idnits 2.17.1 draft-ietf-dnsop-nsec-aggressiveuse-02.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- -- The document has examples using IPv4 documentation addresses according to RFC6890, but does not use any IPv6 documentation addresses. Maybe there should be IPv6 examples, too? Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year (Using the creation date from RFC4035, updated by this document, for RFC5378 checks: 2002-10-29) -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (September 13, 2016) is 2782 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) ** Downref: Normative reference to an Historic RFC: RFC 5074 ** Obsolete normative reference: RFC 7719 (Obsoleted by RFC 8499) == Outdated reference: A later version (-05) exists of draft-ietf-dnsop-nxdomain-cut-03 Summary: 2 errors (**), 0 flaws (~~), 2 warnings (==), 3 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group K. Fujiwara 3 Internet-Draft JPRS 4 Updates: 4035 (if approved) A. Kato 5 Intended status: Standards Track Keio/WIDE 6 Expires: March 17, 2017 W. Kumari 7 Google 8 September 13, 2016 10 Aggressive use of NSEC/NSEC3 11 draft-ietf-dnsop-nsec-aggressiveuse-02 13 Abstract 15 The DNS relies upon caching to scale; however, the cache lookup 16 generally requires an exact match. This document specifies the use 17 of NSEC/NSEC3 resource records to generate negative answers within a 18 range. This increases performance / decreases latency, decreases 19 resource utilization on both authoritative and recursive servers, and 20 also increases privacy. It may also help increase resilience to 21 certain DoS attacks in some circumstances. 23 This document updates RFC4035 by allowing resolvers to generate 24 negative answers based upon NSEC/NSEC3 records. 26 [ Ed note: Text inside square brackets ([]) is additional background 27 information, answers to frequently asked questions, general musings, 28 etc. They will be removed before publication.This document is being 29 collaborated on in Github at: https://github.com/wkumari/draft-ietf- 30 dnsop-nsec-aggressiveuse. The most recent version of the document, 31 open issues, etc should all be available here. The authors 32 (gratefully) accept pull requests. 34 Known / open issues [To be moved to Github issue tracker]: 36 1. We say things like: "Currently the DNS does ..." - this will not 37 be true after this is deployed, but I'm having a hard time 38 rewording this. "Without the techniques described in this 39 document..." seems klunky. Perhaps "historically?!" 41 ] 43 Status of This Memo 45 This Internet-Draft is submitted in full conformance with the 46 provisions of BCP 78 and BCP 79. 48 Internet-Drafts are working documents of the Internet Engineering 49 Task Force (IETF). Note that other groups may also distribute 50 working documents as Internet-Drafts. The list of current Internet- 51 Drafts is at http://datatracker.ietf.org/drafts/current/. 53 Internet-Drafts are draft documents valid for a maximum of six months 54 and may be updated, replaced, or obsoleted by other documents at any 55 time. It is inappropriate to use Internet-Drafts as reference 56 material or to cite them other than as "work in progress." 58 This Internet-Draft will expire on March 17, 2017. 60 Copyright Notice 62 Copyright (c) 2016 IETF Trust and the persons identified as the 63 document authors. All rights reserved. 65 This document is subject to BCP 78 and the IETF Trust's Legal 66 Provisions Relating to IETF Documents 67 (http://trustee.ietf.org/license-info) in effect on the date of 68 publication of this document. Please review these documents 69 carefully, as they describe your rights and restrictions with respect 70 to this document. Code Components extracted from this document must 71 include Simplified BSD License text as described in Section 4.e of 72 the Trust Legal Provisions and are provided without warranty as 73 described in the Simplified BSD License. 75 Table of Contents 77 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 78 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3 79 3. Problem Statement . . . . . . . . . . . . . . . . . . . . . . 4 80 4. Background . . . . . . . . . . . . . . . . . . . . . . . . . 4 81 5. Proposed Solution . . . . . . . . . . . . . . . . . . . . . . 5 82 5.1. Aggressive Negative Caching . . . . . . . . . . . . . . . 5 83 5.2. NSEC . . . . . . . . . . . . . . . . . . . . . . . . . . 6 84 5.3. NSEC3 . . . . . . . . . . . . . . . . . . . . . . . . . . 