idnits 2.17.1 draft-ietf-p2psip-base-13.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 == Line 1831 has weird spacing: '...Options optio...' == Line 2085 has weird spacing: '...ionType type;...' == Line 2233 has weird spacing: '...te_type typ...' == Line 2294 has weird spacing: '...tyValue ide...' == Line 2597 has weird spacing: '...ionType typ...' == (4 more instances...) -- The document seems to contain a disclaimer for pre-RFC5378 work, and may have content which was first submitted before 10 November 2008. The disclaimer is necessary when there are original authors that you have been unable to contact, or if some do not wish to grant the BCP78 rights to the IETF Trust. If you are able to get all authors (current and original) to grant those rights, you can and should remove the disclaimer; otherwise, the disclaimer is needed and you can ignore this comment. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (March 14, 2011) is 4789 days in the past. Is this intentional? -- Found something which looks like a code comment -- if you have code sections in the document, please surround them with '' and '' lines. 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: 'NodeIdLength' is mentioned on line 1726, but not defined -- Looks like a reference, but probably isn't: '0' on line 4552 == Missing Reference: 'RFC-AAAA' is mentioned on line 6240, but not defined ** Obsolete normative reference: RFC 2818 (Obsoleted by RFC 9110) ** Obsolete normative reference: RFC 2988 (Obsoleted by RFC 6298) ** Downref: Normative reference to an Informational RFC: RFC 3174 ** Obsolete normative reference: RFC 3447 (Obsoleted by RFC 8017) ** Obsolete normative reference: RFC 4347 (Obsoleted by RFC 6347) ** Obsolete normative reference: RFC 4395 (Obsoleted by RFC 7595) ** Obsolete normative reference: RFC 4634 (Obsoleted by RFC 6234) ** Obsolete normative reference: RFC 5245 (Obsoleted by RFC 8445, RFC 8839) ** Obsolete normative reference: RFC 5246 (Obsoleted by RFC 8446) ** Obsolete normative reference: RFC 5389 (Obsoleted by RFC 8489) ** Obsolete normative reference: RFC 5766 (Obsoleted by RFC 8656) == Outdated reference: A later version (-10) exists of draft-ietf-hip-reload-instance-03 == Outdated reference: A later version (-16) exists of draft-ietf-mmusic-ice-tcp-12 == Outdated reference: A later version (-09) exists of draft-ietf-p2psip-concepts-03 == Outdated reference: A later version (-21) exists of draft-ietf-p2psip-sip-05 == Outdated reference: A later version (-05) exists of draft-jiang-p2psip-relay-04 -- Obsolete informational reference (is this intentional?): RFC 5201 (Obsoleted by RFC 7401) -- Obsolete informational reference (is this intentional?): RFC 5785 (Obsoleted by RFC 8615) Summary: 11 errors (**), 0 flaws (~~), 14 warnings (==), 7 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 P2PSIP C. Jennings 3 Internet-Draft Cisco 4 Intended status: Standards Track B. Lowekamp, Ed. 5 Expires: September 15, 2011 Skype 6 E. Rescorla 7 RTFM, Inc. 8 S. Baset 9 H. Schulzrinne 10 Columbia University 11 March 14, 2011 13 REsource LOcation And Discovery (RELOAD) Base Protocol 14 draft-ietf-p2psip-base-13 16 Abstract 18 This specification defines REsource LOcation And Discovery (RELOAD), 19 a peer-to-peer (P2P) signaling protocol for use on the Internet. A 20 P2P signaling protocol provides its clients with an abstract storage 21 and messaging service between a set of cooperating peers that form 22 the overlay network. RELOAD is designed to support a P2P Session 23 Initiation Protocol (P2PSIP) network, but can be utilized by other 24 applications with similar requirements by defining new usages that 25 specify the kinds of data that must be stored for a particular 26 application. RELOAD defines a security model based on a certificate 27 enrollment service that provides unique identities. NAT traversal is 28 a fundamental service of the protocol. RELOAD also allows access 29 from "client" nodes that do not need to route traffic or store data 30 for others. 32 Legal 34 THIS DOCUMENT AND THE INFORMATION CONTAINED THEREIN ARE PROVIDED ON 35 AN "AS IS" BASIS AND THE CONTRIBUTOR, THE ORGANIZATION HE/SHE 36 REPRESENTS OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE 37 IETF TRUST, AND THE INTERNET ENGINEERING TASK FORCE, DISCLAIM ALL 38 WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY 39 WARRANTY THAT THE USE OF THE INFORMATION THEREIN WILL NOT INFRINGE 40 ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS 41 FOR A PARTICULAR PURPOSE. 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 September 15, 2011. 60 Copyright Notice 62 Copyright (c) 2011 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 This document may contain material from IETF Documents or IETF 76 Contributions published or made publicly available before November 77 10, 2008. The person(s) controlling the copyright in some of this 78 material may not have granted the IETF Trust the right to allow 79 modifications of such material outside the IETF Standards Process. 80 Without obtaining an adequate license from the person(s) controlling 81 the copyright in such materials, this document may not be modified 82 outside the IETF Standards Process, and derivative works of it may 83 not be created outside the IETF Standards Process, except to format 84 it for publication as an RFC or to translate it into languages other 85 than English. 87 Table of Contents 89 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 8 90 1.1. Basic Setting . . . . . . . . . . . . . . . . . . . . . 9 91 1.2. Architecture . . . . . . . . . . . . . . . . . . . . . . 10 92 1.2.1. Usage Layer . . . . . . . . . . . . . . . . . . . . 13 93 1.2.2. Message Transport . . . . . . . . . . . . . . . . . 14 94 1.2.3. Storage . . . . . . . . . . . . . . . . . . . . . . 14 95 1.2.4. Topology Plugin . . . . . . . . . . . . . . . . . . 15 96 1.2.5. Forwarding and Link Management Layer . . . . . . . . 15 97 1.3. Security . . . . . . . . . . . . . . . . . . . . . . . . 16 98 1.4. Structure of This Document . . . . . . . . . . . . . . . 17 99 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 17 100 3. Overlay Management Overview . . . . . . . . . . . . . . . . . 20 101 3.1. Security and Identification . . . . . . . . . . . . . . 20 102 3.1.1. Shared-Key Security . . . . . . . . . . . . . . . . 21 103 3.2. Clients . . . . . . . . . . . . . . . . . . . . . . . . 21 104 3.2.1. Client Routing . . . . . . . . . . . . . . . . . . . 22 105 3.2.2. Minimum Functionality Requirements for Clients . . . 22 106 3.3. Routing . . . . . . . . . . . . . . . . . . . . . . . . 23 107 3.4. Connectivity Management . . . . . . . . . . . . . . . . 25 108 3.5. Overlay Algorithm Support . . . . . . . . . . . . . . . 26 109 3.5.1. Support for Pluggable Overlay Algorithms . . . . . . 26 110 3.5.2. Joining, Leaving, and Maintenance Overview . . . . . 26 111 3.6. First-Time Setup . . . . . . . . . . . . . . . . . . . . 28 112 3.6.1. Initial Configuration . . . . . . . . . . . . . . . 28 113 3.6.2. Enrollment . . . . . . . . . . . . . . . . . . . . . 28 114 4. Application Support Overview . . . . . . . . . . . . . . . . 29 115 4.1. Data Storage . . . . . . . . . . . . . . . . . . . . . . 29 116 4.1.1. Storage Permissions . . . . . . . . . . . . . . . . 30 117 4.1.2. Replication . . . . . . . . . . . . . . . . . . . . 31 118 4.2. Usages . . . . . . . . . . . . . . . . . . . . . . . . . 31 119 4.3. Service Discovery . . . . . . . . . . . . . . . . . . . 32 120 4.4. Application Connectivity . . . . . . . . . . . . . . . . 32 121 5. Overlay Management Protocol . . . . . . . . . . . . . . . . . 33 122 5.1. Message Receipt and Forwarding . . . . . . . . . . . . . 33 123 5.1.1. Responsible ID . . . . . . . . . . . . . . . . . . . 33 124 5.1.2. Other ID . . . . . . . . . . . . . . . . . . . . . . 34 125 5.1.3. Private ID . . . . . . . . . . . . . . . . . . . . . 35 126 5.2. Symmetric Recursive Routing . . . . . . . . . . . . . . 36 127 5.2.1. Request Origination . . . . . . . . . . . . . . . . 36 128 5.2.2. Response Origination . . . . . . . . . . . . . . . . 37 129 5.3. Message Structure . . . . . . . . . . . . . . . . . . . 37 130 5.3.1. Presentation Language . . . . . . . . . . . . . . . 38 131 5.3.1.1. Common Definitions . . . . . . . . . . . . . . . 38 132 5.3.2. Forwarding Header . . . . . . . . . . . . . . . . . 41 133 5.3.2.1. Processing Configuration Sequence Numbers . . . . 43 134 5.3.2.2. Destination and Via Lists . . . . . . . . . . . . 44 135 5.3.2.3. Forwarding Options . . . . . . . . . . . . . . . 46 136 5.3.3. Message Contents Format . . . . . . . . . . . . . . 47 137 5.3.3.1. Response Codes and Response Errors . . . . . . . 48 138 5.3.4. Security Block . . . . . . . . . . . . . . . . . . . 50 139 5.4. Overlay Topology . . . . . . . . . . . . . . . . . . . . 53 140 5.4.1. Topology Plugin Requirements . . . . . . . . . . . . 