idnits 2.17.1 draft-andrews-dns-no-response-issue-15.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (November 11, 2015) is 3088 days in the past. Is this intentional? Checking references for intended status: Best Current Practice ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) ** Obsolete normative reference: RFC 5966 (Obsoleted by RFC 7766) Summary: 1 error (**), 0 flaws (~~), 1 warning (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group M. Andrews 3 Internet-Draft ISC 4 Intended status: Best Current Practice November 11, 2015 5 Expires: May 14, 2016 7 A Common Operational Problem in DNS Servers - Failure To Respond. 8 draft-andrews-dns-no-response-issue-15 10 Abstract 12 The DNS is a query / response protocol. Failure to respond or to 13 respond correctly to queries causes both immediate operational 14 problems and long term problems with protocol development. 16 This document identifies a number of common classes of queries that 17 some servers fail to respond too or respond incorrectly to. This 18 document also suggests procedures for TLD and other similar zone 19 operators to apply to help reduce / eliminate the problem. 21 The document does not look at the DNS data itself, just the structure 22 of the responses. 24 Status of This Memo 26 This Internet-Draft is submitted in full conformance with the 27 provisions of BCP 78 and BCP 79. 29 Internet-Drafts are working documents of the Internet Engineering 30 Task Force (IETF). Note that other groups may also distribute 31 working documents as Internet-Drafts. The list of current Internet- 32 Drafts is at http://datatracker.ietf.org/drafts/current/. 34 Internet-Drafts are draft documents valid for a maximum of six months 35 and may be updated, replaced, or obsoleted by other documents at any 36 time. It is inappropriate to use Internet-Drafts as reference 37 material or to cite them other than as "work in progress." 39 This Internet-Draft will expire on May 14, 2016. 41 Copyright Notice 43 Copyright (c) 2015 IETF Trust and the persons identified as the 44 document authors. All rights reserved. 46 This document is subject to BCP 78 and the IETF Trust's Legal 47 Provisions Relating to IETF Documents 48 (http://trustee.ietf.org/license-info) in effect on the date of 49 publication of this document. Please review these documents 50 carefully, as they describe your rights and restrictions with respect 51 to this document. Code Components extracted from this document must 52 include Simplified BSD License text as described in Section 4.e of 53 the Trust Legal Provisions and are provided without warranty as 54 described in the Simplified BSD License. 56 Table of Contents 58 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 59 2. Common queries class that result in non responses. . . . . . 3 60 2.1. EDNS Queries - Version Independent . . . . . . . . . . . 3 61 2.2. EDNS Queries - Version Specific . . . . . . . . . . . . . 4 62 2.3. EDNS Options . . . . . . . . . . . . . . . . . . . . . . 4 63 2.4. EDNS Flags . . . . . . . . . . . . . . . . . . . . . . . 4 64 2.5. DNS Flags . . . . . . . . . . . . . . . . . . . . . . . . 4 65 2.6. Unknown / Unsupported Type Queries . . . . . . . . . . . 5 66 2.7. Unknown DNS opcodes . . . . . . . . . . . . . . . . . . . 5 67 2.8. TCP Queries . . . . . . . . . . . . . . . . . . . . . . . 5 68 3. Remediating . . . . . . . . . . . . . . . . . . . . . . . . . 5 69 4. Firewalls and Load Balancers . . . . . . . . . . . . . . . . 7 70 5. Scrubbing Services . . . . . . . . . . . . . . . . . . . . . 8 71 6. Whole Answer Caches . . . . . . . . . . . . . . . . . . . . . 8 72 7. Response Code Selection . . . . . . . . . . . . . . . . . . . 9 73 8. Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 74 9. Security Considerations . . . . . . . . . . . . . . . . . . . 14 75 10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 14 76 11. Normative References . . . . . . . . . . . . . . . . . . . . 14 77 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 15 79 1. Introduction 81 The DNS [RFC1034], [RFC1035] is a query / response protocol. Failure 82 to respond to queries or to respond incorrectly causes both immediate 83 operational problems and long term problems with protocol 84 development. 86 Failure to respond to a query is indistinguishable from a packet loss 87 without doing a analysis of query response patterns and results in 88 unnecessary additional queries being made by DNS clients and 89 unnecessary delays being introduced to the resolution process. 91 Due to the inability to distinguish between packet loss and 92 nameservers dropping EDNS [RFC6891] queries, packet loss is sometimes 93 misclassified as lack of EDNS support which can lead to DNSSEC 94 validation failures. 