6 85 5.4. Wildcard . . . . . . . . . . . . . . . . . . . . . . . . 6 86 5.5. Consideration on TTL . . . . . . . . . . . . . . . . . . 7 87 6. Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . 7 88 7. Update to RFC 4035 . . . . . . . . . . . . . . . . . . . . . 8 89 8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 9 90 9. Security Considerations . . . . . . . . . . . . . . . . . . . 9 91 10. Implementation Status . . . . . . . . . . . . . . . . . . . . 9 92 11. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 9 93 12. Change History . . . . . . . . . . . . . . . . . . . . . . . 9 94 12.1. Version draft-fujiwara-dnsop-nsec-aggressiveuse-01 . . . 11 95 12.2. Version draft-fujiwara-dnsop-nsec-aggressiveuse-02 . . . 11 96 12.3. Version draft-fujiwara-dnsop-nsec-aggressiveuse-03 . . . 11 97 13. References . . . . . . . . . . . . . . . . . . . . . . . . . 11 98 13.1. Normative References . . . . . . . . . . . . . . . . . . 11 99 13.2. Informative References . . . . . . . . . . . . . . . . . 12 100 Appendix A. Detailed implementation notes . . . . . . . . . . . 12 101 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 13 103 1. Introduction 105 A DNS negative cache currently exists, and is used to cache the fact 106 that a name does not exist. This method of negative caching requires 107 exact matching; this leads to unnecessary additional lookups, 108 increases latency, leads to extra resource utilization on both 109 authoritative and recursive servers, and decreases privacy by leaking 110 queries. 112 This document updates RFC 4035 to allow recursive resolvers to use 113 NSEC/NSEC3 resource records to aggressively cache negative answers. 114 This would allow such resolvers to respond with NXDOMAIN immediately 115 if the name in question falls into a range expressed by a NSEC/NSEC3 116 resource record already in the cache. 118 Aggressive Negative Caching was first proposed in Section 6 of DNSSEC 119 Lookaside Validation (DLV) [RFC5074] in order to find covering NSEC 120 records efficiently. 122 Section 3 of [I-D.vixie-dnsext-resimprove] "Stopping Downward Cache 123 Search on NXDOMAIN" and [I-D.ietf-dnsop-nxdomain-cut] proposed 124 another approach to use NXDOMAIN information effectively. 126 2. Terminology 128 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 129 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 130 document are to be interpreted as described in RFC 2119 [RFC2119]. 132 Many of the specialized terms used in this document are defined in 133 DNS Terminology [RFC7719]. In this document we are using the terms 134 "recursive resolver" or "recursive server" as a more readable 135 alternative to the more formal[RFC7719] "full-service resolver" 137 The key words "Closest Encloser" and "Source of Synthesis" in this 138 document are to be interpreted as described in[RFC4592]. 140 "Closest Encloser" is also defined in NSEC3 [RFC5155], as is "Next 141 closer name". 143 3. Problem Statement 145 The current DNS negative cache caches negative (non-existent) 146 information, and requires an exact match in most instances [RFC2308]. 148 Assume that the (DNSSEC signed) "example.com" zone contains: 150 apple.example.com IN A 192.0.2.1 152 elephant.example.com IN A 192.0.2.2 154 zebra.example.com IN A 192.0.2.3 156 If a recursive resolver gets a query for cat.example.com, it will 157 query the example.com authoritative servers and will get back an NSEC 158 (or NSEC3) record starting that there are no records between apple 159 and elephant. The recursive resolver then knows that cat.example.com 160 does not exist; however, it (currently) does not use the fact that 161 the proof covers a range (apple to elephant) to suppress queries for 162 other labels that fall within this range. This means that if the 163 recursive resolvers gets a query for ball.example.com (or 164 dog.example.com) it will once again go off and query the example.com 165 servers for these names. 167 Apart from wasting bandwidth, this also wastes resources on the 168 recursive server (it needs to keep state for outstanding queries), 169 wastes resources on the authoritative server (it has to answer 170 additional questions), increases latency (the end user has to wait 171 longer than necessary to get back an NXDOMAIN answer), can be used by 172 attackers to cause a DoS (see additional resources), and also has 173 privacy implications (e.g: typos leak out further than necessary). 