53 141 5.4.2. Methods and types for use by topology plugins . . . 54 142 5.4.2.1. Join . . . . . . . . . . . . . . . . . . . . . . 54 143 5.4.2.2. Leave . . . . . . . . . . . . . . . . . . . . . . 55 144 5.4.2.3. Update . . . . . . . . . . . . . . . . . . . . . 55 145 5.4.2.4. RouteQuery . . . . . . . . . . . . . . . . . . . 56 146 5.4.2.5. Probe . . . . . . . . . . . . . . . . . . . . . . 57 147 5.5. Forwarding and Link Management Layer . . . . . . . . . . 59 148 5.5.1. Attach . . . . . . . . . . . . . . . . . . . . . . . 59 149 5.5.1.1. Request Definition . . . . . . . . . . . . . . . 60 150 5.5.1.2. Response Definition . . . . . . . . . . . . . . . 63 151 5.5.1.3. Using ICE With RELOAD . . . . . . . . . . . . . . 63 152 5.5.1.4. Collecting STUN Servers . . . . . . . . . . . . . 64 153 5.5.1.5. Gathering Candidates . . . . . . . . . . . . . . 65 154 5.5.1.6. Prioritizing Candidates . . . . . . . . . . . . . 65 155 5.5.1.7. Encoding the Attach Message . . . . . . . . . . . 66 156 5.5.1.8. Verifying ICE Support . . . . . . . . . . . . . . 66 157 5.5.1.9. Role Determination . . . . . . . . . . . . . . . 66 158 5.5.1.10. Full ICE . . . . . . . . . . . . . . . . . . . . 67 159 5.5.1.11. No-ICE . . . . . . . . . . . . . . . . . . . . . 67 160 5.5.1.12. Subsequent Offers and Answers . . . . . . . . . . 67 161 5.5.1.13. Sending Media . . . . . . . . . . . . . . . . . . 67 162 5.5.1.14. Receiving Media . . . . . . . . . . . . . . . . . 68 163 5.5.2. AppAttach . . . . . . . . . . . . . . . . . . . . . 68 164 5.5.2.1. Request Definition . . . . . . . . . . . . . . . 68 165 5.5.2.2. Response Definition . . . . . . . . . . . . . . . 69 166 5.5.3. Ping . . . . . . . . . . . . . . . . . . . . . . . . 69 167 5.5.3.1. Request Definition . . . . . . . . . . . . . . . 70 168 5.5.3.2. Response Definition . . . . . . . . . . . . . . . 70 169 5.5.4. ConfigUpdate . . . . . . . . . . . . . . . . . . . . 70 170 5.5.4.1. Request Definition . . . . . . . . . . . . . . . 71 171 5.5.4.2. Response Definition . . . . . . . . . . . . . . . 71 172 5.6. Overlay Link Layer . . . . . . . . . . . . . . . . . . . 72 173 5.6.1. Future Overlay Link Protocols . . . . . . . . . . . 73 174 5.6.1.1. HIP . . . . . . . . . . . . . . . . . . . . . . . 74 175 5.6.1.2. ICE-TCP . . . . . . . . . . . . . . . . . . . . . 74 176 5.6.1.3. Message-oriented Transports . . . . . . . . . . . 74 177 5.6.1.4. Tunneled Transports . . . . . . . . . . . . . . . 74 178 5.6.2. Framing Header . . . . . . . . . . . . . . . . . . . 75 179 5.6.3. Simple Reliability . . . . . . . . . . . . . . . . . 76 180 5.6.3.1. Retransmission and Flow Control . . . . . . . . . 77 181 5.6.4. DTLS/UDP with SR . . . . . . . . . . . . . . . . . . 78 182 5.6.5. TLS/TCP with FH, No-ICE . . . . . . . . . . . . . . 78 183 5.6.6. DTLS/UDP with SR, No-ICE . . . . . . . . . . . . . . 79 184 5.7. Fragmentation and Reassembly . . . . . . . . . . . . . . 79 185 6. Data Storage Protocol . . . . . . . . . . . . . . . . . . . . 80 186 6.1. Data Signature Computation . . . . . . . . . . . . . . . 81 187 6.2. Data Models . . . . . . . . . . . . . . . . . . . . . . 82 188 6.2.1. Single Value . . . . . . . . . . . . . . . . . . . . 83 189 6.2.2. Array . . . . . . . . . . . . . . . . . . . . . . . 83 190 6.2.3. Dictionary . . . . . . . . . . . . . . . . . . . . . 84 191 6.3. Access Control Policies . . . . . . . . . . . . . . . . 84 192 6.3.1. USER-MATCH . . . . . . . . . . . . . . . . . . . . . 85 193 6.3.2. NODE-MATCH . . . . . . . . . . . . . . . . . . . . . 85 194 6.3.3. USER-NODE-MATCH . . . . . . . . . . . . . . . . . . 85 195 6.3.4. NODE-MULTIPLE . . . . . . . . . . . . . . . . . . . 85 196 6.4. Data Storage Methods . . . . . . . . . . . . . . . . . . 86 197 6.4.1. Store . . . . . . . . . . . . . . . . . . . . . . . 86 198 6.4.1.1. Request Definition . . . . . . . . . . . . . . . 86 199 6.4.1.2. Response Definition . . . . . . . . . . . . . . . 90 200 6.4.1.3. Removing Values . . . . . . . . . . . . . . . . . 92 201 6.4.2. Fetch . . . . . . . . . . . . . . . . . . . . . . . 92 202 6.4.2.1. Request Definition . . . . . . . . . . . . . . . 93 203 6.4.2.2. Response Definition . . . . . . . . . . . . . . . 95 204 6.4.3. Stat . . . . . . . . . . . . . . . . . . . . . . . . 95 205 6.4.3.1. Request Definition . . . . . . . . . . . . . . . 96 206 6.4.3.2. Response Definition . . . . . . . . . . . . . . . 96 207 6.4.4. Find . . . . . . . . . . . . . . . . . . . . . . . . 98 208 6.4.4.1. Request Definition . . . . . . . . . . . . . . . 98 209 6.4.4.2. Response Definition . . . . . . . . . . . . . . . 99 210 6.4.5. Defining New Kinds . . . . . . . . . . . . . . . . . 100 211 7. Certificate Store Usage . . . . . . . . . . . . . . . . . . . 100 212 8. TURN Server Usage . . . . . . . . . . . . . . . . . . . . . . 101 213 9. Chord Algorithm . . . . . . . . . . . . . . . . . . . . . . . 103 214 9.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . 104 215 9.2. Hash Function . . . . . . . . . . . . . . . . . . . . . 104 216 9.3. Routing . . . . . . . . . . . . . . . . . . . . . . . . 104 217 9.4. Redundancy . . . . . . . . . . . . . . . . . . . . . . . 105 218 9.5. Joining . . . . . . . . . . . . . . . . . . . . . . . . 105 219 9.6. Routing Attaches . . . . . . . . . . . . . . . . . . . . 106 220 9.7. Updates . . . . . . . . . . . . . . . . . . . . . . . . 106 221 9.7.1. Handling Neighbor Failures . . . . . . . . . . . . . 108 222 9.7.2. Handling Finger Table Entry Failure . . . . . . . . 109 223 9.7.3. Receiving Updates . . . . . . . . . . . . . . . . . 109 224 9.7.4. Stabilization . . . . . . . . . . . . . . . . . . . 110 225 9.7.4.1. Updating neighbor table . . . . . . . . . . . . . 110 226 9.7.4.2. Refreshing finger table . . . . . . . . . . . . . 110 227 9.7.4.3. Adjusting finger table size . . . . . . . . . . . 111 228 9.7.4.4. Detecting partitioning . . . . . . . . . . . . . 112 229 9.8. Route query . . . . . . . . . . . . . . . . . . . . . . 112 230 9.9. Leaving . . . . . . . . . . . . . . . . . . . . . . . . 113 232 10. Enrollment and Bootstrap . . . . . . . . . . . . . . . . . . 114 233 10.1. Overlay Configuration . . . . . . . . . . . . . . . . . 114 234 10.1.1. Relax NG Grammar . . . . . . . . . . . . . . . . . . 120 235 10.2. Discovery Through Configuration Server . . . . . . . . . 122 236 10.3. Credentials . . . . . . . . . . . . . . . . . . . . . . 123 237 10.3.1. Self-Generated Credentials . . . . . . . . . . . . . 124 238 10.4. Searching for a Bootstrap Node . . . . . . . . . . . . . 124 239 10.5. Contacting a Bootstrap Node . . . . . . . . . . . . . . 125 240 11. Message Flow Example . . . . . . . . . . . . . . . . . . . . 125 241 12. Security Considerations . . . . . . . . . . . . . . . . . . . 131 242 12.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . 131 243 12.2. Attacks on P2P Overlays . . . . . . . . . . . . . . . . 132 244 12.3. Certificate-based Security . . . . . . . . . . . . . . . 132 245 12.4. Shared-Secret Security . . . . . . . . . . . . . . . . . 133 246 12.5. Storage Security . . . . . . . . . . . . . . . . . . . . 134 247 12.5.1. Authorization . . . . . . . . . . . . . . . . . . . 134 248 12.5.2. Distributed Quota . . . . . . . . . . . . . . . . . 135 249 12.5.3. Correctness . . . . . . . . . . . . . . . . . . . . 135 250 12.5.4. Residual Attacks . . . . . . . . . . . . . . . . . . 135 251 12.6. Routing Security . . . . . . . . . . . . . . . . . . . . 136 252 12.6.1. Background . . . . . . . . . . . . . . . . . . . . . 136 253 12.6.2. Admissions Control . . . . . . . . . . . . . . . . . 137 254 12.6.3. Peer Identification and Authentication . . . . . . . 137 255 12.6.4. Protecting the Signaling . . . . . . . . . . . . . . 138 256 12.6.5. Residual Attacks . . . . . . . . . . . . . . . . . . 138 257 13. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 139 258 13.1. Well-Known URI Registration . . . . . . . . . . . . . . 139 259 13.2. Port Registrations . . . . . . . . . . . . . . . . . . . 139 260 13.3. Overlay Algorithm Types . . . . . . . . . . . . . . . . 140 261 13.4. Access Control Policies . . . . . . . . . . . . . . . . 140 262 13.5. Application-ID . . . . . . . . . . . . . . . . . . . . . 140 263 13.6. Data Kind-ID . . . . . . . . . . . . . . . . . . . . . . 141 264 13.7. Data Model . . . . . . . . . . . . . . . . . . . . . . . 141 265 13.8. Message Codes . . . . . . . . . . . . . . . . . . . . . 141 266 13.9. Error Codes . . . . . . . . . . . . . . . . . . . . . . 142 267 13.10. Overlay Link Types . . . . . . . . . . . . . . . . . . . 143 268 13.11. Overlay Link Protocols . . . . . . . . . . . . . . . . . 143 269 13.12. Forwarding Options . . . . . . . . . . . . . . . . . . . 144 270 13.13. Probe Information Types . . . . . . . . . . . . . . . . 144 271 13.14. Message Extensions . . . . . . . . . . . . . . . . . . . 