96 Allowing servers which fail to respond to queries to remain results 97 in developers being afraid to deploy implementations of recent 98 standards. Such servers need to be identified and corrected / 99 replaced. 101 The DNS has response codes that cover almost any conceivable query 102 response. A nameserver should be able to respond to any conceivable 103 query using them. 105 Unless a nameserver is under attack, it should respond to all queries 106 directed to it as a result of following delegations. Additionally 107 code should not assume that there isn't a delegation to the server 108 even if it is not configured to serve the zone. Broken delegations 109 are a common occurrence in the DNS and receiving queries for zones 110 that you are not configured for is not a necessarily a indication 111 that you are under attack. Parent zone operators are supposed to 112 regularly check that the delegating NS records are consistent with 113 those of the delegated zone and to correct them when they are not 114 [RFC1034]. If this was being done regularly, the instances of broken 115 delegations would be much lower. 117 When a nameserver is under attack it may wish to drop packets. A 118 common attack is to use a nameserver as a amplifier by sending 119 spoofed packets. This is done because response packets are bigger 120 than the queries and big amplification factors are available 121 especially if EDNS is supported. Limiting the rate of responses is 122 reasonable when this is occurring and the client should retry. This 123 however only works if legitimate clients are not being forced to 124 guess whether EDNS queries are accept or not. While there is still a 125 pool of servers that don't respond to EDNS requests, clients have no 126 way to know if the lack of response is due to packet loss, EDNS 127 packets not being supported or rate limiting due to the server being 128 under attack. Mis-classifications of server characteristics are 129 unavoidable when rate limiting is done. 131 2. Common queries class that result in non responses. 133 There are three common query classes that result in non responses 134 today. These are EDNS queries, queries for unknown (unallocated) or 135 unsupported types, and filtering of TCP queries. 137 2.1. EDNS Queries - Version Independent 139 Identifying servers that fail to respond to EDNS queries can be done 140 by first identifying that the server responds to regular DNS queries, 141 followed by a series of otherwise identical responses using EDNS, 142 then making the original query again. A series of EDNS queries is 143 needed as at least one DNS implementation responds to the first EDNS 144 query with FORMERR but fails to respond to subsequent queries from 145 the same address for a period until a regular DNS query is made. The 146 EDNS query should specify a UDP buffer size of 512 bytes to avoid 147 false classification of not supporting EDNS due to response packet 148 size. 150 If the server responds to the first and last queries but fails to 151 respond to most or all of the EDNS queries, it is probably faulty. 152 The test should be repeated a number of times to eliminate the 153 likelihood of a false positive due to packet loss. 155 Firewalls may also block larger EDNS responses but there is no easy 156 way to check authoritative servers to see if the firewall is 157 misconfigured. 159 2.2. EDNS Queries - Version Specific 161 Some servers respond correctly to EDNS version 0 queries but fail to 162 respond to EDNS queries with version numbers that are higher than 163 zero. Servers should respond with BADVERS to EDNS queries with 164 version numbers that they do not support. 166 Some servers respond correctly to EDNS version 0 queries but fail to 167 set QR=1 when responding to EDNS versions they do not support. Such 168 answers are discarded or treated as requests. 170 2.3. EDNS Options 172 Some servers fail to respond to EDNS queries with EDNS options set. 173 Unknown EDNS options are supposed to be ignored by the server 174 [RFC6891]. 176 2.4. EDNS Flags 178 Some servers fail to respond to EDNS queries with EDNS flags set. 179 Server should ignore EDNS flags there do not understand and should 180 not add them to the response [RFC6891]. 182 2.5. DNS Flags 184 Some servers fail to respond to DNS queries with various DNS flags 185 set, regardless of whether they are defined or still reserved. At 186 the time of writing there are servers that fail to respond to queries 187 with the AD bit set to 1 and servers that fail to respond to queries 188 with the last reserved flag bit set. 190 2.6. Unknown / Unsupported Type Queries 192 Identifying servers that fail to respond to unknown or unsupported 193 types can be done by making an initial DNS query for an A record, 194 making a number of queries for an unallocated type, them making a 195 query for an A record again. IANA maintains a registry of allocated 196 types. 