idnits 2.17.1 draft-ietf-sidr-ltamgmt-00.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 == Line 254 has weird spacing: '...ntifier unc...' == Line 257 has weird spacing: '... points repl...' == Line 330 has weird spacing: '...lagname boole...' == Line 360 has weird spacing: '...tagname tagva...' -- The document date (November 9, 2010) is 4909 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) == Missing Reference: 'I-D.sidr-arch' is mentioned on line 120, but not defined == Missing Reference: 'I-D.sidr-res-cert-prof' is mentioned on line 124, but not defined == Missing Reference: 'ID.sidr-res-cert-prof' is mentioned on line 454, but not defined == Unused Reference: 'RFC2119' is defined on line 1076, but no explicit reference was found in the text == Unused Reference: 'RFC3513' is defined on line 1079, but no explicit reference was found in the text == Unused Reference: 'RFC5396' is defined on line 1090, but no explicit reference was found in the text ** Obsolete normative reference: RFC 3513 (Obsoleted by RFC 4291) Summary: 1 error (**), 0 flaws (~~), 11 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Secure Inter-Domain Routing M. Reynolds 3 Internet-Draft S. Kent 4 Intended status: Standards Track BBN 5 Expires: May 13, 2011 November 9, 2010 7 Local Trust Anchor Management for the Resource Public Key Infrastructure 8 10 Status of this Memo 12 This Internet-Draft is submitted to IETF in full conformance with the 13 provisions of BCP 78 and BCP 79. 15 Internet-Drafts are working documents of the Internet Engineering 16 Task Force (IETF), its areas, and its working groups. Note that 17 other groups may also distribute working documents as 18 Internet-Drafts. 20 Internet-Drafts are draft documents valid for a maximum of six months 21 and may be updated, replaced, or obsoleted by other documents at any 22 time. It is inappropriate to use Internet-Drafts as reference 23 material or to cite them other than as "work in progress." 25 The list of current Internet-Drafts can be accessed at 26 http://www.ietf.org/1id-abstracts.html 28 The list of Internet-Draft Shadow Directories can be accessed at 29 http://www.ietf.org/shadow.html 31 This Internet-Draft will expire on May 13, 2011. 33 Copyright and License Notice 35 Copyright (c) 2010 IETF Trust and the persons identified as the 36 document authors. All rights reserved. 38 This document is subject to BCP 78 and the IETF Trust's Legal 39 Provisions Relating to IETF Documents 40 (http://trustee.ietf.org/license-info) in effect on the date of 41 publication of this document. Please review these documents 42 carefully, as they describe your rights and restrictions with respect 43 to this document. Code Components extracted from this document must 44 include Simplified BSD License text as described in Section 4.e of 45 the Trust Legal Provisions and are provided without warranty as 46 described in the Simplified BSD License. 48 Abstract 50 This document describes a facility to enable a relying party (RP) to 51 manage trust anchors (TAs) in the context of the Resource Public Key 52 Infrastructure (RPKI). It is common to allow an RP to import TA 53 material in the form of self-signed certificates. The facility 54 described in this document allows an RP to impose constraints on such 55 TAs. Because this mechanism is designed to operate in the RPKI 56 context, the relevant constraints are the RFC 3779 extensions that 57 bind address spaces and/or autonomous system (AS) numbers to 58 entities. The primary motivation for this facility is to enable an RP 59 to ensure that resource allocation information that it has acquired 60 via some trusted channel is not overridden by the information 61 acquired from the RPKI repository system or by the putative TAs that 62 the RP imports. Specifically, the mechanism allows an RP to specify a 63 set of bindings between public key identifiers and RFC 3779 extension 64 data and will override any conflicting bindings expressed via the 65 putative TAs and the certificates downloaded from the RPKI repository 66 system. Although this mechanism is designed for local use by an RP, 67 an entity that is accorded administrative control over a set of RPs 68 may use this mechanism to convey its view of the RPKI to a set of RPs 69 within its jurisdiction. The means by which this latter use case is 70 effected is outside the scope of this document. 72 Table of Contents 74 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 4 75 1.1 Terminology . . . . . . . . . . . . . . . . . . . . . . . . 5 76 2 Overview of Certificate Processing . . . . . . . . . . . . . . . 5 77 2.1 Target Certificate Processing . . . . . . . . . . . . . . . 5 78 2.2 Perforation . . . . . . . . . . . . . . . . . . . . . . . . 5 79 2.3 TA Re-parenting . . . . . . . . . . . . . . . . . . . . . . 6 80 2.4 Paracertificates . . . . . . . . . . . . . . . . . . . . . 6 81 3 Format of the constraints file . . . . . . . . . . . . . . . . . 8 82 3.1 Relying party subsection . . . . . . . . . . . . . . . . . 8 83 3.2 Flags subsection . . . . . . . . . . . . . . . . . . . . . 8 84 3.3 Tags subsection . . . . . . . . . . . . . . . . . . . . . . 9 85 3.3.1 Xvalidity_dates tag . . . . . . . . . . . . . . . . 10 86 3.3.2 Xcrldp tag . . . . . . . . . . . . . . . . . . . . . 10 87 3.3.3 Xcp tag . . . . . . . . . . . . . . . . . . . . . . 11 88 3.3.4 Xaia tag . . . . . . . . . . . . . . . . . . . . . . 11 89 3.4 Blocks subsection . . . . . . . . . . . . . . . . . . . . 12 90 4 Certificate Processing Algorithm . . . . . . . . . . . . . . . 13 91 4.1 Proofreading algorithm . . . . . . . . . . . . . . . . . 14 92 4.2 TA processing algorithm . . . . . . . . . . . . . . . . . 15 93 4.2.1 Preparatory processing (stage 0) . . . . . . . . . . 16 94 4.2.2 Target processing (stage 1) . . . . . . . . . . . . 17 95 4.2.3 Ancestor processing (stage 2) . . . . . . . . . . . 18 96 4.2.4 Tree processing (stage 3) . . . . . . . . . . . . . 19 97 4.2.5 TA re-parenting (stage 4) . . . . . . . . . . . . . 20 98 4.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . 21 99 5 Implications for Path Discovery . . . . . . . . . . . . . . . 21 100 5.1 Two answers . . . . . . . . . . . . . . . . . . . . . . . 21 101 5.2 One answer . . . . . . . . . . . . . . . . . . . . . . . 22 102 5.3 No answer . . . . . . . . . . . . . . . . . . . . . . . . 22 103 6 Implications for Revocation . . . . . . . . . . . . . . . . . 22 104 6.1 No state bits set . . . . . . . . . . . . . . . . . . . . 22 105 6.2 ORIGINAL state bit set . . . . . . . . . . . . . . . . . 23 106 6.3 PARA state bit set . . . . . . . . . . . . . . . . . . . 23 107 6.4 Both ORIGINAL and PARA state bits set . . . . . . . . . . 23 108 7 Security Considerations . . . . . . . . . . . . . . . . . . . 24 109 8 IANA Considerations . . . . . . . . . . . . . . . . . . . . . 24 110 9 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 24 111 10 References . . . . . . . . . . . . . . . . . . . . . . . . . 24 112 10.1 Normative References . . . . . . . . . . . . . . . . . . 24 113 10.2 Informative References . . . . . . . . . . . . . . . . . 25 114 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 25 115 Appendix A: Sample Constraints File . . . . . . . . . . . . . . . 26 116 Appendix B: Optional Sorting Algorithm for Ancestor Processing . 27 118 1 Introduction 120 The Resource Public Key Infrastructure (RPKI) [I-D.sidr-arch] is a 121 PKI in which certificates are issued to facilitate management of IP 122 addresses and autonomous system number resources. Such resources are 123 expressed in the form of X.509v3 "resource" certificates with 124 extensions as defined by RFC 3779 [I-D.sidr-res-cert-prof]. 125 Validation of a resource certificate is preceded by path discovery. 