144 272 13.15. reload URI Scheme . . . . . . . . . . . . . . . . . . . 145 273 13.15.1. URI Registration . . . . . . . . . . . . . . . . . . 145 274 14. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 146 275 15. References . . . . . . . . . . . . . . . . . . . . . . . . . 146 276 15.1. Normative References . . . . . . . . . . . . . . . . . . 146 277 15.2. Informative References . . . . . . . . . . . . . . . . . 148 278 Appendix A. Change Log . . . . . . . . . . . . . . . . . . . . . 151 279 A.1. Changes since draft-ietf-p2psip-reload-12 . . . . . . . 151 281 Appendix B. Routing Alternatives . . . . . . . . . . . . . . . . 152 282 B.1. Iterative vs Recursive . . . . . . . . . . . . . . . . . 152 283 B.2. Symmetric vs Forward response . . . . . . . . . . . . . 152 284 B.3. Direct Response . . . . . . . . . . . . . . . . . . . . 153 285 B.4. Relay Peers . . . . . . . . . . . . . . . . . . . . . . 154 286 B.5. Symmetric Route Stability . . . . . . . . . . . . . . . 154 287 Appendix C. Why Clients? . . . . . . . . . . . . . . . . . . . . 155 288 C.1. Why Not Only Peers? . . . . . . . . . . . . . . . . . . 155 289 C.2. Clients as Application-Level Agents . . . . . . . . . . 156 290 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 156 292 1. Introduction 294 This document defines REsource LOcation And Discovery (RELOAD), a 295 peer-to-peer (P2P) signaling protocol for use on the Internet. It 296 provides a generic, self-organizing overlay network service, allowing 297 nodes to efficiently route messages to other nodes and to efficiently 298 store and retrieve data in the overlay. RELOAD provides several 299 features that are critical for a successful P2P protocol for the 300 Internet: 302 Security Framework: A P2P network will often be established among a 303 set of peers that do not trust each other. RELOAD leverages a 304 central enrollment server to provide credentials for each peer 305 which can then be used to authenticate each operation. This 306 greatly reduces the possible attack surface. 308 Usage Model: RELOAD is designed to support a variety of 309 applications, including P2P multimedia communications with the 310 Session Initiation Protocol [I-D.ietf-p2psip-sip]. RELOAD allows 311 the definition of new application usages, each of which can define 312 its own data types, along with the rules for their use. This 313 allows RELOAD to be used with new applications through a simple 314 documentation process that supplies the details for each 315 application. 317 NAT Traversal: RELOAD is designed to function in environments where 318 many if not most of the nodes are behind NATs or firewalls. 319 Operations for NAT traversal are part of the base design, 320 including using ICE to establish new RELOAD or application 321 protocol connections. 323 High Performance Routing: The very nature of overlay algorithms 324 introduces a requirement that peers participating in the P2P 325 network route requests on behalf of other peers in the network. 326 This introduces a load on those other peers, in the form of 327 bandwidth and processing power. RELOAD has been defined with a 328 simple, lightweight forwarding header, thus minimizing the amount 329 of effort required by intermediate peers. 331 Pluggable Overlay Algorithms: RELOAD has been designed with an 332 abstract interface to the overlay layer to simplify implementing a 333 variety of structured (e.g., distributed hash tables) and 334 unstructured overlay algorithms. This specification also defines 335 how RELOAD is used with the Chord DHT algorithm, which is 336 mandatory to implement. Specifying a default "must implement" 337 overlay algorithm promotes interoperability, while extensibility 338 allows selection of overlay algorithms optimized for a particular 339 application. 341 These properties were designed specifically to meet the requirements 342 for a P2P protocol to support SIP. This document defines the base 343 protocol for the distributed storage and location service, as well as 344 critical usages for NAT traversal and security. The SIP Usage itself 345 is described separately in [I-D.ietf-p2psip-sip]. RELOAD is not 346 limited to usage by SIP and could serve as a tool for supporting 347 other P2P applications with similar needs. RELOAD is also based on 348 the concepts introduced in [I-D.ietf-p2psip-concepts]. 350 1.1. Basic Setting 352 In this section, we provide a brief overview of the operational 353 setting for RELOAD. See the concepts 354 document[I-D.ietf-p2psip-concepts] for more details. A RELOAD 355 Overlay Instance consists of a set of nodes arranged in a connected 356 graph. Each node in the overlay is assigned a numeric Node-ID which, 357 together with the specific overlay algorithm in use, determines its 358 position in the graph and the set of nodes it connects to. The 359 figure below shows a trivial example which isn't drawn from any 360 particular overlay algorithm, but was chosen for convenience of 361 representation. 363 +--------+ +--------+ +--------+ 364 | Node 10|--------------| Node 20|--------------| Node 30| 365 +--------+ +--------+ +--------+ 366 | | | 367 | | | 368 +--------+ +--------+ +--------+ 369 | Node 40|--------------| Node 50|--------------| Node 60| 370 +--------+ +--------+ +--------+ 371 | | | 372 | | | 373 +--------+ +--------+ +--------+ 374 | Node 70|--------------| Node 80|--------------| Node 90| 375 +--------+ +--------+ +--------+ 376 | 377 | 378 +--------+ 379 | Node 85| 380 |(Client)| 381 +--------+ 383 Because the graph is not fully connected, when a node wants to send a 384 message to another node, it may need to route it through the network. 385 For instance, Node 10 can talk directly to nodes 20 and 40, but not 386 to Node 70. In order to send a message to Node 70, it would first 387 send it to Node 40 with instructions to pass it along to Node 70. 388 Different overlay algorithms will have different connectivity graphs, 389 but the general idea behind all of them is to allow any node in the 390 graph to efficiently reach every other node within a small number of 391 hops. 393 The RELOAD network is not only a messaging network. It is also a 394 storage network. Records are stored under numeric addresses which 395 occupy the same space as node identifiers. Peers are responsible for 396 storing the data associated with some set of addresses as determined 397 by their Node-ID. For instance, we might say that every peer is 398 responsible for storing any data value which has an address less than 399 or equal to its own Node-ID, but greater than the next lowest 400 Node-ID. Thus, Node-20 would be responsible for storing values 401 11-20. 403 RELOAD also supports clients. These are nodes which have Node-IDs 404 but do not participate in routing or storage. For instance, in the 405 figure above Node 85 is a client. It can route to the rest of the 406 RELOAD network via Node 80, but no other node will route through it 407 and Node 90 is still responsible for all addresses between 81-90. We 408 refer to non-client nodes as peers. 410 Other applications (for instance, SIP) can be defined on top of 411 RELOAD and use these two basic RELOAD services to provide their own 412 services. 414 1.2. Architecture 416 RELOAD is fundamentally an overlay network. The following figure 417 shows the layered RELOAD architecture. 419 Application 421 +-------+ +-------+ 422 | SIP | | XMPP | ... 423 | Usage | | Usage | 424 +-------+ +-------+ 425 ------------------------------------ Messaging Service Boundary 426 +------------------+ +---------+ 427 | Message |<--->| Storage | 428 | Transport | +---------+ 429 +------------------+ ^ 430 ^ ^ | 431 | v v 432 | +-------------------+ 433 | | Topology | 434 | | Plugin | 435 | +-------------------+ 436 | ^ 437 v v 438 +------------------+ 439 | Forwarding & | 440 | Link Management | 441 +------------------+ 442 ------------------------------------ Overlay Link Service Boundary 443 +-------+ +------+ 444 |TLS | |DTLS | ... 445 +-------+ +------+ 447 The major components of RELOAD are: 449 Usage Layer: Each application defines a RELOAD usage; a set of data 450 kinds and behaviors which describe how to use the services 451 provided by RELOAD. These usages all talk to RELOAD through a 452 common Message Transport Service. 454 Message Transport: Handles end-to-end reliability, manages request 455 state for the usages, and forwards Store and Fetch operations to 456 the Storage component. Delivers message responses to the 457 component initiating the request. 459 Storage: The Storage component is responsible for processing 460 messages relating to the storage and retrieval of data. It talks 461 directly to the Topology Plugin to manage data replication and 462 migration, and it talks to the Message Transport component to send 463 and receive messages. 