198 If the server responds to the first and last queries but fails to 199 respond to the queries for the unallocated type, it is probably 200 faulty. The test should be repeated a number of times to eliminate 201 the likelihood of a false positive due to packet loss. 203 2.7. Unknown DNS opcodes 205 The use of previously undefined opcodes is to be expected. Since the 206 DNS was first defined two new opcodes have been added, UPDATE and 207 NOTIFY. 209 NOTIMP is the expected rcode to an unknown / unimplemented opcode. 211 Note: while new opcodes will most probably use the current layout 212 structure for the rest of the message there is no requirement than 213 anything other than the DNS header match. 215 2.8. TCP Queries 217 All DNS servers are supposed to respond to queries over TCP 218 [RFC5966]. Firewalls that drop TCP connection attempts rather that 219 resetting the connect attempt or send a ICMP/ICMPv6 administratively 220 prohibited message introduce excessive delays to the resolution 221 process. 223 Whether a server accepts TCP connections can be tested by first 224 checking that it responds to UDP queries to confirm that it is up and 225 operating, then attempting the same query over TCP. An additional 226 query should be made over UDP if the TCP connection attempt fails to 227 confirm that the server under test is still operating. 229 3. Remediating 231 While the first step in remediating this problem is to get the 232 offending nameserver code corrected, there is a very long tail 233 problem with DNS servers in that it can often take over a decade 234 between the code being corrected and a nameserver being upgraded with 235 corrected code. With that in mind it is requested that TLD, and 236 other similar zone operators, take steps to identify and inform their 237 customers, directly or indirectly through registrars, that they are 238 running such servers and that the customers need to correct the 239 problem. 241 TLD operators are being asked to do this as they, due to the nature 242 of running a TLD and the hierarchical nature of the DNS, have access 243 to a large numbers of nameserver names as well as contact details for 244 the registrants of those nameservers. One can construct lists of 245 nameservers from other sources and that has been done to survey the 246 state of the Internet, but that doesn't give you the contact details 247 necessary to inform the operators. The SOA RNAME is often invalid 248 and whois data is obscured and / or not available which makes it 249 infeasible for others to do this. 251 TLD operators should construct a list of servers child zones are 252 delegated to along with a delegated zone name. This name shall be 253 the query name used to test the server as it is supposed to exist. 255 For each server the TLD operator shall make an SOA query of the 256 delegated zone name. This should result in the SOA record being 257 returned in the answer section. If the SOA record is not returned 258 but some other response is returned, this is a indication of a bad 259 delegation and the TLD operator should take whatever steps it 260 normally takes to rectify a bad delegation. If more that one zone is 261 delegated to the server, it should choose another zone until it finds 262 a zone which responds correctly or it exhausts the list of zones 263 delegated to the server. 265 If the server fails to get a response to a SOA query, the TLD 266 operator should make an A query as some nameservers fail to respond 267 to SOA queries but respond to A queries. If it gets no response to 268 the A query, another delegated zone should be queried for as some 269 nameservers fail to respond to zones they are not configured for. If 270 subsequent queries find a responding zone, all delegation to this 271 server need to be checked and rectified using the TLD's normal 272 procedures. 274 Having identified a working tuple the TLD 275 operator should now check that the server responds to EDNS, Unknown 276 Query Type and TCP tests as described above. If the TLD operator 277 finds that server fails any of the tests, the TLD operator shall take 278 steps to inform the operator of the server that they are running a 279 faulty nameserver and that they need to take steps to correct the 280 matter. The TLD operator shall also record the 281 for follow-up testing. 