126 Path discovery is effected by constructing a certificate path 127 (upward) from a target certificate to a trust anchor. Path validation 128 proceeds from the TA in question to the target certificate, using the 129 public key from each certificate along the path to verify the 130 signature of its subordinate certificate. In the RPKI it is 131 anticipated that one or more putative TAs, aligned with the resource 132 allocation hierarchy, will be available in the form of self-signed 133 certificates configured by an RP. There are circumstances under which 134 an RP may wish to override the resource specifications obtained 135 through the RPKI distributed repository system [I-D.sidr-repos- 136 struct]. This document describes a mechanism by which an RP may 137 override any conflicting information expressed via the putative TAs 138 and the certificates downloaded from the RPKI repository system. 140 To effect this local control, this document calls for a relying party 141 to specify a set of bindings between public key identifiers and 142 resources (IP resources and/or AS number resources) through a text 143 file known as a constraints file. The constraints expressed in this 144 file then take precedence over any competing claims expressed by 145 resource certificates acquired from the distributed repository 146 system. (The means by which a relying party acquires the key 147 identifier and the RFC 3779 extension data used to populate the 148 constraints file is outside the scope of this document.) The relying 149 party also may use a local publication point (the root of a local 150 directory tree that is made available as if it were a remote 151 repository) as a source of certificates and CRLs (and other RPKI 152 signed objects, e.g. ROAs and manifests) that do not appear in the 153 RPKI repository system. 155 In order to allow reuse of existing, standard path validation 156 mechanisms, the RP-imposed constraints are realized by having the RP 157 itself represented as the only TA known in the local certificate 158 validation context. To ensure that all RPKI certificates can be 159 validated relative to this TA, this RP TA certificate must contain 160 all-encompassing resource allocations, i.e. 0/0 for IPv4, 0::/0 for 161 IPv6 and 0-4294967295 for AS numbers. Thus, a conforming 162 implementation of this mechanism must be able to cause a self-signed 163 certification authority (CA) certificate to be created with a locally 164 generated key pair. It also must be able to issue CA certificates 165 subordinate to this TA. Finally, a conforming implementation of this 166 mechanism must process the constraints file and modify certificates 167 as needed in order to enforce the constraints asserted in the file. 169 The remainder of this document describes in detail the types of 170 certificate modification that may occur, the semantics of the 171 constraints file, and the implications of certificate modification on 172 path discovery and revocation. 174 1.1 Terminology 176 It is assumed that the reader is familiar with the terms and concepts 177 described in "Internet X.509 Public Key Infrastructure Certificate 178 and Certificate Revocation List (CRL) Profile" [RFC5280] and "X.509 179 Extensions for IP Addresses and AS Identifiers" [RFC3779]. 181 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 182 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 183 document are to be interpreted as described in RFC 2119. 185 2 Overview of Certificate Processing 187 The fundamental aspect of the facility described in this document is 188 one of certificate modification. The constraints file, described in 189 more detail in the next section, contains assertions about resources 190 that are to be specially processed. As a result of this processing, 191 certificates in the local copy of the RPKI repository are transformed 192 into new certificates satisfying the resource constraints so 193 specified. This enables the RP to override conflicting assertions 194 about resource holdings as acquired from the RPKI repository system. 195 Three forms of certificate modification can occur. 197 2.1 Target Certificate Processing 199 If a certificate is acquired from the RPKI repository system and it's 200 SKI is listed in the constraints file, it will be reissued directly 201 under the RP TA certificate, with (possibly) modified RFC 3779 202 extensions. The modified extensions will include any RFC 3779 data 203 expressed in the constraints file. In Section 4.2, target certificate 204 processing corresponds to stage one of the algorithm. 206 2.2 Perforation 208 Any certificate acquired from the RPKI repository that contains an 209 RFC 3779 extension that intersects the resource data in the 210 constraints file will be reissued directly under the RP TA, with 211 modified RFC 3779 extensions. We refer to the process of modifying 212 the RFC 3779 extension in an affected certificate as "perforation" 213 (because the process will create "holes" in these extensions). The 214 modified extensions will exclude any RFC 3779 data expressed in the 215 constraints file. In the certificate processing algorithm described 216 in Section 4.2, perforation corresponds to stage two of the algorithm 217 ("ancestor processing") and also to stage three of the algorithm 218 ("tree processing"). 220 2.3 TA Re-parenting 222 For consistency, all valid, self-signed certificates that would have 223 been regarded as TAs in the public RPKI certificate hierarchy, e.g. 224 self-signed certificates issued by IANA or the RIRs, will be re- 225 issued under the RP TA certificate. This processing is done even 226 though all but one of these certificates might not intersect any 227 resources specified in the constraints file. We refer to this 228 reissuance as "re-parenting" since the Issuer (parent) of the 229 certificate has been changed. In the certificate processing algorithm 230 described in Section 4.2, TA re-parenting corresponds to stage four 231 of the algorithm. 233 2.4 Paracertificates 235 If a certificate is subject to any of the three forms of processing 236 just described, that certificate will be referred to as an "original" 237 certificate and the processed (output) certificate will be referred 238 to as a paracertificate. When an original certificate is transformed 239 into a paracertificate all the fields and extensions from the 240 original certificate will be retained, except as indicated in Table 241 1, below. 243 Original Certificate Field Action 245 Version unchanged 246 Serial number created per note A 247 Signature replaced if needed 248 with RP's signing alg 249 Issuer replaced with RP's name 250 Validity dates replaced per note B 251 Subject unchanged 252 Subject public key info unchanged 253 Extensions 254 Subject key identifier unchanged 255 Key usage unchanged 256 Basic constraints unchanged 257 CRL distribution points replaced per note B 258 Certificate policy replaced per note B 259 Authority info access replaced per note B 260 Authority key ident replaced with RP's 261 IP address block modified as described 262 AS number block modified as described 263 Subject info access unchanged 264 All other extensions unchanged 265 Signature Algorithm same as above 266 Signature value new 268 Table 1 Certificate Field Modifications 270 Note A. The serial number will be created by concatenating the 271 current time (the number of seconds since Jan 1, 1970) with a count 272 of the certificates created in the current run. 274 Note B. These fields are derived (as described in section 3.3 below) 275 from parameters in the constraints file (if present); otherwise, they 276 take on values from the certificates from which the paracertificates 277 are derived. 279 3 Format of the constraints file 281 This section describes a general model for the syntax of the 282 constraints file. The model described below is nominal; 283 implementations need not match details of this model as presented, 284 but the external behavior of implementations MUST correspond to the 285 externally observable characteristics of this model in order to be 286 compliant. 