465 Topology Plugin: The Topology Plugin is responsible for implementing 466 the specific overlay algorithm being used. It uses the Message 467 Transport component to send and receive overlay management 468 messages, to the Storage component to manage data replication, and 469 directly to the Forwarding Layer to control hop-by-hop message 470 forwarding. This component closely parallels conventional routing 471 algorithms, but is more tightly coupled to the Forwarding Layer 472 because there is no single "routing table" equivalent used by all 473 overlay algorithms. 475 Forwarding and Link Management Layer: Stores and implements the 476 routing table by providing packet forwarding services between 477 nodes. It also handles establishing new links between nodes, 478 including setting up connections across NATs using ICE. 480 Overlay Link Layer: Responsible for actually transporting traffic 481 directly between nodes. Each such protocol includes the 482 appropriate provisions for per-hop framing or hop-by-hop ACKs 483 required by unreliable transports. TLS [RFC5246] and DTLS 484 [RFC4347] are the currently defined "link layer" protocols used by 485 RELOAD for hop-by-hop communication. New protocols MAY be 486 defined, as described in Section 5.6.1 and Section 10.1. As this 487 document defines only TLS and DTLS, we use those terms throughout 488 the remainder of the document with the understanding that some 489 future specification may add new overlay link layers. 491 To further clarify the roles of the various layers, this figure 492 parallels the architecture with each layer's role from an overlay 493 perspective and implementation layer in the internet: 495 | Internet Model | 496 Real | Equivalent | Reload 497 Internet | in Overlay | Architecture 498 -------------+-----------------+------------------------------------ 499 | | +-------+ +-------+ 500 | Application | | SIP | | XMPP | ... 501 | | | Usage | | Usage | 502 | | +-------+ +-------+ 503 | | ---------------------------------- 504 | |+------------------+ +---------+ 505 | Transport || Message |<--->| Storage | 506 | || Transport | +---------+ 507 | |+------------------+ ^ 508 | | ^ ^ | 509 | | | v v 510 Application | | | +-------------------+ 511 | (Routing) | | | Topology | 512 | | | | Plugin | 513 | | | +-------------------+ 514 | | | ^ 515 | | v v 516 | Network | +------------------+ 517 | | | Forwarding & | 518 | | | Link Management | 519 | | +------------------+ 520 | | ---------------------------------- 521 Transport | Link | +-------+ +------+ 522 | | |TLS | |DTLS | ... 523 | | +-------+ +------+ 524 -------------+-----------------+------------------------------------ 525 Network | 526 | 527 Link | 529 1.2.1. Usage Layer 531 The top layer, called the Usage Layer, has application usages, such 532 as the SIP Registration Usage [I-D.ietf-p2psip-sip], that use the 533 abstract Message Transport Service provided by RELOAD. The goal of 534 this layer is to implement application-specific usages of the generic 535 overlay services provided by RELOAD. The usage defines how a 536 specific application maps its data into something that can be stored 537 in the overlay, where to store the data, how to secure the data, and 538 finally how applications can retrieve and use the data. 540 The architecture diagram shows both a SIP usage and an XMPP usage. A 541 single application may require multiple usages; for example a 542 softphone application may also require a voicemail usage. An usage 543 may define multiple kinds of data that are stored in the overlay and 544 may also rely on kinds originally defined by other usages. 546 Because the security and storage policies for each kind are dictated 547 by the usage defining the kind, the usages may be coupled with the 548 Storage component to provide security policy enforcement and to 549 implement appropriate storage strategies according to the needs of 550 the usage. The exact implementation of such an interface is outside 551 the scope of this specification. 553 1.2.2. Message Transport 555 The Message Transport component provides a generic message routing 556 service for the overlay. The Message Transport layer is responsible 557 for end-to-end message transactions, including retransmissions. Each 558 peer is identified by its location in the overlay as determined by 559 its Node-ID. A component that is a client of the Message Transport 560 can perform two basic functions: 562 o Send a message to a given peer specified by Node-ID or to the peer 563 responsible for a particular Resource-ID. 564 o Receive messages that other peers sent to a Node-ID or Resource-ID 565 for which the receiving peer is responsible. 567 All usages rely on the Message Transport component to send and 568 receive messages from peers. For instance, when a usage wants to 569 store data, it does so by sending Store requests. Note that the 570 Storage component and the Topology Plugin are themselves clients of 571 the Message Transport, because they need to send and receive messages 572 from other peers. 574 The Message Transport Service is similar to those described as 575 providing "Key based routing" (KBR), although as RELOAD supports 576 different overlay algorithms (including non-DHT overlay algorithms) 577 that calculate keys in different ways, the actual interface must 578 accept Resource Names rather than actual keys. 580 1.2.3. Storage 582 One of the major functions of RELOAD is to allow nodes to store data 583 in the overlay and to retrieve data stored by other nodes or by 584 themselves. The Storage component is responsible for processing data 585 storage and retrieval messages. For instance, the Storage component 586 might receive a Store request for a given resource from the Message 587 Transport. It would then query the appropriate usage before storing 588 the data value(s) in its local data store and sending a response to 589 the Message Transport for delivery to the requesting node. 590 Typically, these messages will come from other nodes, but depending 591 on the overlay topology, a node might be responsible for storing data 592 for itself as well, especially if the overlay is small. 594 A peer's Node-ID determines the set of resources that it will be 595 responsible for storing. However, the exact mapping between these is 596 determined by the overlay algorithm in use. The Storage component 597 will only receive a Store request from the Message Transport if this 598 peer is responsible for that Resource-ID. The Storage component is 599 notified by the Topology Plugin when the Resource-IDs for which it is 600 responsible change, and the Storage component is then responsible for 601 migrating resources to other peers, as required. 603 1.2.4. Topology Plugin 605 RELOAD is explicitly designed to work with a variety of overlay 606 algorithms. In order to facilitate this, the overlay algorithm 607 implementation is provided by a Topology Plugin so that each overlay 608 can select an appropriate overlay algorithm that relies on the common 609 RELOAD core protocols and code. 611 The Topology Plugin is responsible for maintaining the overlay 612 algorithm Routing Table, which is consulted by the Forwarding and 613 Link Management Layer before routing a message. When connections are 614 made or broken, the Forwarding and Link Management Layer notifies the 615 Topology Plugin, which adjusts the routing table as appropriate. The 616 Topology Plugin will also instruct the Forwarding and Link Management 617 Layer to form new connections as dictated by the requirements of the 618 overlay algorithm Topology. The Topology Plugin issues periodic 619 update requests through Message Transport to maintain and update its 620 Routing Table. 622 As peers enter and leave, resources may be stored on different peers, 623 so the Topology Plugin also keeps track of which peers are 624 responsible for which resources. As peers join and leave, the 625 Topology Plugin instructs the Storage component to issue resource 626 migration requests as appropriate, in order to ensure that other 627 peers have whatever resources they are now responsible for. The 628 Topology Plugin is also responsible for providing for redundant data 629 storage to protect against loss of information in the event of a peer 630 failure and to protect against compromised or subversive peers. 632 1.2.5. Forwarding and Link Management Layer 634 The Forwarding and Link Management Layer is responsible for getting a 635 message to the next peer, as determined by the Topology Plugin. This 636 Layer establishes and maintains the network connections as required 637 by the Topology Plugin. This layer is also responsible for setting 638 up connections to other peers through NATs and firewalls using ICE, 639 and it can elect to forward traffic using relays for NAT and firewall 640 traversal. 