175 4. Background 177 DNSSEC [RFC4035] and [RFC5155] both provide "authenticated denial of 178 existence"; this is a cryptographic proof that the queried for name 179 does not exist, accomplished by providing a (DNSSEC secured) record 180 containing the names which appear alphabetically before and after the 181 queried for name. In the example above, if the (DNSSEC validating) 182 recursive server were to query for lion.example.com it would receive 183 a (signed) NSEC/NSEC3 record stating that there are no labels between 184 "elephant" and "zebra". This is a signed, cryptographic proof that 185 these names are the ones before and after the queried for label. As 186 lion.example.com falls within this range, the recursive server knows 187 that lion.example.com really does not exist. This document specifies 188 that this NSEC/NSEC3 record should be used to generate negative 189 answers for any queries that the recursive server receives that fall 190 within the range covered by the record (for the TTL for the record). 192 [RFC4035]; Section 4.5 states: 194 For a zone signed with NSEC, it would be possible to use the 195 information carried in NSEC resource records to indicate the non- 196 existence of a range of names. However, such use is discouraged by 197 Section 4.5 of RFC4035. It is recommended that readers read RFC4035 198 in its entirety for a better understanding. At the root of the 199 concern is that new records could have been added to the zone during 200 the TTL of the NSEC record, and that generating negative responses 201 from the NSEC record would hide these. We believe this 202 recommendation can be relaxed because lookups for the specific name 203 could have come in during the normal negative cache time and so 204 operators should have no expectation that an added name would work 205 immediately. We think that the TTL of the NSEC record is the 206 authoritative statement of how quickly a name can start working 207 within a zone. 209 5. Proposed Solution 211 5.1. Aggressive Negative Caching 213 Section 4.5 of [RFC4035] shows that "In theory, a resolver could use 214 wildcards or NSEC RRs to generate positive and negative responses 215 (respectively) until the TTL or signatures on the records in question 216 expire. However, it seems prudent for resolvers to avoid blocking 217 new authoritative data or synthesizing new data on their own. 218 Resolvers that follow this recommendation will have a more consistent 219 view of the namespace". 221 This document relaxes this this restriction, as follows: 223 +--------------------------------------------------------------+ 224 | Once the records are validated, DNSSEC enabled validating | 225 | resolvers MAY use NSEC/NSEC3 resource records to generate | 226 | negative responses until their effective TTLs or signatures | 227 | for those records expire. | 228 +--------------------------------------------------------------+ 230 If the validating resolver's cache has sufficient information to 231 validate the query, the resolver SHOULD use NSEC/NSEC3/wildcard 232 records aggressively. Otherwise, it MUST fall back to send the query 233 to the authoritative DNS servers. 235 If the query name has the matching NSEC/NSEC3 RR proving the 236 information requested does not exist, the resolver may respond with a 237 NODATA (empty) answer. 239 5.2. NSEC 241 Implementations SHOULD enable aggressive use of NSEC by default. 242 Implementations SHOULD provide a configuration switch to disable 243 aggressive use of NSEC and allow it to be enabled or disabled per 244 domain. 246 The validating resolver needs to check the existence of an NSEC RR 247 matching/covering the source of synthesis and an NSEC RR covering the 248 query name. 250 If the validating resolver's cache contains an NSEC RR covering the 251 source of synthesis and the covering NSEC RR of the query name, the 252 resolver may respond with NXDOMAIN error immediately. 254 5.3. NSEC3 256 NSEC3 aggressive negative caching is more difficult. If the zone is 257 signed with NSEC3, the validating resolver needs to check the 258 existence of non-terminals and wildcards which derive from query 259 names. 261 If the validating resolver's cache contains an NSEC3 RR matching the 262 closest encloser, an NSEC3 RR covering the next closer name, and an 263 NSEC3 RR covering the source of synthesis, it is possible for the 264 resolver to respond with NXDOMAIN immediately. 