288 The constraints file consists of four logical subsections: the 289 replying party subsection, the flags subsection, the tags subsection 290 and the blocks subsection. The relying party subsection and the 291 blocks subsection are REQUIRED and MUST be present; the flags and 292 tags subsections are OPTIONAL. Each subsection is described in more 293 detail below. Note that the semicolon (;) character acts as the 294 comment character, to enable annotating constraints files. All 295 characters from a semicolon to the end of that line are ignored. In 296 addition, lines consisting only of whitespace are ignored. The 297 subsections MUST occur in the order indicated. An example constraints 298 file is given in Appendix A. 300 3.1 Relying party subsection 302 The relying party subsection is a REQUIRED subsection of the 303 constraints file. It MUST be the first subsection of the constraints 304 file, and it MUST consist of two lines of the form: 306 PRIVATEKEYMETHOD value [ ... value ] 307 TOPLEVELCERTIFICATE value 309 The first line provides guidance to the certificate processing 310 algorithm on the method that will be used to gain access to the RP's 311 private key. This line consists of the string literal 312 PRIVATEKEYMETHOD, followed by one or more whitespace delimited string 313 values. These values are passed to the certificate processing 314 algorithm as described below. Note that this entry, as for all 315 entries in the constraints file, is case sensitive. 317 The second line of this subsection consists of the string literal 318 TOPLEVELCERTIFICATE, followed by exactly one string value. This value 319 is the name of a file containing the relying party's TA certificate. 320 The file name is passed to the certificate processing algorithm as 321 described below. 323 3.2 Flags subsection 325 The flags subsection of the constraints file is an OPTIONAL 326 subsection. If present it MUST immediately follow the relying party 327 subsection. The flags subsection consists of one or more lines of the 328 form 330 CONTROL flagname booleanvalue 332 Each such line is referred to as a control line. Each control line 333 MUST contain exactly three whitespace delimited strings. The first 334 string MUST be the literal CONTROL. The second string MUST be one of 335 the following three literals: 337 resource_nounion 338 intersection_always 339 treegrowth 341 The third string denotes a boolean value, and MUST be one of the 342 literals TRUE or FALSE. Control flags influence the global operation 343 of the certificate processing algorithm; the semantics of the flags 344 is described in detail in Section 4.2. Note that each flag has a 345 default value, so that if the corresponding CONTROL line does not 346 appear in the constraints file, the algorithm flag is considered to 347 take the corresponding default value. The default value for each flag 348 is FALSE. Thus, if any flag is not named in a control line it takes 349 the value FALSE. Further, if the flags subsection is absent, all 350 three flags take the value FALSE. 352 3.3 Tags subsection 354 The tags subsection is an OPTIONAL subsection in the constraints 355 file. If present it MUST immediately follow the relying party 356 subsection (if the flags subsection is absent) or the flags 357 subsection (if it is present). The tags subsection consists of one or 358 more lines of the form 360 TAG tagname tagvalue [ ... tagvalue ] 362 Each such line is referred to as a tag line. Each tag line MUST 363 consist of at least three whitespace delimited string values, the 364 first of which must be the literal TAG. The second string value gives 365 the name of the tag, and subsequent string(s) give the value(s) of 366 the tag. The tag name MUST be one of the following four string 367 literals: 369 Xvalidity_dates 370 Xcrldp 371 Xcp 372 Xaia 374 The purpose of the tag lines is to provide an indication of the means 375 by which paracertificate fields, specifically those indicated above 376 under "Note B", are constructed. Each tag has a default, so that if 377 the corresponding tag line is not present in the constraints file, 378 the default behavior is used when constructing the paracertificates. 379 The syntax and semantics of each tag line is described next. 381 Note that the tag lines are considered to be global; the action of 382 each tag line (or the default action, if that tag line is not 383 present) applies to all paracertificates that are created as part of 384 the certificate processing algorithm. 386 3.3.1 Xvalidity_dates tag 388 This tag line is used to control the value of the notBefore and 389 notAfter fields in paracertificates. If this tag line is specified 390 and there is a single tagvalue which is the literal string C, the 391 paracertificate validity interval is copied from the original 392 certificate validity interval from which it is derived. If this tag 393 is specified and there is a single tagvalue which is the literal 394 string R, the paracertificate validity interval is copied from the 395 validity interval of the relying party's top level (TA) certificate. 396 If this tag is specified and the tagvalue is neither of these 397 literals, then exactly two tagvalues MUST be specified. Each must be 398 a Generalized Time string of the form YYYYMMDDHHMMSSZ. The first 399 tagvalue is assigned to the notBefore field and the second tagvalue 400 is assigned to the notAfter field. It MUST be the case that the 401 tagvalues may be parsed as valid Generalized Time strings such that 402 notBefore is less than notAfter, and also such that notAfter 403 represents a time in the future (i.e., the paracertificate has not 404 already expired). 406 If this tag line is not present in the constraints file the default 407 behavior is to copy the validity interval from the original 408 certificate to the corresponding paracertificate. 410 3.3.2 Xcrldp tag 412 This tag line is used to control the value of the CRL distribution 413 point extension in paracertificates. If this tag line is specified 414 and there is a single tagvalue that is the string literal C, the 415 CRLDP of the paracertificate is copied from the CRLDP of the original 416 certificate from which it is derived. If this tag line is specified 417 and there is a single tagvalue that is the string literal R, the 418 CRLDP of the paracertificate is copied from the CRLDP of the RP's top 419 level certificate. If this tag line is specified and there is a 420 single tagvalue that is not one of these two reserved literals, or if 421 there is more than one tagvalue, then each tagvalue is interpreted as 422 a URL that will be placed in the CRLDP sequence in the 423 paracertificate. 425 If this tag line is not present in the constraints file the default 426 behavior is to copy the CRLDP from the original certificate to the 427 corresponding paracertificate. 429 3.3.3 Xcp tag 431 This tag line is used to control the value of the policyQualifierId 432 field in paracertificates. If this tag line is specified there MUST 433 be exactly one tagvalue. If the tagvalue is the string literal C, the 434 paracertificate value is copied from the value in the corresponding 435 original certificate. If the tagvalue is the string literal R, the 436 paracertificate value is copied from the value in the RP's top level 437 TA certificate. If the tagvalue is the string literal D, the 438 paracertificate value is set to the default OID. If the tagvalue is 439 not one of these reserved string literals, then the tagvalue MUST be 440 an OID specified using the standard dotted notation. The value in the 441 paracertificate's policyQualifierId field is set to this OID. Note 442 the RFC 5280 specifies that only a single policy may be specified in 443 a certificate, so only a single tagvalue is permitted in this tag 444 line, even though the CertificatePolicy field is an ASN.1 sequence. 446 If this tag line is not specified the default behavior is to use the 447 default OID in creating the paracertificate. 449 This option permits the RP to convert a value of the 450 policyQualifierId field in a certificate (that would not be in 451 conformance with the RPKI CP) to a conforming value in the 452 paracertificate. This conversion enables use of RPKI validation 453 software that checks the policy field against that specified in the 454 RPKI CP [ID.sidr-res-cert-prof]. 