642 This layer provides a generic interface that allows the topology 643 plugin to control the overlay and resource operations and messages. 644 Since each overlay algorithm is defined and functions differently, we 645 generically refer to the table of other peers that the overlay 646 algorithm maintains and uses to route requests (neighbors) as a 647 Routing Table. The Topology Plugin actually owns the Routing Table, 648 and forwarding decisions are made by querying the Topology Plugin for 649 the next hop for a particular Node-ID or Resource-ID. If this node 650 is the destination of the message, the message is delivered to the 651 Message Transport. 653 This layer also utilizes a framing header to encapsulate messages as 654 they are forwarding along each hop. This header aids reliability 655 congestion control, flow control, etc. It has meaning only in the 656 context of that individual link. 658 The Forwarding and Link Management Layer sits on top of the Overlay 659 Link Layer protocols that carry the actual traffic. This 660 specification defines how to use DTLS and TLS protocols to carry 661 RELOAD messages. 663 1.3. Security 665 RELOAD's security model is based on each node having one or more 666 public key certificates. In general, these certificates will be 667 assigned by a central server which also assigns Node-IDs, although 668 self-signed certificates can be used in closed networks. These 669 credentials can be leveraged to provide communications security for 670 RELOAD messages. RELOAD provides communications security at three 671 levels: 673 Connection Level: Connections between peers are secured with TLS, 674 DTLS, or potentially some to be defined future protocol. 675 Message Level: Each RELOAD message must be signed. 676 Object Level: Stored objects must be signed by the creating peer. 678 These three levels of security work together to allow peers to verify 679 the origin and correctness of data they receive from other peers, 680 even in the face of malicious activity by other peers in the overlay. 681 RELOAD also provides access control built on top of these 682 communications security features. Because the peer responsible for 683 storing a piece of data can validate the signature on the data being 684 stored, the responsible peer can determine whether a given operation 685 is permitted or not. 687 RELOAD also provides an optional shared secret based admission 688 control feature using shared secrets and TLS-PSK. In order to form a 689 TLS connection to any node in the overlay, a new node needs to know 690 the shared overlay key, thus restricting access to authorized users 691 only. This feature is used together with certificate-based access 692 control, not as a replacement for it. It is typically used when 693 self-signed certificates are being used but would generally not be 694 used when the certificates were all signed by an enrollment server. 696 1.4. Structure of This Document 698 The remainder of this document is structured as follows. 700 o Section 2 provides definitions of terms used in this document. 701 o Section 3 provides an overview of the mechanisms used to establish 702 and maintain the overlay. 703 o Section 4 provides an overview of the mechanism RELOAD provides to 704 support other applications. 705 o Section 5 defines the protocol messages that RELOAD uses to 706 establish and maintain the overlay. 707 o Section 6 defines the protocol messages that are used to store and 708 retrieve data using RELOAD. 709 o Section 7 defines the Certificate Store Usage that is fundamental 710 to RELOAD security. 711 o Section 8 defines the TURN Server Usage needed to locate TURN 712 servers for NAT traversal. 713 o Section 9 defines a specific Topology Plugin using Chord. 714 o Section 10 defines the mechanisms that new RELOAD nodes use to 715 join the overlay for the first time. 716 o Section 11 provides an extended example. 718 2. Terminology 720 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 721 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 722 document are to be interpreted as described in RFC 2119 [RFC2119]. 724 We use the terminology and definitions from the Concepts and 725 Terminology for Peer to Peer SIP [I-D.ietf-p2psip-concepts] draft 726 extensively in this document. Other terms used in this document are 727 defined inline when used and are also defined below for reference. 729 DHT: A distributed hash table. A DHT is an abstract hash table 730 service realized by storing the contents of the hash table across 731 a set of peers. 733 Overlay Algorithm: An overlay algorithm defines the rules for 734 determining which peers in an overlay store a particular piece of 735 data and for determining a topology of interconnections amongst 736 peers in order to find a piece of data. 738 Overlay Instance: A specific overlay algorithm and the collection of 739 peers that are collaborating to provide read and write access to 740 it. There can be any number of overlay instances running in an IP 741 network at a time, and each operates in isolation of the others. 743 Peer: A host that is participating in the overlay. Peers are 744 responsible for holding some portion of the data that has been 745 stored in the overlay and also route messages on behalf of other 746 hosts as required by the Overlay Algorithm. 748 Client: A host that is able to store data in and retrieve data from 749 the overlay but which is not participating in routing or data 750 storage for the overlay. 752 Kind: A kind defines a particular type of data that can be stored in 753 the overlay. Applications define new Kinds to story the data they 754 use. Each Kind is identified with a unique integer called a 755 Kind-ID. 757 Node: We use the term "Node" to refer to a host that may be either a 758 Peer or a Client. Because RELOAD uses the same protocol for both 759 clients and peers, much of the text applies equally to both. 760 Therefore we use "Node" when the text applies to both Clients and 761 Peers and the more specific term (i.e. client or peer) when the 762 text applies only to Clients or only to Peers. 764 Node-ID: A fixed-length value that uniquely identifies a node. 765 Node-IDs of all 0s and all 1s are reserved and are invalid Node- 766 IDs. A value of zero is not used in the wire protocol but can be 767 used to indicate an invalid node in implementations and APIs. The 768 Node-ID of all 1s is used on the wire protocol as a wildcard. 770 Resource: An object or group of objects associated with a string 771 identifier. See "Resource Name" below. 773 Resource Name: The potentially human readable name by which a 774 resource is identified. In unstructured P2P networks, the 775 resource name is sometimes used directly as a Resource-ID. In 776 structured P2P networks the resource name is typically mapped into 777 a Resource-ID by using the string as the input to hash function. 778 A SIP resource, for example, is often identified by its AOR which 779 is an example of a Resource Name. 781 Resource-ID: A value that identifies some resources and which is 782 used as a key for storing and retrieving the resource. Often this 783 is not human friendly/readable. One way to generate a Resource-ID 784 is by applying a mapping function to some other unique name (e.g., 785 user name or service name) for the resource. The Resource-ID is 786 used by the distributed database algorithm to determine the peer 787 or peers that are responsible for storing the data for the 788 overlay. In structured P2P networks, Resource-IDs are generally 789 fixed length and are formed by hashing the resource name. In 790 unstructured networks, resource names may be used directly as 791 Resource-IDs and may be variable lengths. 793 Connection Table: The set of nodes to which a node is directly 794 connected. This includes nodes with which Attach handshakes have 795 been done but which have not sent any Updates. 797 Routing Table: The set of peers which a node can use to route 798 overlay messages. In general, these peers will all be on the 799 connection table but not vice versa, because some peers will have 800 Attached but not sent updates. Peers may send messages directly 801 to peers that are in the connection table but may only route 802 messages to other peers through peers that are in the routing 803 table. 805 Destination List: A list of IDs through which a message is to be 806 routed. A single Node-ID is a trivial form of destination list. 808 Usage: A usage is an application that wishes to use the overlay for 809 some purpose. Each application wishing to use the overlay defines 810 a set of data kinds that it wishes to use. The SIP usage defines 811 the location data kind. 813 The term "maximum request lifetime" is the maximum time a request 814 will wait for a response; it defaults to 15 seconds. The term 815 "successor replacement hold-down time" is the amount of time to wait 816 before starting replication when a new successor is found; it 817 defaults to 30 seconds. 