283 If repeated attempts to inform and get the customer to correct / 284 replace the faulty server are unsuccessful the TLD operator shall 285 remove all delegations to said server from the zone. 287 It will also be necessary for TLD operators to repeat the scans 288 periodically. It is recommended that this be performed monthly 289 backing off to bi-annually once the numbers of faulty servers found 290 drops off to less than 1 in 100000 servers tested. Follow-up tests 291 for faulty servers still need to be performed monthly. 293 Some operators claim that they can't perform checks at registration 294 time. If a check is not performed at registration time, it needs to 295 be performed within a week of registration in order to detect faulty 296 servers swiftly. 298 Checking of delegations by TLD operators should be nothing new as 299 they have been required from the very beginnings of DNS to do this 300 [RFC1034]. Checking for compliance of nameserver operations should 301 just be a extension of such testing. 303 It is recommended that TLD operators setup a test web page which 304 performs the tests the TLD operator performs as part of their regular 305 audits to allow nameserver operators to test that they have correctly 306 fixed their servers. Such tests should be rate limited to avoid 307 these pages being a denial of service vector. 309 4. Firewalls and Load Balancers 311 Firewalls and load balancers can affect the externally visible 312 behaviour of a nameserver. Tests for conformance need to be done 313 from outside of any firewall so that the system as a whole is tested. 315 Firewalls and load balancers should not drop DNS packets that they 316 don't understand. They should either pass through the packets or 317 generate an appropriate error response. 319 Requests for unknown query types are not attacks and should not be 320 treated as such. 322 Requests with unassigned flags set (DNS or EDNS) are not attacks and 323 should not be treated as such. The behaviour for unassigned is to 324 ignore them in the request and to not set them in the response. All 325 dropping DNS / EDNS packets with unassigned flags does is make it 326 harder to deploy extensions that make use of them due to the need to 327 reconfigure / update firewalls. 329 Requests with unknown EDNS options are not an attack and should not 330 be treated as such. The correct behaviour for unknown EDNS options 331 is to ignore them. 333 Requests with unknown EDNS versions are not a attack and should not 334 be treated as such. The correct behaviour for unknown EDNS versions 335 is to return BADVERS along with the highest EDNS version the server 336 supports. All dropping EDNS packets does is break EDNS version 337 negotiation. 339 Firewalls should not assume that there will only be a single response 340 message to a requests. There have been proposals to use EDNS to 341 signal that multiple DNS messages be returned rather than a single 342 UDP message that is fragmented at the IP layer. 344 5. Scrubbing Services 346 Scrubbing services, like firewalls, can affect the externally visible 347 behaviour of a nameserver. If you use a scrubbing service, you 348 should check that legitimate queries are not being blocked. 350 Scrubbing services, unlike firewalls, are also turned on and off in 351 response to denial of service attacks. One needs to take care when 352 choosing a scrubbing service and ask questions like: 354 Do they pass unknown DNS query types? 356 Do they pass unknown EDNS versions? 358 Do they pass unknown EDNS options? 360 Do they pass unknown EDNS flags? 362 Do they pass requests with unknown DNS opcodes? 364 Do they pass requests with the remaining reserved DNS header flag 365 bit set? 367 All of these are not attack vectors but some scrubbing services treat 368 them as such. 370 6. Whole Answer Caches 372 Whole answer caches can return the wrong response to a query if they 373 do not take all of the query into account. This has implications 374 when testing and with overall protocol compliance. 376 e.g. There are whole answer caches that ignore the EDNS version 377 field which results in incorrect answers to non EDNS version 0 378 queries being returned if they were proceeded by a EDNS version 0 379 query for the same name and type. 381 7. Response Code Selection 383 Choosing the correct response code when fixing a nameserver is 384 important. Just because a type is not implemented does not mean that 385 NOTIMP is the correct response code to return. Response codes need 386 to be chosen considering how clients will handle them. 388 For unimplemented opcodes NOTIMP is the expected response code. 389 Additionally a new opcode could change the message format by 390 extending the header or changing the structure of the records etc. 391 This may result in FORMERR being returned though NOTIMP would be more 392 correct. 