456 3.3.4 Xaia tag 458 This tag line is used to control the value of the Authority 459 Information Access (AIA) extension in the paracertificate. If this 460 tag line is present then it MUST have exactly one tagvalue. If this 461 tagvalue is the string literal C, then the AIA field in the 462 paracertificate is copied from the AIA field in the original 463 certificate from which it is derived. If this tag line is present and 464 the tagvalue is not the reserved string literal, then the tagvalue 465 MUST be a URL. This URL is set as the AIA extension of the 466 paracertificates that are created. 468 If this tag line is not specified the default behavior is to use copy 469 the AIA field from the original certificate to the AIA field of the 470 paracertificate. 472 3.4 Blocks subsection 474 The blocks subsection is a REQUIRED subsection of the constraints 475 file. If the tags subsection is present, the blocks subsection MUST 476 appear immediately after it. If the tags subsection is absent, but 477 the flags subsection is present, the block subsection MUST appear 478 immediately after it. Otherwise, the blocks subsection MUST appear 479 immediately after the relying party subsection. The blocks subsection 480 consists of one or more blocks, known as target blocks. A target 481 block is used to specify an association between a certificate (given 482 by a hash of its public key information) and a set of resource 483 assertions. Each target block contains four regions, an SKI region, 484 an IPv4 region, an IPv6 region and an AS number region. All regions 485 are REQUIRED to be present in a target block. 487 The SKI region contains a single line beginning with the string 488 literal SKI and followed by forty hexadecimal characters giving the 489 subject key identifier of a certificate, known as the target 490 certificate. The hex character string MAY contain embedded whitespace 491 or colon characters (included to improve readability), which are 492 ignored. The IPv4 region consists of a line containing only the 493 string literal IPv4. This line is followed by zero or more lines 494 containing IPv4 prefixes in the format described in RFC 3779. The 495 IPv6 region consists of a line containing only the string literal 496 IPv6, followed by zero or more lines containing IPv6 prefixes using 497 the format described in RFC 3513. (The presence of the IPv4 and IPv6 498 literals is to simplify parsing of the constraints file.) Finally, 499 the AS number region consists of a line containing only the string 500 literal AS#, followed by zero or more lines containing AS numbers 501 (one per line). The AS numbers are specified in decimal notation as 502 recommended in RFC 5396. A target block is terminated by either the 503 end of the constraints file, or by the beginning of the next target 504 block, as signaled by its opening SKI region line. An example target 505 block is shown below. See also the complete constraints file example 506 given in Appendix A. Note that whitespace, as always, is ignored. 508 SKI 00:12:33:44:00:BA:BA:DE:EB:EE:00:99:88:77:66:55:44:33:22:11 509 IPv4 510 10.2.3/24 511 10.8/16 512 IPv6 513 1:2:3:4:5:6/112 514 AS# 515 123 516 567 518 The blocks subsection MUST contain at least one target block. Note 519 that it is OPTIONAL that the SKI refer to a certificate that is known 520 or resolvable within the context of the local RPKI repository. Also, 521 there is no REQUIRED or implied ordering of target blocks within the 522 block subsection. As a result of the fact that blocks may occur in 523 any order, it MAY result that the outcome of processing a constraints 524 file depends on the order in which target blocks occur within the 525 constraints file. The next section of this document contains a 526 detailed description of the certificate processing algorithm. 528 4 Certificate Processing Algorithm 530 The section describes the certificate processing algorithm through 531 which paracertificates are created from original certificates in the 532 local RPKI repository. For the purposes of describing this algorithm, 533 it will be assumed that certificates may be persistently associated 534 with state (or metadata) information. This state information will be 535 further construed as having the form of any array of named bits that 536 are associated with each certificate. No specific implementation of 537 this functionality is mandated by this document. Any implementation 538 that provides the indicated functionality is acceptable, and need not 539 actually consist of a bit field associated with each certificate. 541 The state bits used in certificate processing are 543 NOCHAIN 544 ORIGINAL 545 PARA 546 TARGET 548 If the NOCHAIN bit is set, this indicates that a full path between 549 the given certificate and a TA has not yet been discovered. If the 550 ORIGINAL bit is set, this indicates that the certificate is question 551 has been processed by some part of the processing algorithm described 552 in Section 4.2. If it was processed as part of stage one processing, 553 as described in section 4.2.2, the TARGET bit will also be set. 554 Finally, any paracertificate will have the PARA bit set. 556 At the beginning of algorithm processing each certificate in the 557 local RPKI repository has the ORIGINAL, PARA and TARGET bits clear. 558 If a certificate has a complete, validated path to a TA, or is itself 559 a TA, then that certificate will have the NOCHAIN bit clear, 560 otherwise it will have the NOCHAIN bit set. As the certificate 561 processing algorithm is executed, the bit state of original 562 certificates may changed. In addition, since the certificate 563 processing algorithm may also be creating paracertificates, it is 564 responsible for actively setting or clearing the state of these four 565 bits on those paracertificates. 567 The certificate processing algorithm consists of two sub-algorithms: 569 "proofreading" and "TA processing". Conceptually, the proofreading 570 sub-algorithm performs syntactic checks on the constraints file, 571 while the TA processing sub-algorithm performs the actual certificate 572 transformation processing. If the proofreading sub-algorithm does not 573 succeed in parsing the constraints file, the TA processing sub- 574 algorithm is not executed. Note also that if the constraints file is 575 not present, neither sub-algorithm is executed and the local RPKI 576 repository is not modified. Each of the constituent algorithms will 577 now be described in detail. 579 4.1 Proofreading algorithm 581 The goal of the proofreading algorithm is to check the constraints 582 file for syntactic errors, such as missing REQUIRED subsections, or 583 malformed addresses such as 1.2.300/24. It also performs a set of 584 heuristic checks, such as checking for prefixes that are too large 585 (larger than /8). The proofreading algorithm SHOULD also examine 586 resource regions (IPv4, IPv6 and AS# regions) within the blocks 587 subsection, and reorder such resources within a region in ascending 588 numeric order. On encountering any error the proofreading algorithm 589 SHOULD provide an error message indicating the line on which the 590 error occurred as well as informative text that is sufficiently 591 descriptive as to allow the user to identify and correct the error. 592 An implementation of the proofreading algorithm MUST NOT assume that 593 is has access to the local RPKI repository (even read-only access). 594 An implementation of the proofreading algorithm MUST NOT alter the 595 local RPKI repository in any way; it also MUST NOT change any of the 596 state/metadata information associated with certificates in that 597 repository. (Recall that the processing described here is creating a 598 copy of that local repository.) Finally, the proofreading algorithm 599 MAY produce a transformed output file containing the same syntactic 600 information as in the text version of the constraints file, so long 601 as the format of the transformed file is understood by the TA 602 processing algorithm. 