819 3. Overlay Management Overview 821 The most basic function of RELOAD is as a generic overlay network. 822 Nodes need to be able to join the overlay, form connections to other 823 nodes, and route messages through the overlay to nodes to which they 824 are not directly connected. This section provides an overview of the 825 mechanisms that perform these functions. 827 3.1. Security and Identification 829 Every node in the RELOAD overlay is identified by a Node-ID. The 830 Node-ID is used for three major purposes: 832 o To address the node itself. 833 o To determine its position in the overlay topology when the overlay 834 is structured. 835 o To determine the set of resources for which the node is 836 responsible. 838 Each node has a certificate [RFC5280] containing a Node-ID, which is 839 unique within an overlay instance. 841 The certificate serves multiple purposes: 843 o It entitles the user to store data at specific locations in the 844 Overlay Instance. Each data kind defines the specific rules for 845 determining which certificates can access each Resource-ID/Kind-ID 846 pair. For instance, some kinds might allow anyone to write at a 847 given location, whereas others might restrict writes to the owner 848 of a single certificate. 849 o It entitles the user to operate a node that has a Node-ID found in 850 the certificate. When the node forms a connection to another 851 peer, it uses this certificate so that a node connecting to it 852 knows it is connected to the correct node (technically: a (D)TLS 853 association with client authentication is formed.) In addition, 854 the node can sign messages, thus providing integrity and 855 authentication for messages which are sent from the node. 856 o It entitles the user to use the user name found in the 857 certificate. 859 If a user has more than one device, typically they would get one 860 certificate for each device. This allows each device to act as a 861 separate peer. 863 RELOAD supports multiple certificate issuance models. The first is 864 based on a central enrollment process which allocates a unique name 865 and Node-ID and puts them in a certificate for the user. All peers 866 in a particular Overlay Instance have the enrollment server as a 867 trust anchor and so can verify any other peer's certificate. 869 In some settings, a group of users want to set up an overlay network 870 but are not concerned about attack by other users in the network. 871 For instance, users on a LAN might want to set up a short term ad hoc 872 network without going to the trouble of setting up an enrollment 873 server. RELOAD supports the use of self-generated, self-signed 874 certificates. When self-signed certificates are used, the node also 875 generates its own Node-ID and username. The Node-ID is computed as a 876 digest of the public key, to prevent Node-ID theft; however this 877 model is still subject to a number of known attacks (most notably 878 Sybil attacks [Sybil]) and can only be safely used in closed networks 879 where users are mutually trusting. 881 The general principle here is that the security mechanisms (TLS and 882 message signatures) are always used, even if the certificates are 883 self-signed. This allows for a single set of code paths in the 884 systems with the only difference being whether certificate 885 verification is required to chain to a single root of trust. 887 3.1.1. Shared-Key Security 889 RELOAD also provides an admission control system based on shared 890 keys. In this model, the peers all share a single key which is used 891 to authenticate the peer-to-peer connections via TLS-PSK/TLS-SRP. 893 3.2. Clients 895 RELOAD defines a single protocol that is used both as the peer 896 protocol and as the client protocol for the overlay. This simplifies 897 implementation, particularly for devices that may act in either role, 898 and allows clients to inject messages directly into the overlay. 900 We use the term "peer" to identify a node in the overlay that routes 901 messages for nodes other than those to which it is directly 902 connected. Peers typically also have storage responsibilities. We 903 use the term "client" to refer to nodes that do not have routing or 904 storage responsibilities. When text applies to both peers and 905 clients, we will simply refer such devices as "nodes." 907 RELOAD's client support allows nodes that are not participating in 908 the overlay as peers to utilize the same implementation and to 909 benefit from the same security mechanisms as the peers. Clients 910 possess and use certificates that authorize the user to store data at 911 certain locations in the overlay. The Node-ID in the certificate is 912 used to identify the particular client as a member of the overlay and 913 to authenticate its messages. 915 In RELOAD, unlike some other designs, clients are not a first-class 916 concept. From the perspective of a peer, a client is simply a node 917 which has not yet sent any Updates or Joins. It might never do so 918 (if it's a client) or it might eventually do so (if it's just a node 919 that's taking a long time to join). The routing and storage rules 920 for RELOAD provide for correct behavior by peers regardless of 921 whether other nodes attached to them are clients or peers. Of 922 course, a client implementation must know that it intends to be a 923 client, but this localizes complexity only to that node. 925 For more discussion of the motivation for RELOAD's client support, 926 see Appendix C. 928 3.2.1. Client Routing 930 Clients may insert themselves in the overlay in two ways: 932 o Establish a connection to the peer responsible for the client's 933 Node-ID in the overlay. Then requests may be sent from/to the 934 client using its Node-ID in the same manner as if it were a peer, 935 because the responsible peer in the overlay will handle the final 936 step of routing to the client. This may require a TURN relay in 937 cases where NATs or firewalls prevent a client from forming a 938 direct connections with its responsible peer. Note that clients 939 that choose this option MUST process Update messages from the 940 peer. Those updates can indicate that the peer no longer is 941 responsible for the Client's Node-ID. The client then MUST form a 942 connection to the appropriate peer. Failure to do so will result 943 in the client no longer receiving messages. 944 o Establish a connection with an arbitrary peer in the overlay 945 (perhaps based on network proximity or an inability to establish a 946 direct connection with the responsible peer). In this case, the 947 client will rely on RELOAD's Destination List feature to ensure 948 reachability. The client can initiate requests, and any node in 949 the overlay that knows the Destination List to its current 950 location can reach it, but the client is not directly reachable 951 using only its Node-ID. If the client is to receive incoming 952 requests from other members of the overlay, the Destination List 953 required to reach it must be learnable via other mechanisms, such 954 as being stored in the overlay by a usage. 956 3.2.2. Minimum Functionality Requirements for Clients 958 A node may act as a client simply because it does not have the 959 resources or even an implementation of the topology plugin required 960 to act as a peer in the overlay. In order to exchange RELOAD 961 messages with a peer, a client must meet a minimum level of 962 functionality. Such a client must: 964 o Implement RELOAD's connection-management operations that are used 965 to establish the connection with the peer. 966 o Implement RELOAD's data retrieval methods (with client 967 functionality). 968 o Be able to calculate Resource-IDs used by the overlay. 969 o Possess security credentials required by the overlay it is 970 implementing. 972 A client speaks the same protocol as the peers, knows how to 973 calculate Resource-IDs, and signs its requests in the same manner as 974 peers. While a client does not necessarily require a full 975 implementation of the overlay algorithm, calculating the Resource-ID 976 requires an implementation of the appropriate algorithm for the 977 overlay. 979 3.3. Routing 981 This section will discuss the requirements RELOAD's routing 982 capabilities must meet, then describe the routing features in the 983 protocol, and then provide a brief overview of how they are used. 984 Appendix B discusses some alternative designs and the tradeoffs that 985 would be necessary to support them. 987 RELOAD's routing capabilities must meet the following requirements: 989 NAT Traversal: RELOAD must support establishing and using 990 connections between nodes separated by one or more NATs, including 991 locating peers behind NATs for those overlays allowing/requiring 992 it. 993 Clients: RELOAD must support requests from and to clients that do 994 not participate in overlay routing. 