394 In general, for unimplemented type codes Name Error (NXDOMAIN) and 395 NOERROR (no data) are the expected response codes. A server is not 396 supposed to serve a zone which contains unsupported types ([RFC1034]) 397 so the only thing left is return if the QNAME exists or not. NOTIMP 398 and REFUSED are not useful responses as they force the clients to try 399 all the authoritative servers for a zone looking for a server which 400 will answer the query. 402 Meta queries type may be the exception but these need to be thought 403 about on a case by case basis. 405 If you support EDNS and get a query with an unsupported EDNS version, 406 the correct response is BADVERS [RFC6891]. 408 If you do not support EDNS at all, FORMERR and NOTIMP are the 409 expected error codes. That said a minimal EDNS server implementation 410 just requires parsing the OPT records and responding with an empty 411 OPT record. There is no need to interpret any EDNS options present 412 in the request as unsupported options are expected to be ignored 413 [RFC6891]. 415 8. Testing 417 This first set of tests cover basic DNS server behaviour and all 418 servers should pass these tests. 420 Verify the server is configured for the zone: 422 dig +noedns +noad +norec soa $zone @$server 424 expect: status: NOERROR 425 expect: SOA record 426 expect: flag: aa to be present 428 Check that TCP queries work: 430 dig +noedns +noad +norec +tcp soa $zone @$server 432 expect: status: NOERROR 433 expect: SOA record 434 expect: flag: aa to be present 436 Check that queries for an unknown type to work: 438 dig +noedns +noad +norec type1000 $zone @$server 440 expect: status: NOERROR 441 expect: an empty answer section. 442 expect: flag: aa to be present 444 Check that queries with CD=1 work: 446 dig +noedns +noad +norec +cd soa $zone @$server 448 expect: status: NOERROR 449 expect: SOA record to be present 450 expect: flag: aa to be present 452 Check that queries with AD=1 work: 454 dig +noedns +norec +ad soa $zone @$server 456 expect: status: NOERROR 457 expect: SOA record to be present 458 expect: flag: aa to be present 460 Check that queries with the last unassigned DNS header flag to work: 462 dig +noedns +noad +norec +zflag soa $zone @$server 464 expect: status: NOERROR 465 expect: SOA record to be present 466 expect: MBZ to not be in the response 467 expect: flag: aa to be present 469 MBZ (Must Be Zero) presence indicates the flag bit has been copied. 471 Check that new opcodes are handled: 473 dig +noedns +noad +opcode=15 +norec soa $zone @$server 475 expect: status: NOTIMP 476 expect: SOA record to not be present 477 expect: flag: aa to NOT be present 478 The next set of test cover various aspects of EDNS behaviour. If any 479 of these tests succeed, then all of them should succeed. There are 480 servers that support EDNS but fail to handle plain EDNS queries 481 correctly so a plain EDNS query is not a good indicator of lack of 482 EDNS support. 484 Check that plain EDNS queries work: 486 dig +nocookie +edns=0 +noad +norec soa $zone @$server 488 expect: status: NOERROR 489 expect: SOA record to be present 490 expect: OPT record to be present 491 expect: EDNS Version 0 in response 492 expect: flag: aa to be present 494 +nocookie disables sending a EDNS COOKIE option in which is on by 495 default. 497 Check that EDNS version 1 queries work (EDNS supported): 499 dig +nocookie +edns=1 +noednsneg +noad +norec soa $zone @$server 501 expect: status: BADVERS 502 expect: SOA record to not be present 503 expect: OPT record to be present 504 expect: EDNS Version 0 in response 505 expect: flag: aa to NOT be present 507 (Only EDNS Version 0 is currently defined so the response should 508 always be a 0 version. This will change when EDNS version 1 is 509 defined.) 511 Check that EDNS queries with an unknown option work (EDNS supported): 513 dig +nocookie +edns=0 +noad +norec +ednsopt=100 soa $zone @$server 515 expect: status: NOERROR 516 expect: SOA record to be present 517 expect: OPT record to be present 518 expect: OPT=100 to not be present 519 expect: EDNS Version 0 in response 520 expect: flag: aa to be present 522 Check that EDNS queries with unknown flags work (EDNS supported): 524 dig +nocookie +edns=0 +noad +norec +ednsflags=0x40 soa $zone @$server 526 expect: status: NOERROR 527 expect: SOA record to be present 528 expect: OPT record to be present 529 expect: MBZ not to be present 530 expect: EDNS Version 0 in response 531 expect: flag: aa to be present 533 MBZ (Must Be Zero) presence indicates the flag bit has been copied. 