604 The proofreading algorithm performs the following syntactic checks on 605 the constraints file. It checks for the presence of the REQUIRED 606 relying party subsection and the REQUIRED blocks subsection. It 607 checks that the order of the two, three or four subsections is as 608 stated above. It checks that the relying party subsection conforms to 609 the specification given in section 3.1 above. If present, it checks 610 that the tags and flags subsections conform to the specifications in 611 sections 3.2 and 3.3 above. It then checks the blocks subsection. It 612 splits the blocks subsection into constituent target blocks, as 613 delimited by the SKI region line(s), and verifies that at least one 614 target block is present. It verifies that each SKI region line 615 contains exactly forty hexadecimal digits and contains no additional 616 characters other than whitespace or colon characters. For each target 617 block, it then verifies the presence of the IPv4, IPv6 and AS# 618 regions, and also verifies that at least one such resource is 619 present. For each IPv4 prefix, IPv6 prefix and autonomous system 620 number given, it checks that the indicated resource is syntactically 621 valid according to the appropriate RFC definition, as described in 622 section 3.4. It also verifies that no IPv4 resource has a prefix 623 larger than /8. The proofreading algorithm SHOULD performing 624 reordering within each of the three resource regions so that stated 625 resource occur in ascending numerical order. If the proofreading 626 algorithm has performed any reordering of information it MAY 627 overwrite the constraints file. If it does so, however, it MUST 628 preserve all information contained within the file, including 629 information that is not parsed (such as comments). If the 630 proofreading algorithm has performed any reordering of information 631 but has not overwritten the constraints file, it MAY produce a 632 transformed output file, as described above. If the proofreading 633 algorithm has performed any reordering of information, but has 634 neither overwritten the constraints file nor produced a transformed 635 output file, it MUST provide an error message to the user indicating 636 what reordering was performed. 638 4.2 TA processing algorithm 640 The TA processing algorithm acts on the constraints file (or the 641 output file produced by the proofreading algorithm) and the contents 642 of the local RPKI repository to produce paracertificates for the 643 purpose of enforcing the resource allocations as expressed in the 644 constraints file. The TA processing algorithm operates in five 645 stages, a preparatory stage (stage 0), target processing (stage 1), 646 ancestor processing (stage 2), tree processing (state 3) and TA re- 647 parenting (stage 4). Conceptually, during the preparatory stage the 648 constraints (or proofreader output) file is read and a set of 649 internal RP, tag and flag variables are set based on the contents of 650 that file. (If the constraint file has not specified one or more of 651 the tags and/or flags, those tags and flags are set to default 652 values). During target processing all certificates specified by a 653 target block are processed, and the resources for those certificates 654 are (potentially) expanded; for each target found a new 655 paracertificate is manufactured with its various fields set, as shown 656 in Table 1, using the values of the internal variables set in the 657 preparatory stage and also, of course, the fields of the original 658 certificate (and, potentially, fields of the RP's TA certificate). In 659 stage 2 (ancestor) processing, all ancestors of the each target 660 certificate are found, and the claimed resources are then removed 661 (perforated). A new paracertificate with these diminished resources 662 is crafted, with its fields generated based on internal variable 663 settings, original certificate field values, and, potentially, the 664 fields of the RP's TA certificate. In tree processing (stage 3), the 665 entire local RPKI repository is searching for any other certificates 666 that have resources that intersect a target resource, and that were 667 not otherwise processed during a preceding stage. Perforation is 668 again performed for any such intersecting certificates, and 669 paracertificates created as in stage 2. Finally, in the fourth and 670 last stage, TA re-parenting, any TA certificates in the local RPKI 671 repository that have not already been processed are now re-parented 672 under the RP's TA certificate. This transformation will create 673 paracertificates; however, these paracertificates may have RFC 3779 674 resources that were not altered during algorithm processing. The 675 final output of algorithm processing will be threefold. First, the 676 state/metadata information on some (original) certificates in the 677 repository MAY be altered. Second, paracertificates will be created, 678 with the appropriate metadata, and entered into the repository. 679 Finally, the TA processing algorithm SHOULD produce a human readable 680 log of its actions, indicating which paracertificates were created 681 and why. The remainder of this section describes the processing 682 stages of the algorithm in detail. 684 4.2.1 Preparatory processing (stage 0) 686 During preparatory processing, the constraints file, or the 687 corresponding output file of the proofreader algorithm, is read. 688 Internal variables are set corresponding to each tag and flag, if 689 present, or to their defaults, if absent. Internal variables are also 690 set corresponding to the PRIVATEKEYMETHOD value string(s) and the 691 TOPLEVELCERTIFICATE string. The TA processing algorithm is queried to 692 determine if it supports the indicated private key access 693 methodology. This query is performed in an implementation-specific 694 manner. In particular, an implementation is free to vacuously return 695 success to this query. The TA processing algorithm next uses the 696 value string for the TOPLEVELCERTIFICATE to locate this certificate, 697 again in an implementation=specific manner. The certificate in 698 question may already be present in the local RPKI repository, or it 699 may be located elsewhere. The implementation is also free to create 700 the top level certificate at this time, and then assign to this 701 newly-created certificate the name indicated. It is necessary only 702 that, at the conclusion of this processing, a valid trust anchor 703 certificate for the relying party has been created or otherwise 704 obtained. 706 Some form of access to the RP's private key and top level certificate 707 are required for subsequent correct operation of the algorithm. 708 Therefore, stage 0 processing MUST terminate if one or both 709 conditions are not satisfied. In the error case, the implementation 710 SHOULD provide an error message of sufficient detail that the user 711 can correct the error(s). If stage 0 processing does not succeed, no 712 further stages of TA processing are executed. 714 4.2.2 Target processing (stage 1) 716 During target processing, the TA processing algorithm reads all 717 target blocks in the constraints file or corresponding proofreader 718 output file. It then processes each target block in the order 719 specified in the file. In the description that follows, except where 720 noted, the operation of the algorithm on a single target block will 721 be described. Note, however, that all stage 1 processing is executed 722 before any processing in subsequent stages is performed. 724 The algorithm first obtains the SKI region of the target block. It 725 then locates, in an implementation-dependent manner, the certificate 726 the SKI extension field of which contains that value. Note that if 727 paracertificates have been created by virtue of previous target 728 blocks being processed, those paracertificates are not searched in 729 attempting to locate a certificate with a matching SKI; only original 730 certificates are searched. If more than one original certificate is 731 found matching this SKI, there are two possible scenarios. If a 732 resource holder has two certificates issued by the same CA, with 733 overlapping validity intervals and the same key, but distinct subject 734 names (typically, by virtue of the SerialNumber parts being 735 different), then these two certificates are both consider to be 736 (distinct) targets, and are both processed. If, however, a resource 737 holder has certificates issued by two different CAs, containing 738 different resources, but using the same key, there is no unambiguous 739 method to decide which of the certificates is intended as the target. 