995 Client promotion: RELOAD must support clients that become peers at a 996 later point as determined by the overlay algorithm and deployment. 997 Low state: RELOAD's routing algorithms must not require 998 significant state to be stored on intermediate peers. 999 Return routability in unstable topologies: At some points in 1000 times, different nodes may have inconsistent information about the 1001 connectivity of the routing graph. In all cases, the response to 1002 a request needs to delivered to the node that sent the request and 1003 not to some other node. 1005 RELOAD's routing provides three mechanisms designed to assist in 1006 meeting these needs: 1008 Destination Lists: While in principle it is possible to just 1009 inject a message into the overlay with a bare Node-ID as the 1010 destination, RELOAD provides a source routing capability in the 1011 form of "Destination Lists". A "Destination List provides a list 1012 of the nodes through which a message must flow. 1013 Via Lists: In order to allow responses to follow the same path as 1014 requests, each message also contains a "Via List", which is added 1015 to by each node a message traverses. This via list can then be 1016 inverted and used as a destination list for the response. 1017 RouteQuery: The RouteQuery method allows a node to query a peer 1018 for the next hop it will use to route a message. This method is 1019 useful for diagnostics and for iterative routing. 1021 The basic routing mechanism used by RELOAD is Symmetric Recursive. 1022 We will first describe symmetric recursive routing and then discuss 1023 its advantages in terms of the requirements discussed above. 1025 Symmetric recursive routing requires that a message follow a path 1026 through the overlay to the destination without returning to the 1027 originating node: each peer forwards the message closer to its 1028 destination. The return path of the response is then the same path 1029 followed in reverse. For example, a message following a route from A 1030 to Z through B and X: 1032 A B X Z 1033 ------------------------------- 1035 ----------> 1036 Dest=Z 1037 ----------> 1038 Via=A 1039 Dest=Z 1040 ----------> 1041 Via=A, B 1042 Dest=Z 1044 <---------- 1045 Dest=X, B, A 1046 <---------- 1047 Dest=B, A 1048 <---------- 1049 Dest=A 1051 Note that the preceding Figure does not indicate whether A is a 1052 client or peer: A forwards its request to B and the response is 1053 returned to A in the same manner regardless of A's role in the 1054 overlay. 1056 This figure shows use of full via-lists by intermediate peers B and 1057 X. However, if B and/or X are willing to store state, then they may 1058 elect to truncate the lists, save that information internally (keyed 1059 by the transaction id), and return the response message along the 1060 path from which it was received when the response is received. This 1061 option requires greater state to be stored on intermediate peers but 1062 saves a small amount of bandwidth and reduces the need for modifying 1063 the message en route. Selection of this mode of operation is a 1064 choice for the individual peer; the techniques are interoperable even 1065 on a single message. The figure below shows B using full via lists 1066 but X truncating them to X1 and saving the state internally. 1068 A B X Z 1069 ------------------------------- 1071 ----------> 1072 Dest=Z 1073 ----------> 1074 Via=A 1075 Dest=Z 1076 ----------> 1077 Dest=Z, X1 1079 <---------- 1080 Dest=X,X1 1081 <---------- 1082 Dest=B, A 1083 <---------- 1084 Dest=A 1086 RELOAD also supports a basic Iterative routing mode (where the 1087 intermediate peers merely return a response indicating the next hop, 1088 but do not actually forward the message to that next hop themselves). 1089 Iterative routing is implemented using the RouteQuery method, which 1090 requests this behavior. Note that iterative routing is selected only 1091 by the initiating node. 1093 3.4. Connectivity Management 1095 In order to provide efficient routing, a peer needs to maintain a set 1096 of direct connections to other peers in the Overlay Instance. Due to 1097 the presence of NATs, these connections often cannot be formed 1098 directly. Instead, we use the Attach request to establish a 1099 connection. Attach uses ICE [RFC5245] to establish the connection. 1100 It is assumed that the reader is familiar with ICE. 1102 Say that peer A wishes to form a direct connection to peer B. It 1103 gathers ICE candidates and packages them up in an Attach request 1104 which it sends to B through usual overlay routing procedures. B does 1105 its own candidate gathering and sends back a response with its 1106 candidates. A and B then do ICE connectivity checks on the candidate 1107 pairs. The result is a connection between A and B. At this point, A 1108 and B can add each other to their routing tables and send messages 1109 directly between themselves without going through other overlay 1110 peers. 1112 There is one special case in which Attach cannot be used: when a 1113 peer is joining the overlay and is not connected to any peers. In 1114 order to support this case, some small number of "bootstrap nodes" 1115 typically need to be publicly accessible so that new peers can 1116 directly connect to them. Section 10 contains more detail on this. 1118 In general, a peer needs to maintain connections to all of the peers 1119 near it in the Overlay Instance and to enough other peers to have 1120 efficient routing (the details depend on the specific overlay). If a 1121 peer cannot form a connection to some other peer, this isn't 1122 necessarily a disaster; overlays can route correctly even without 1123 fully connected links. However, a peer should try to maintain the 1124 specified link set and if it detects that it has fewer direct 1125 connections, should form more as required. This also implies that 1126 peers need to periodically verify that the connected peers are still 1127 alive and if not try to reform the connection or form an alternate 1128 one. 1130 3.5. Overlay Algorithm Support 1132 The Topology Plugin allows RELOAD to support a variety of overlay 1133 algorithms. This specification defines a DHT based on Chord [Chord], 1134 which is mandatory to implement, but the base RELOAD protocol is 1135 designed to support a variety of overlay algorithms. 1137 3.5.1. Support for Pluggable Overlay Algorithms 1139 RELOAD defines three methods for overlay maintenance: Join, Update, 1140 and Leave. However, the contents of those messages, when they are 1141 sent, and their precise semantics are specified by the actual overlay 1142 algorithm; RELOAD merely provides a framework of commonly-needed 1143 methods that provides uniformity of notation (and ease of debugging) 1144 for a variety of overlay algorithms. 1146 3.5.2. Joining, Leaving, and Maintenance Overview 1148 When a new peer wishes to join the Overlay Instance, it must have a 1149 Node-ID that it is allowed to use and a set of credentials which 1150 match that Node-ID. When an enrollment server is used that Node-ID 1151 will be in the certificate the node received from the enrollment 1152 server. The details of the joining procedure are defined by the 1153 overlay algorithm, but the general steps for joining an Overlay 1154 Instance are: 1156 o Forming connections to some other peers. 1157 o Acquiring the data values this peer is responsible for storing. 1158 o Informing the other peers which were previously responsible for 1159 that data that this peer has taken over responsibility. 1161 The first thing the peer needs to do is to form a connection to some 1162 "bootstrap node". Because this is the first connection the peer 1163 makes, these nodes must have public IP addresses so that they can be 1164 connected to directly. Once a peer has connected to one or more 1165 bootstrap nodes, it can form connections in the usual way by routing 1166 Attach messages through the overlay to other nodes. Once a peer has 1167 connected to the overlay for the first time, it can cache the set of 1168 nodes it has connected to with public IP addresses for use as future 1169 bootstrap nodes. 1171 Once a peer has connected to a bootstrap node, it then needs to take 1172 up its appropriate place in the overlay. This requires two major 1173 operations: 1175 o Forming connections to other peers in the overlay to populate its 1176 Routing Table. 1177 o Getting a copy of the data it is now responsible for storing and 1178 assuming responsibility for that data. 1180 The second operation is performed by contacting the Admitting Peer 1181 (AP), the node which is currently responsible for that section of the 1182 overlay. 