266 If a covering NSEC3 RR has Opt-Out flag, the covering NSEC3 RR does 267 not prove the non-existence of the domain name and the aggressive 268 negative caching is not possible for the domain name. 270 A validating resolver implementation MAY support aggressive use of 271 NSEC3. If it does aggressive use of NSEC3, it SHOULD provide a 272 configuration switch to disable aggressive use of NSEC3 and allow it 273 to be enabled or disabled for specific zones. 275 5.4. Wildcard 277 The last paragraph of RFC 4035 Section 4.5 discusses aggressive use 278 of a cached deduced wildcard (as well as aggressive use of NSEC) and 279 recommends that it is not relied upon. 281 Just like the case for the aggressive use of NSEC discussed in this 282 draft, we revise this recommendation. As long as the resolver knows 283 a name would not exist without the wildcard match, it can answer a 284 query for that name using the cached deduced wildcard, and it may be 285 justified for performance and other benefits. 287 Such aggressive use of cached deduced wildcard can be employed 288 independently from aggressive use of NSEC. But, it will be more 289 effective when both are enabled since the resolver can determine the 290 name subject to wildcard would not otherwise exist more efficiently. 292 Furthermore, when aggressive use of NSEC is enabled, the aggressive 293 use of cached deduced wildcard will be more effective. 295 An implementation MAY support aggressive use of wildcards. It SHOULD 296 provide a configuration switch to disable aggressive use of 297 wildcards. 299 5.5. Consideration on TTL 301 The TTL value of negative information is especially important, 302 because newly added domain names cannot be used while the negative 303 information is effective. Section 5 of RFC 2308 states that the 304 maximum number of negative cache TTL value is 3 hours (10800). It is 305 RECOMMENDED that resolvers limit the maximum effective TTL value of 306 negative responses (NSEC/NSEC3 RRs) to this same value. 308 6. Benefits 310 The techniques described in this document provide a number of 311 benefits, including (in no specific order): 313 Latency By answering directly from cache, recursive resolvers can 314 immediately inform clients that the name they are looking for does 315 not exist, improving the user experience. 317 Decreased recursive server load By answering negative queries from 318 the cache, recursive servers avoid having send a query and wait 319 for a response. In addition to decreasing the bandwidth used, it 320 also means that the server does not need to allocate and maintain 321 state, thereby decreasing memory and CPU load. 323 Decreased authorative server load Because recursive servers can 324 answer (negative) queries without asking the authoritative server, 325 the authoritative servers receive less queries. This decreases 326 the authoritative server bandwidth, queries per second and CPU 327 utilization. 329 The scale of the benefit depends upon multiple factors, including the 330 query distribution. For example, currently around 65% of queries to 331 Root Name servers result in NXDOMAIN responses; this technique will 332 eliminate a sizable quantity of these. 334 [ Editor note: There has been some discussion on if this document 335 should discuss this attack and mitigation. The authors think that 336 this is useful / important, but some participants feel that it 337 oversells the DoS mitigation benefit. Please let us know if the 338 below is helpful. Also, the below description is not as clear as it 339 could be - it's been tricky to balance readability, correctness and 340 conciseness. Text gratefully accepted... ] 342 The technique described in this document may also mitigate so-called 343 "random QNAME attacks", in which attackers send many queries for 344 random sub-domains to recursive resolvers. As the recursive server 345 will not have the answers cached it has to ask the authoritative 346 servers for each random query, leading to a DoS on the authoritative 347 (and often recursive) servers. Aggressive NSEC may help mitigate 348 these attacks by allowing the recursive to answer directly from cache 349 for any random queries which fall within already requested ranges. 