740 In this latter case the algorithm MUST issue a warning to that 741 effect, mark the target block in question as unavailable for 742 processing by subsequent stages and proceed to the next target block. 743 If no certificate is found then the algorithm SHOULD issue a warning 744 to that effect and proceed to process the next target block. 746 If a single original certificate is found matching the indicated SKI, 747 then the algorithm takes the following actions. First, it sets the 748 ORIGINAL state bit for the certificate found. Second, it sets the 749 TARGET state bit for the certificate found. Third, it extracts the 750 RFC 3779 resources from the certificate. If the global 751 resource_nounion flag is TRUE, it compares the extracted certificate 752 resources with the resources specified in the constraints file. If 753 the two resource sets are different, the algorithm SHOULD issue a 754 warning noting the difference. An output resource set is then formed 755 that is identical to the resource set extracted from the certificate. 756 If, however, the resource_nounion flag is FALSE, then the output 757 resource set is calculated by forming the union of the resources 758 extracted from the certificate and the resources specified for this 759 target block in the constraints file. A paracertificate is then 760 constructed according to Table 1, using fields from the original 761 certificate, the tags that had been set during stage 0, and, if 762 necessary, fields from the RP's TA certificate. The RFC 3779 763 resources of the paracertificate are equated to the derived output 764 resource set. The PARA state bit is set for the newly created 765 paracertificate. 767 4.2.3 Ancestor processing (stage 2) 769 The goal of ancestor processing is to discover all ancestors of 770 target certificates and remove from those ancestors the resources 771 specified in the target blocks corresponding to the targets being 772 processed. Note that it is possible that, for a given chain from a 773 target certificate to a trust anchor, another target might be 774 encountered. This is handled by removing all the target resources of 775 all descendents. The set of all targets that are descendants of the 776 given certificate is formed. The union of all the target resources of 777 the corresponding target blocks is computed, and this union in then 778 removed from the shared ancestor. 780 In detail, the algorithm is as follows. First, all original target 781 certificates processed during stage 1 processing are collected. 782 Second, any such certificates that have the NOCHAIN state bit set are 783 eliminated from the collection. (Note that, as a result of 784 eliminating such certificates, the resulting collection may be empty, 785 in which case this stage of algorithm processing terminates, and 786 processing advances to stage 3.) Next, an implementation MAY sort the 787 collection. The optional sorting algorithm is described in Appendix 788 B. Note that all stage 2 processing is completed before any stage 3 789 processing. 791 Two levels of nested iteration are performed. The outer iteration is 792 effected over all certificates in the collection; the inner iteration 793 is over all ancestors of the designated certificate being processed. 794 The first certificate in the collection is chosen, and a resource set 795 R is initialized based on the resources of the target block for that 796 certificate (since the certificate is in collection, it must be a 797 target certificate, and thus correspond to a target block). The 798 parent of the certificate is then located using ordinary path 799 discovery over original certificates only. The ancestor's certificate 800 resources A are then extracted. These resources are then perforated 801 with respect to R. That is, an output set of resources is created by 802 forming the intersection I of A and R, and then taking the set 803 difference A - I as the output resources. A paracertificate is then 804 created containing resources tat are these output resources, and 805 containing other fields and extensions from the original certificate 806 (and possibly the RP's TA certificate) according to the procedure 807 given in Table 1. The PARA state bit is set on this paracertificate 808 and the ORIGINAL state bit is set on A. If A is also a target 809 certificate, as indicated by its TARGET state bit being set, then 810 there will already have been a paracertificate created for it. This 811 previous paracertificate is destroyed in favor of the newly created 812 paracertificate. In this case also, the set R is augmented by adding 813 into it the set of resources of the target block for A. The algorithm 814 then proceeds to process the parent of A. This inner iteration 815 continues until the self-signed certificate at the root of the path 816 is encountered and processed. The outer iteration then continues by 817 clearing R and proceeding to the next certificate in the target 818 collection. 820 Note that ancestor processing has the potential for order dependency 821 as mentioned earlier in this document. If sorting is not implemented, 822 or if the sorting algorithm fails to completely process the 823 collection of target certificates because the allotted maximum number 824 of iterations has been realized, it may be the case that an ancestor 825 of a certificate logically occurs before that certificate in the 826 collection. Whenever an existing paracertificate is replaced by a 827 newly created paracertificate during ancestor processing, the 828 algorithm SHOULD alert the user, and SHOULD log sufficient detail 829 such that the user is able to determine which resources were 830 perforated from the original certificate in order to create the (new) 831 paracertificate. 833 In addition, implementations MUST provide for conflict detection and 834 notification during ancestor processing. In particular, if a 835 certificate is encountered two or more times during any part of the 836 ancestor processing algorithm, and the modifications dictated by the 837 ancestor processing algorithm are in conflict, the implementation 838 MUST refrain from processing that certificate. Further, the 839 implementation MUST present the user with an error message that 840 contains enough detail that the user can locate those directives in 841 the constraints file that are creating the conflict. For example, 842 during one stage of the processing algorithm it may be directed that 843 resources R1 be added to a certificate C, while during a different 844 stage of the processing algorithm it may be directed that resources 845 R2 be removed from certificate C. If the resource sets R1 and R2 have 846 a non-empty intersection, that is a conflict. 848 4.2.4 Tree processing (stage 3) 850 The goal of tree processing is to locate other certificates the 851 resources of which might conflict with the resources allocated to a 852 target by virtue of their being mentioned in the constraints file. In 853 this stage of processing, certificates that are not ancestors of any 854 target are considered. In detail, the algorithm used is as follows. 855 First, all target certificates are again collected. Second, all 856 target certificates that have the NOCHAIN state bit set are 857 eliminated from this collection. Third, if the intersection_always 858 global flag is set, those target blocks that occur in the constraints 859 file, but that did not correspond to a certificate in the local 860 repository, are also added to the collection. In tree processing, 861 unlike ancestor processing, this collection is not sorted. An 862 iteration is now performed over each certificate (or set of target 863 block resources) in the collection. Note that the collection may be 864 empty, in which case this stage of algorithm processing terminates, 865 and processing advances to stage 4. Note also that all stage 3 866 processing is performed before any stage 4 processing. 