large End Site: A larger multi-homed end site might run one or more 660 caches, arranging them in a hierarchy of client caches, each 661 fetching from a serving cache which is closer to the global RPKI. 662 They might configure fall-back peerings to up-stream ISP caches. 664 ISP Backbone: A large ISP would likely have one or more redundant 665 caches in each major PoP, and these caches would fetch from each 666 other in an ISP-dependent topology so as not to place undue load 667 on the global RPKI publication infrastructure. 669 Experience with large DNS cache deployments has shown that complex 670 topologies are ill-advised as it is easy to make errors in the graph, 671 e.g. not maintaining a loop-free condition. 673 Of course, these are illustrations and there are other possible 674 deployment strategies. It is expected that minimizing load on the 675 global RPKI servers will be a major consideration. 677 To keep load on global RPKI services from unnecessary peaks, it is 678 recommended that primary caches which load from the distributed 679 global RPKI not do so all at the same times, e.g. on the hour. 680 Choose a random time, perhaps the ISP's AS number modulo 60 and 681 jitter the inter-fetch timing. 683 9. Error Codes 685 This section contains a preliminary list of error codes. The authors 686 expect additions to this section during development of the initial 687 implementations. Eventually, these error codes will probably need to 688 reside in an IANA registry. 690 0: Reserved. 692 1: Internal Error: The party reporting the error experienced some 693 kind of internal error unrelated to protocol operation (ran out of 694 memory, a coding assertion failued, etcetera). 696 2: No Data Available: The cache believes itself to be in good 697 working order, but is unable to answer either a Serial Query or a 698 Reset Query because it has no useful data available at this time. 699 This is likely to be a temporary error, and most likely indicates 700 that the cache has not yet completed pulling down an initial 701 current data set from the global RPKI system after some kind of 702 event that invalidated whatever data it might have previously held 703 (reboot, network partition, etcetera). 705 10. Security Considerations 707 As this document describes a security protocol, many aspects of 708 security interest are described in the relevant sections. This 709 section points out issues which may not be obvious in other sections. 711 Cache Validation: In order for a collection of caches as described 712 in Section 8 to guarantee a consistent view, they need to be given 713 consistent trust anchors to use in their internal validation 714 process. Distribution of a consistent trust anchor is assumed to 715 be out of band. 717 Cache Peer Identification: As the router initiates an ssh transport 718 session to a cache which it identifies by either IP address or 719 fully qualified domain name, a DNS or address spoofing attack 720 could make the correct cache unreachable. No session would be 721 established, as the authorization keys would not match. 723 Transport Security: The RPKI relies on object, not server or 724 transport, trust. I.e. the IANA root trust anchor is distributed 725 to all caches through some out of band means, and can then be used 726 by each cache to validate certificates and ROAs all the way down 727 the tree. The inter-cache relationships are based on this object 728 security model, hence the inter-cache transport can be lightly 729 protected. 731 But this protocol document assumes that the routers can not do the 732 validation cryptography. Hence the last link, from cache to 733 router, is secured by server authentication and transport level 734 security. This is dangerous, as server authentication and 735 transport have very different threat models than object security. 737 So the strength of the trust relationship and the transport 738 between the router(s) and the cache(s) are critical. You're 739 betting your routing on this. 741 While we can not say the cache must be on the same LAN, if only 742 due to the issue of an enterprise wanting to off-load the cache 743 task to their upstream ISP(s), locality, trust, and control are 744 very critical issues here. The cache(s) really SHOULD be as 745 close, in the sense of controlled and protected (against DDoS, 746 MITM) transport, to the router(s) as possible. It also SHOULD be 747 topologically close so that a minimum of validated routing data 748 are needed to bootstrap a router's access to a cache. 750 11. Glossary 752 The following terms are used with special meaning: 754 Global RPKI: The authoritative data of the RPKI are published in a 755 distributed set of servers at the IANA, RIRs, NIRs, and ISPs, see 756 [I-D.ietf-sidr-repos-struct]. 758 Non-authorative Cache: A coalesced copy of the RPKI which is 759 periodically fetched/refreshed directly or indirectly from the 760 global RPKI using the [rcynic] protocol/tools 762 Cache: The rcynic system is used to gather the distributed data of 763 the RPKI into a validated cache. Trusting this cache further is a 764 matter between the provider of the cache and a relying party. 766 Serial Number: A 32-bit monotonically increasing, cardinal which 767 wraps from 2^32-1 to 0. It denotes the logical version of a 768 cache. A cache increments the value by one when it successfully 769 updates its data from a parent cache or from primary RPKI data. 770 As a cache is rcynicing, new incoming data, and implicit deletes, 771 are marked with the new serial but MUST not be sent until the 772 fetch is complete. A serial number is not commensurate between 773 caches, nor need it be maintained across resets of the cache 774 server. See [RFC1982] on DNS Serial Number Arithmetic for too 775 much detail on serial number arithmetic. 777 12. IANA Considerations 779 This document requests the IANA to create a registry for PDU types. 781 This document requests the IANA to create a registry for Error Codes. 783 In addition, a registry for Version Numbers would be needed if new 784 Version Number is defined in a new RFC. 786 Note to RFC Editor: this section may be replaced on publication as an 787 RFC. 789 13. Acknowledgments 791 The authors wish to thank Steve Bellovin, Rex Fernando, Russ Housley, 792 Pradosh Mohapatra, Sandy Murphy, Megumi Ninomiya, Robert Raszuk, John 793 Scudder, Ruediger Volk, David Ward, and Bert Wijnen. 795 14. References 797 14.1. Normative References 799 [RFC1982] Elz, R. and R. Bush, "Serial Number Arithmetic", RFC 1982, 800 August 1996. 802 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 803 Requirement Levels", RFC 2119, BCP 14, March 1997. 805 [RFC4252] Ylonen, T. and C. Lonvick, "The Secure Shell (SSH) 806 Authentication Protocol", RFC 4252, January 2006. 808 [rcynic] Austein, R., "rcynic protocol", 809 . 811 14.2. Informative References 813 [I-D.ietf-sidr-arch] 814 Lepinski, M. and S. Kent, "An Infrastructure to Support 815 Secure Internet Routing", draft-ietf-sidr-arch-06 (work in 816 progress), March 2009. 818 [I-D.ietf-sidr-repos-struct] 819 Huston, G., Loomans, R., and G. Michaelson, "A Profile for 820 Resource Certificate Repository Structure", 821 draft-ietf-sidr-repos-struct-01 (work in progress), 822 October 2008. 824 [RFC1996] Vixie, P., "A Mechanism for Prompt Notification of Zone 825 Changes (DNS NOTIFY)", RFC 1996, August 1996. 827 Authors' Addresses 829 Randy Bush 830 Internet Initiative Japan, Inc. 831 5147 Crystal Springs 832 Bainbridge Island, Washington 98110 833 US 835 Phone: +1 206 780 0431 x1 836 Email: randy@psg.com 838 Rob Austein 839 Internet Systems Consortium 840 950 Charter Street 841 Redwood City, CA 94063 842 USA 844 Email: sra@isc.org