350 The effectiveness of this depends upon a number of factors, including 351 if the attacker is making his queries through recursive resolvers 352 (e.g to hide his source), the number of entries in the zone, the TTL, 353 if the zone is using NSEC, if the attacker is setting the CD bit, 354 etc. In the ideal case, authoritative servers under attack will need 355 to answer somewhere between number_of_entries_in_zone queries and 2 * 356 number_of_entries_in_zone queries from each recursive server. This 357 is because there are as many "holes" between labels as there are 358 labels in a zone. If the random query falls in range for which 359 recursive server does not have an NSEC record cached, it will send a 360 query to the authoritative server, and so it will send approximately 361 the same number of queries as there are "holes" between entries. If 362 the random queries happen to be for names which exist in the zone, 363 the recursive will send those as well. 365 7. Update to RFC 4035 367 Section 4.5 of [RFC4035] shows that "In theory, a resolver could use 368 wildcards or NSEC RRs to generate positive and negative responses 369 (respectively) until the TTL or signatures on the records in question 370 expire. However, it seems prudent for resolvers to avoid blocking 371 new authoritative data or synthesizing new data on their own. 372 Resolvers that follow this recommendation will have a more consistent 373 view of the namespace". 375 The paragraph is updated as follows: 377 +--------------------------------------------------------------+ 378 | Once the records are validated, DNSSEC enabled recursive | 379 | resolvers MAY use wildcards and NSEC/NSEC3 resource records | 380 | to generate (positive and) negative responses until their | 381 | effective TTLs or signatures for those records expire. | 382 +--------------------------------------------------------------+ 384 8. IANA Considerations 386 This document has no IANA actions. 388 9. Security Considerations 390 Newly registered resource records may not be used immediately. 391 However, choosing suitable TTL value and negative cache TTL value 392 (SOA MINIMUM field) will mitigate the delay concern, and it is not a 393 security problem. 395 It is also suggested to limit the maximum TTL value of NSEC / NSEC3 396 resource records in the negative cache to, for example, 10800 seconds 397 (3hrs), to mitigate this issue. Implementations which comply with 398 this proposal are recommended to have a configurable maximum value of 399 NSEC RRs in the negative cache. 401 Aggressive use of NSEC / NSEC3 resource records without DNSSEC 402 validation may cause security problems. It is highly recommended to 403 apply DNSSEC validation. 405 10. Implementation Status 407 Unbound supports aggressive negative caching. 409 11. Acknowledgments 411 The authors gratefully acknowledge DLV [RFC5074] author Samuel Weiler 412 and the Unbound developers. 414 The authors would like to specifically thank Tatuya JINMEI for 415 extensive review and comments, and also Mark Andrews, Stephane 416 Bortzmeyer, Casey Deccio, Alexander Dupuy, Olafur Gudmundsson, Bob 417 Harold, Shumon Huque, Pieter Lexis and Matthijs Mekking. 419 12. Change History 421 RFC Editor: Please remove this section prior to publication. 423 -01 to -02: 425 o Added Section 6 - Benefits (as suggested by Jinmei). 427 o Removed Appendix B (Jinmei) 429 o Replaced "full-service" with "validating" (where applicable) 431 o Integrated other comments from Jinmei from https://www.ietf.org/ 432 mail-archive/web/dnsop/current/msg17875.html 434 o Integrated comment from co-authors, including re-adding parts of 435 Appendix B, terminology, typos. 437 o Tried to explain under what conditions this may actually mitigate 438 attacks. 440 -00 to -01: 442 o Comments from DNSOP meeting in Berlin. 444 o Changed intended status to Standards Track (updates RFC 4035) 446 o Added a section "Updates to RFC 4035" 448 o Some language clarification / typo / cleanup 450 o Cleaned up the TTL section a bit. 452 o Removed Effects section, Additional proposal section, and pseudo 453 code. 455 o Moved "mitigation of random subdomain attacks" to Appendix. 457 From draft-fujiwara-dnsop-nsec-aggressiveuse-03 -> draft-ietf-dnsop- 458 nsec-aggressiveuse 460 o Document adopted by DNSOP WG. 462 o Adoption comments 464 o Changed main purpose to performance 466 o Use NSEC3/Wildcard keywords 468 o Improved wordings (from good comments) 470 o Simplified pseudo code for NSEC3 472 o Added Warren as co-author. 474 o Reworded much of the problem statement 476 o Reworked examples to better explain the problem / solution. 