868 Given a certificate or target resource block, each top level original 869 TA certificate is examined. If that TA certificate has an 870 intersection with the target block resources, then the certificate is 871 perforated with respect to those resources. A paracertificate is 872 created based on the contents of the original certificate (and 873 possibly the RP's TA certificate, as indicated in Table 1) using the 874 perforated resources. The ORIGINAL state bit is set on the original 875 certificate processed in this manner, and the PARA state bit is set 876 on the paracertificate just created. An inner iteration then begins 877 on the descendants of the original certificate just processed. There 878 are two ways in which this iteration may proceed. If the treegrowth 879 global flag is clear, then examination of the children proceeds until 880 all children are exhausted, or until one child is found with 881 intersecting resources. If the treegrowth global flag is set, all 882 children are examined. Since a transfer of resources may be in 883 process such that more than one child possesses intersecting 884 resources, it is RECOMMENDED that the treegrowth flag be set. The 885 inner iteration proceeds until all descendants have been examined and 886 no further intersecting resources are found. The outer iteration then 887 continues with the next certificate or target resource block in the 888 collection. Note that unlike ancestor processing, there is no concept 889 of a potentially cumulating resource collection R; only the resources 890 in the target block are used for perforation. 892 4.2.5 TA re-parenting (stage 4) 894 In the final stage of TA algorithm processing, all TA certificates 895 (other than the RP's TA certificate) that have not already been 896 processed in a previous stage are now processed. It will be the case 897 that all such unprocessed TA certificates have no intersection with 898 any target resource blocks. As such, in creating the corresponding 899 paracertificates, the output resource set is identical to the input 900 resource set. Other transformations as described in Table 1 are 901 performed. The original TA certificates have the ORIGINAL state bit 902 set; the newly created paracertificates have the PARA state bit set. 903 Note that once stage four processing is completely, only a single TA 904 certificate will remain in an unprocessed state, namely the relying 905 party's own TA certificate. 907 4.3 Discussion 909 The algorithm described in this document effectively creates two 910 coexisting certificate hierarchies: the original certificate 911 hierarchy and the paracertificate hierarchy. Note that original 912 certificates are not removed during any of the processing described 913 in the previous section. Some original certificates may move from 914 having no state bits set (or only the NOCHAIN state bit set) to 915 having one or both of the ORIGINAL and TARGET state bits set. In 916 addition, the NOCHAIN state bit will still be set if it was set 917 before any processing. The paracertificate hierarchy, however, is 918 intended to supersede the original hierarchy for the purposes of ROA 919 validation. The presence of two hierarchies has implications for the 920 handling of path discovery, and also for the handling of revocation. 921 If one thinks of a certificate as being "named" by its SKI, then 922 there can now be two certificates with the same name, one an original 923 certificate and the other a paracertificate. The next two sections 924 discuss the implications of this duality in detail. Before 925 proceeding, it is worth noting that even without the existence of the 926 paracertificate hierarchy, cases may exist in which two or more 927 original certificates have the same SKI. As noted earlier, in Section 928 4.2.2, these cases may be subdivided into the case in which such 929 certificates are distinguishable by virtue of having different 930 subject names, but identical issuers and resource sets, versus all 931 other cases. In the distinguishable case, the path discovery 932 algorithm treats the original certificates as separate certificates, 933 and processes them separately. In all other cases, the original 934 certificates should be treated as indistinguishable, and path 935 validation should fail. 937 5 Implications for Path Discovery 939 Path discovery proceeds from a child certificate C by asking for a 940 parent certificate P such that the AKI of C is equal to the SKI of P. 941 With one hierarchy this question would produce at most one answer. 942 With two hierarchies, the original certificate hierarchy and the 943 paracertificate hierarchy, the question may produce two answers, one 944 answer, or no answer. Each of these cases is considered in turn. 946 5.1 Two answers 948 In this case, it SHOULD be the case that one of the matches is a 949 certificate with the ORIGINAL state bit set and the PARA state bit 950 clear, while the other match inversely has the ORIGINAL state bit 951 clear and the PARA state bit set. If any other combination of 952 ORIGINAL and PARA state bits obtains, the path discovery algorithm 953 MUST alert the user. In addition, the path discovery algorithm SHOULD 954 refrain from attempting to make a choice as to which of the two 955 certificates is the putative parent. In the no-error case, with the 956 state bits are as indicated, the certificate with the PARA state bit 957 set is chosen as the parent P. Note this means, in effect, that all 958 children of the original certificate have been re-parented under the 959 paracertificate. 961 5.2 One answer 963 If the matching certificate has neither the ORIGINAL state bit set 964 nor the PARA state bit set, this certificate is the parent. If the 965 matching certificate has the PARA state bit set but the ORIGINAL 966 state bit not set, this certificate is the parent. (This situation 967 would arise, for example, if the original certificate had been 968 revoked by its issuer but the paracertificate had not been revoked by 969 the RP.) If the matching certificate has the ORIGINAL state bit set 970 but the PARA state bit not set, this is not an error but it is a 971 situation in which path discovery MUST be forced to fail. The parent 972 P MUST be set to NULL, and the NOCHAIN state bit must be set on C and 973 all its descendants; the user SHOULD be warned. Even if the RP has 974 revoked the paracertificate, the original certificate MAY persist. 975 Forcing path discovery to unsuccessfully terminate is a reflection of 976 the RP's preference for path discovery to fail as opposed to using 977 the original hierarchy. Finally, if the matching certificate has both 978 the ORIGINAL and PARA state bits set, this is an error. The parent P 979 MUST be set to NULL, and the user MUST be warned. 981 5.3 No answer 983 This situation occurs when C has no parent in either the original 984 hierarchy or the paracertificate hierarchy. In this case the parent P 985 is NULL and path discovery terminates unsuccessfully. The NOCHAIN 986 state bit must be set on C and all its descendants. 988 6 Implications for Revocation 990 In a standard implementation of revocation in a PKI, a valid CRL 991 names a (sibling) certificate by serial number. That certificate is 992 revoked and is purged from the local RPKI repository. In the 993 mechanism described in this document, the original certificate 994 hierarchy and the paracertificate hierarchy are closely related. It 995 can thus be asked how revocation is handled in the presence of these 996 two hierarchies, in particular with regard to whether changes in one 997 of the hierarchies triggers corresponding changes in the other 998 hierarchy. There are four cases. 1000 6.1 No state bits set 1002 If the CRL names a certificate that has neither the ORIGINAL state 1003 bit set nor the PARA state bit set, revocation proceeds normally. All 1004 children of the revoked certificate have their state modified so that 1005 the NOCHAIN state bit is set. 1007 6.2 ORIGINAL state bit set 1009 If the CRL names a certificate with the ORIGINAL state bit set and 1010 the PARA state bit clear, then this certificate is revoked as usual. 