535 Check that EDNS version 1 queries with unknown flags work (EDNS 536 supported): 538 dig +nocookie +edns=1 +noednsneg +noad +norec +ednsflags=0x40 soa \ 539 $zone @$server 541 expect: status: BADVERS 542 expect: SOA record to NOT be present 543 expect: OPT record to be present 544 expect: MBZ not to be present 545 expect: EDNS Version 0 in response 546 expect: flag: aa to NOT be present 548 +noednsneg disables EDNS version negotiation in DiG; MBZ (Must Be 549 Zero) presence indicates the flag bit has been copied. 551 Check that EDNS version 1 queries with unknown options work (EDNS 552 supported): 554 dig +nocookie +edns=1 +noednsneg +noad +norec +ednsopt=100 soa \ 555 $zone @$server 557 expect: status: BADVERS 558 expect: SOA record to NOT be present 559 expect: OPT record to be present 560 expect: OPT=100 to NOT be present 561 expect: EDNS Version 0 in response 562 expect: flag: aa to be present 564 +noednsneg disables EDNS version negotiation in DiG. 566 Check that a DNSSEC queries work (EDNS supported): 568 dig +nocookie +edns=0 +noad +norec +dnssec soa $zone @$server 570 expect: status: NOERROR 571 expect: SOA record to be present 572 expect: OPT record to be present 573 expect: DO=1 to be present if a RRSIG is in the response 574 expect: EDNS Version 0 in response 575 expect: flag: aa to be present 577 DO=1 should be present if RRSIGs are returned as they indicate that 578 the server supports DNSSEC. Servers that support DNSSEC are supposed 579 to copy the DO bit from the request to the response as per [RFC3225]. 581 Check that EDNS version 1 DNSSEC queries work (EDNS supported): 583 dig +nocookie +edns=1 +noednsneg +noad +norec +dnssec soa \ 584 $zone @$server 586 expect: status: BADVERS 587 expect: SOA record to not be present 588 expect: OPT record to be present 589 expect: DO=1 to be present if the EDNS version 0 DNSSEC query test 590 returned DO=1 591 expect: EDNS Version 0 in response 592 expect: flag: aa to NOT be present 594 +noednsneg disables EDNS version negotiation in DiG. 596 Check that EDNS queries with multiple defined EDNS options work. 598 dig +edns=0 +noad +norec +cookie +nsid +expire +subnet=0.0.0.0/0 \ 599 soa $zone @$server 601 expect: status: NOERROR 602 expect: SOA record to be present 603 expect: OPT record to be present 604 expect: EDNS Version 0 in response 605 expect: flag: aa to be present 607 If EDNS is not supported by the nameserver, we expect a response to 608 all the above queries. That response may be a FORMERR or NOTIMP 609 error response or the OPT record may just be ignored. 611 It is advisable to run all the above tests in parallel so as to 612 minimise the delays due to multiple timeouts when the servers do not 613 respond. 615 The above tests use dig from BIND 9.11.0 which is still in 616 development. 618 9. Security Considerations 620 Testing protocol compliance can potentially result in false reports 621 of attempts to break services from Intrusion Detection Services and 622 firewalls. None of the tests listed above should break nominally 623 EDNS compliant servers. None of the tests above should break non 624 EDNS servers. All the tests above are well formed, though not 625 necessarily common, DNS queries. 627 Relaxing firewall settings to ensure EDNS compliance could 628 potentially expose a critical implementation flaw in the nameserver. 629 Nameservers should be tested for conformance before relaxing firewall 630 settings. 632 10. IANA Considerations 634 IANA / ICANN needs to consider what tests, if any, from above that it 635 should add to the zone maintenance procedures for zones under its 636 control including pre-delegation checks. Otherwise this document has 637 no actions for IANA. 639 11. Normative References 641 [RFC1034] Mockapetris, P., "Domain names - concepts and facilities", 642 STD 13, RFC 1034, DOI 10.17487/RFC1034, November 1987, 643 . 645 [RFC1035] Mockapetris, P., "Domain names - implementation and 646 specification", STD 13, RFC 1035, DOI 10.17487/RFC1035, 647 November 1987, . 649 [RFC3225] Conrad, D., "Indicating Resolver Support of DNSSEC", RFC 650 3225, DOI 10.17487/RFC3225, December 2001, 651 . 653 [RFC5966] Bellis, R., "DNS Transport over TCP - Implementation 654 Requirements", RFC 5966, DOI 10.17487/RFC5966, August 655 2010, . 657 [RFC6891] Damas, J., Graff, M., and P. Vixie, "Extension Mechanisms 658 for DNS (EDNS(0))", STD 75, RFC 6891, DOI 10.17487/ 659 RFC6891, April 2013, 660 . 662 Author's Address 664 M. Andrews 665 Internet Systems Consortium 666 950 Charter Street 667 Redwood City, CA 94063 668 US 670 Email: marka@isc.org