478 12.1. Version draft-fujiwara-dnsop-nsec-aggressiveuse-01 480 o Added reference to DLV [RFC5074] and imported some sentences. 482 o Added Aggressive Negative Caching Flag idea. 484 o Added detailed algorithms. 486 12.2. Version draft-fujiwara-dnsop-nsec-aggressiveuse-02 488 o Added reference to [I-D.vixie-dnsext-resimprove] 490 o Added considerations for the CD bit 492 o Updated detailed algorithms. 494 o Moved Aggressive Negative Caching Flag idea into Additional 495 Proposals 497 12.3. Version draft-fujiwara-dnsop-nsec-aggressiveuse-03 499 o Added "Partial implementation" 501 o Section 4,5,6 reorganized for better representation 503 o Added NODATA answer in Section 4 505 o Trivial updates 507 o Updated pseudo code 509 13. References 511 13.1. Normative References 513 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 514 Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/ 515 RFC2119, March 1997, 516 . 518 [RFC2308] Andrews, M., "Negative Caching of DNS Queries (DNS 519 NCACHE)", RFC 2308, DOI 10.17487/RFC2308, March 1998, 520 . 522 [RFC4035] Arends, R., Austein, R., Larson, M., Massey, D., and S. 523 Rose, "Protocol Modifications for the DNS Security 524 Extensions", RFC 4035, DOI 10.17487/RFC4035, March 2005, 525 . 527 [RFC4592] Lewis, E., "The Role of Wildcards in the Domain Name 528 System", RFC 4592, DOI 10.17487/RFC4592, July 2006, 529 . 531 [RFC5074] Weiler, S., "DNSSEC Lookaside Validation (DLV)", RFC 5074, 532 DOI 10.17487/RFC5074, November 2007, 533 . 535 [RFC5155] Laurie, B., Sisson, G., Arends, R., and D. Blacka, "DNS 536 Security (DNSSEC) Hashed Authenticated Denial of 537 Existence", RFC 5155, DOI 10.17487/RFC5155, March 2008, 538 . 540 [RFC7719] Hoffman, P., Sullivan, A., and K. Fujiwara, "DNS 541 Terminology", RFC 7719, DOI 10.17487/RFC7719, December 542 2015, . 544 13.2. Informative References 546 [I-D.ietf-dnsop-nxdomain-cut] 547 Bortzmeyer, S. and S. Huque, "NXDOMAIN really means there 548 is nothing underneath", draft-ietf-dnsop-nxdomain-cut-03 549 (work in progress), May 2016. 551 [I-D.vixie-dnsext-resimprove] 552 Vixie, P., Joffe, R., and F. Neves, "Improvements to DNS 553 Resolvers for Resiliency, Robustness, and Responsiveness", 554 draft-vixie-dnsext-resimprove-00 (work in progress), June 555 2010. 557 Appendix A. Detailed implementation notes 559 o Previously, cached negative responses were indexed by QNAME, 560 QCLASS, QTYPE, and the setting of the CD bit (see RFC 4035, 561 Section 4.7), and only queries matching the index key would be 562 answered from the cache. With aggressive negative caching, the 563 validator, in addition to checking to see if the answer is in its 564 cache before sending a query, checks to see whether any cached and 565 validated NSEC record denies the existence of the sought 566 record(s). Using aggressive negative caching, a validator will 567 not make queries for any name covered by a cached and validated 568 NSEC record. Furthermore, a validator answering queries from 569 clients will synthesize a negative answer whenever it has an 570 applicable validated NSEC in its cache unless the CD bit was set 571 on the incoming query. (Imported from Section 6 of [RFC5074]). 573 o Implementing aggressive negative caching suggests that a validator 574 will need to build an ordered data structure of NSEC and NSEC3 575 records for each signer domain name of NSEC / NSEC3 records in 576 order to efficiently find covering NSEC / NSEC3 records. Call the 577 table as NSEC_TABLE. (Imported from Section 6.1 of [RFC5074] and 578 expanded.) 580 o The aggressive negative caching may be inserted at the cache 581 lookup part of the recursive resolvers. 583 o If errors happen in aggressive negative caching algorithm, 584 resolvers MUST fall back to resolve the query as usual. "Resolve 585 the query as usual" means that the resolver must process the query 586 as though it does not implement aggressive negative caching. 588 Authors' Addresses 590 Kazunori Fujiwara 591 Japan Registry Services Co., Ltd. 592 Chiyoda First Bldg. East 13F, 3-8-1 Nishi-Kanda 593 Chiyoda-ku, Tokyo 101-0065 594 Japan 596 Phone: +81 3 5215 8451 597 Email: fujiwara@jprs.co.jp 599 Akira Kato 600 Keio University/WIDE Project 601 Graduate School of Media Design, 4-1-1 Hiyoshi 602 Kohoku, Yokohama 223-8526 603 Japan 605 Phone: +81 45 564 2490 606 Email: kato@wide.ad.jp 608 Warren Kumari 609 Google 610 1600 Amphitheatre Parkway 611 Mountain View, CA 94043 612 US 614 Email: warren@kumari.net