1011 If this original certificate also has the TARGET state bit set, then 1012 the corresponding paracertificate (if it exists) is not revoked; if 1013 this original certificate has the TARGET state bit clear, then the 1014 corresponding paracertificate is revoked as well. Note that since all 1015 the children of the original certificate have been re-parented to be 1016 children of the corresponding paracertificate, as described above, 1017 the revocation algorithm MUST NOT set the NOCHAIN state bit on these 1018 children unless the paracertificate is also revoked. Note also that 1019 if the original certificate is revoked but the paracertificate is not 1020 revoked, the paracertificate retains its PARA state bit. This is to 1021 ensure that path discovery proceeds preferentially through the 1022 paracertificate hierarchy, as described above. 1024 6.3 PARA state bit set 1026 If the CRL names a certificate with the PARA state bit set and the 1027 ORIGINAL state bit clear, this CRL must have been issued, perforce, 1028 by the RP itself. This is because all the paracertificates are 1029 children of the RP's TA certificate. (Recall that a TA is not revoked 1030 via a CRL; it is merely removed from the repository.) The 1031 paracertificate is revoked and all children of the paracertificate 1032 have the NOCHAIN state bit set. No action is taken on the 1033 corresponding original certificate; in particular, its ORIGINAL state 1034 bit is not cleared. 1036 Note that the serial numbers of paracertificates are synthesized 1037 according to the procedure given in Table 1, rather than being 1038 assigned by an algorithm under the control of the (original) issuer. 1040 6.4 Both ORIGINAL and PARA state bits set 1042 This is an error. The revocation algorithm MUST alert the user and 1043 take no further action. 1045 7 Security Considerations 1047 The goal of the algorithm described in this document is to enable an 1048 RP to impose its own constraints on its view of the RPKI, which 1049 itself is a security function. An RP using a constraints file is 1050 trusting the assertions made in that file. Errors in the constraints 1051 file used by an RP can undermine the security offered by the RPKI, to 1052 that RP. In particular, since the paracertificate hierarchy is 1053 intended to trump the original certificate hierarchy for the purposes 1054 of path discovery, an improperly constructed paracertificate 1055 hierarchy could validate origin attestations that would otherwise be 1056 invalid, or could declare as invalid origin attestations that would 1057 otherwise be valid. As a result, an RP must carefully consider the 1058 security implications of the constraints file being used. 1060 8 IANA Considerations 1062 [Note to IANA, to be removed prior to publication: there are no IANA 1063 considerations stated in this version of the document.] 1065 9 Acknowledgements 1067 The authors would like to acknowledge the significant contributions 1068 of Charles Gardiner, who was the original author of an internal 1069 version of this document, and who contributed significantly to its 1070 evolution into the current version. 1072 10 References 1074 10.1 Normative References 1076 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1077 Requirement Levels", BCP 14, RFC 2119, March 1997. 1079 [RFC3513] Hinden, R., and S. Deering, "Internet Protocol Version 6 1080 (IPv6) Addressing Architecture", RFC 3513, April 2003. 1082 [RFC3779] Lynn, C., Kent, S., and K. Seo, "X.509 Extensions for IP 1083 Addresses and AS Identifiers", RFC 3779, June 2004. 1085 [RFC5280] Cooper, D., Santesson, S., Farrell, S., Boeyen, S., 1086 Housley, R., and W. Polk, "Internet X.509 Public Key 1087 Infrastructure Certificate and Certificate Revocation 1088 List (CRL) Profile", RFC 5280, May 2008. 1090 [RFC5396] Huston, G., and G. Michaelson, "Textual Representation of 1091 Autonomous System (AS) Numbers", RFC 5396, December 2008. 1093 [I-D. sidr-arch] 1094 Lepinski, M. and S. Kent, "An Infrastructure to Support 1095 Secure Internet Routing", draft-ietf-sidr-arch-11.txt 1096 (work in progress), September 2010. 1098 [I-D. sidr-repos-struct] 1099 Huston, G., Loomans, R., and G. Michaelson, "A Profile 1100 for Resource Certificate Policy Structure", draft-ietf- 1101 sidr-repos-struct-05.txt (work in progress), October 1102 2010. 1104 [I-D. sidr-res-cert-prof] 1105 Huston, G., Michaelson, G., and R. Loomans, "A Profile 1106 for X.509 PKIX Resource Certificates", draft-ietf-sidr- 1107 res-certs-19.txt (work in progress), October 2010. 1109 10.2 Informative References 1111 None. 1113 Authors' Addresses 1115 Stephen Kent 1116 BBN Technologies 1117 10 Moulton St. 1118 Cambridge, MA 02138 1120 Email: kent@bbn.com 1122 Mark Reynolds 1123 BBN Technologies 1124 10 Moulton St. 1125 Cambridge, MA 02138 1127 Email: mreynold@bbn.com 1129 Appendix A: Sample Constraints File 1131 ; 1132 ; Sample constraints file for TBO LTA Test Corporation. 1133 ; 1134 ; TBO manages its own local (10.x.x.x) address space 1135 ; via the target blocks in this file. 1136 ; 1138 ; 1139 ; Relying party subsection. TBO uses ssh-agent as 1140 ; a software cryptographic agent. 1141 ; 1143 PRIVATEKEYMETHOD OBO(ssh-agent) 1144 TOPLEVELCERTIFICATE tbomaster.cer 1146 ; 1147 ; Flags subsection 1148 ; 1149 ; Always use the resources in this file to augment 1150 ; certificate resources. 1151 ; Always process resource conflicts in the tree, even 1152 ; if the target certificate is missing. 1153 ; Always search the entire tree. 1154 ; 1156 CONTROL resource_nounion FALSE 1157 CONTROL intersection_always TRUE 1158 CONTROL treegrowth TRUE 1160 ; 1161 ; Tags subsection 1162 ; 1163 ; Copy the original cert's validity dates. 1164 ; Use the default policy OID. 1165 ; Use our own CRLDP. 1166 ; Use our own AIA. 1167 ; 1169 TAG Xvalidity_dates C 1170 TAG Xcp D 1171 TAG Xcrldp rsync://tbo_lta_test.com/pub/CRLs 1172 TAG Xaia rsync://tbo_lta_test.com/pub/repos 1174 ; 1175 ; Block subsection 1176 ; 1178 ; 1179 ; First block: TBO corporate 1180 ; 1182 SKI 00112233445566778899998877665544332211 1183 IPv4 1184 10.2.3/24 1185 10.8/16 1186 IPv6 1187 2000:2:3:4:5:6/112 1188 AS# 1189 60123 1190 5507 1192 ; 1193 ; Second block: TBO LTA Test enforcement division 1194 ; 1196 SKI 653420AF758421CF600029FF857422AA6833299F 1197 IPv4 1198 10.2.8/24 1199 10.47/16 1200 IPv6 1201 AS# 1202 60124 1204 ; 1205 ; Third block: TBO LTA Test Acceptance Corporation 1206 ; Quality financial services since sometime 1207 ; late yesterday. 1208 ; 1210 SKI 19:82:34:90:8b:a0:9c:ef:00:af:a0:98:23:09:82:4b:ef:ab:98:09 1211 IPv4 1212 10.3.3/24 1213 IPv6 1214 AS# 1215 60125 1217 ; End of TBO constraints file 1219 Appendix B: Optional Sorting Algorithm for Ancestor Processing 1221 Sorting is performed in an effort to eliminate any order dependencies 1222 in ancestor processing, as described in section 4.2.3 of this 1223 document. The sorting algorithm does this by rearranging the 1224 processing of certificates such that if A is an ancestor of B, B is 1225 processed before A. The sorting algorithm is an OPTIONAL part of 1226 ancestor processing. Sorting proceeds as follows. The collection 1227 created at the beginning of ancestor processing is traversed and any 1228 certificate in the collection that is visited as a result of path 1229 discovery is temporarily marked. After the traversal, all unmarked 1230 certificates are moved to the beginning of the collection. The 1231 remaining marked certificates are unmarked, and a traversal again 1232 performed through this sub-collection of previously marked 1233 certificates. The sorting algorithm proceeds iteratively until all 1234 certificates have been sorted or until a predetermined fixed number 1235 of iterations has been performed. (Eight is suggested as a munificent 1236 value for the upper bound, since the number of sorting steps need 1237 should be no greater than the maximum depth of the tree). Finally, 1238 the ancestor processing algorithm is applied in turn to each 1239 certificate in the remaining sorted collection. If the sorting 1240 algorithm fails to converge, that is if the maximum number of 1241 iterations has been reached and unsorted certificates remain, the 1242 implementation SHOULD warn the user.