1184 The details of this operation depend mostly on the overlay algorithm 1185 involved, but a typical case would be: 1187 1. JP (Joining Peer) sends a Join request to AP (Admitting Peer) 1188 announcing its intention to join. 1189 2. AP sends a Join response. 1190 3. AP does a sequence of Stores to JP to give it the data it will 1191 need. 1192 4. AP does Updates to JP and to other peers to tell it about its own 1193 routing table. At this point, both JP and AP consider JP 1194 responsible for some section of the Overlay Instance. 1195 5. JP makes its own connections to the appropriate peers in the 1196 Overlay Instance. 1198 After this process is completed, JP is a full member of the Overlay 1199 Instance and can process Store/Fetch requests. 1201 Note that the first node is a special case. When ordinary nodes 1202 cannot form connections to the bootstrap nodes, then they are not 1203 part of the overlay. However, the first node in the overlay can 1204 obviously not connect to other nodes. In order to support this case, 1205 potential first nodes (which must also serve as bootstrap nodes 1206 initially) must somehow be instructed (perhaps by configuration 1207 settings) that they are the entire overlay, rather than not part of 1208 it. 1210 Note that clients do not perform either of these operations. 1212 3.6. First-Time Setup 1214 Previous sections addressed how RELOAD works once a node has 1215 connected. This section provides an overview of how users get 1216 connected to the overlay for the first time. RELOAD is designed so 1217 that users can start with the name of the overlay they wish to join 1218 and perhaps a username and password, and leverage that into having a 1219 working peer with minimal user intervention. This helps avoid the 1220 problems that have been experienced with conventional SIP clients 1221 where users are required to manually configure a large number of 1222 settings. 1224 3.6.1. Initial Configuration 1226 In the first phase of the process, the user starts out with the name 1227 of the overlay and uses this to download an initial set of overlay 1228 configuration parameters. The node does a DNS SRV lookup on the 1229 overlay name to get the address of a configuration server. It can 1230 then connect to this server with HTTPS to download a configuration 1231 document which contains the basic overlay configuration parameters as 1232 well as a set of bootstrap nodes which can be used to join the 1233 overlay. 1235 If a node already has the valid configuration document that it 1236 received by some out of band method, this step can be skipped. 1238 3.6.2. Enrollment 1240 If the overlay is using centralized enrollment, then a user needs to 1241 acquire a certificate before joining the overlay. The certificate 1242 attests both to the user's name within the overlay and to the Node- 1243 IDs which they are permitted to operate. In that case, the 1244 configuration document will contain the address of an enrollment 1245 server which can be used to obtain such a certificate. The 1246 enrollment server may (and probably will) require some sort of 1247 username and password before issuing the certificate. The enrollment 1248 server's ability to restrict attackers' access to certificates in the 1249 overlay is one of the cornerstones of RELOAD's security. 1251 4. Application Support Overview 1253 RELOAD is not intended to be used alone, but rather as a substrate 1254 for other applications. These applications can use RELOAD for a 1255 variety of purposes: 1257 o To store data in the overlay and retrieve data stored by other 1258 nodes. 1259 o As a discovery mechanism for services such as TURN. 1260 o To form direct connections which can be used to transmit 1261 application-level messages without using the overlay. 1263 This section provides an overview of these services. 1265 4.1. Data Storage 1267 RELOAD provides operations to Store and Fetch data. Each location in 1268 the Overlay Instance is referenced by a Resource-ID. However, each 1269 location may contain data elements corresponding to multiple kinds 1270 (e.g., certificate, SIP registration). Similarly, there may be 1271 multiple elements of a given kind, as shown below: 1273 +--------------------------------+ 1274 | Resource-ID | 1275 | | 1276 | +------------+ +------------+ | 1277 | | Kind 1 | | Kind 2 | | 1278 | | | | | | 1279 | | +--------+ | | +--------+ | | 1280 | | | Value | | | | Value | | | 1281 | | +--------+ | | +--------+ | | 1282 | | | | | | 1283 | | +--------+ | | +--------+ | | 1284 | | | Value | | | | Value | | | 1285 | | +--------+ | | +--------+ | | 1286 | | | +------------+ | 1287 | | +--------+ | | 1288 | | | Value | | | 1289 | | +--------+ | | 1290 | +------------+ | 1291 +--------------------------------+ 1293 Each kind is identified by a Kind-ID, which is a code point either 1294 assigned by IANA or allocated out of a private range. As part of the 1295 kind definition, protocol designers may define constraints, such as 1296 limits on size, on the values which may be stored. For many kinds, 1297 the set may be restricted to a single value; some sets may be allowed 1298 to contain multiple identical items while others may only have unique 1299 items. Note that a kind may be employed by multiple usages and new 1300 usages are encouraged to use previously defined kinds where possible. 1301 We define the following data models in this document, though other 1302 usages can define their own structures: 1304 single value: There can be at most one item in the set and any value 1305 overwrites the previous item. 1307 array: Many values can be stored and addressed by a numeric index. 1309 dictionary: The values stored are indexed by a key. Often this key 1310 is one of the values from the certificate of the peer sending the 1311 Store request. 1313 In order to protect stored data from tampering, by other nodes, each 1314 stored value is digitally signed by the node which created it. When 1315 a value is retrieved, the digital signature can be verified to detect 1316 tampering. 1318 4.1.1. Storage Permissions 1320 A major issue in peer-to-peer storage networks is minimizing the 1321 burden of becoming a peer, and in particular minimizing the amount of 1322 data which any peer is required to store for other nodes. RELOAD 1323 addresses this issue by only allowing any given node to store data at 1324 a small number of locations in the overlay, with those locations 1325 being determined by the node's certificate. When a peer uses a Store 1326 request to place data at a location authorized by its certificate, it 1327 signs that data with the private key that corresponds to its 1328 certificate. Then the peer responsible for storing the data is able 1329 to verify that the peer issuing the request is authorized to make 1330 that request. Each data kind defines the exact rules for determining 1331 what certificate is appropriate. 1333 The most natural rule is that a certificate authorizes a user to 1334 store data keyed with their user name X. This rule is used for all 1335 the kinds defined in this specification. Thus, only a user with a 1336 certificate for "alice@example.org" could write to that location in 1337 the overlay. However, other usages can define any rules they choose, 1338 including publicly writable values. 1340 The digital signature over the data serves two purposes. First, it 1341 allows the peer responsible for storing the data to verify that this 1342 Store is authorized. Second, it provides integrity for the data. 1344 The signature is saved along with the data value (or values) so that 1345 any reader can verify the integrity of the data. Of course, the 1346 responsible peer can "lose" the value but it cannot undetectably 1347 modify it. 1349 The size requirements of the data being stored in the overlay are 1350 variable. For instance, a SIP AOR and voicemail differ widely in the 1351 storage size. RELOAD leaves it to the Usage and overlay 1352 configuration to limit size imbalance of various kinds. 1354 4.1.2. Replication 1356 Replication in P2P overlays can be used to provide: 1358 persistence: if the responsible peer crashes and/or if the storing 1359 peer leaves the overlay 1360 security: to guard against DoS attacks by the responsible peer or 1361 routing attacks to that responsible peer 1362 load balancing: to balance the load of queries for popular 1363 resources. 1365 A variety of schemes are used in P2P overlays to achieve some of 1366 these goals. Common techniques include replicating on neighbors of 1367 the responsible peer, randomly locating replicas around the overlay, 1368 or replicating along the path to the responsible peer. 1370 The core RELOAD specification does not specify a particular 1371 replication strategy. Instead, the first level of replication 1372 strategies are determined by the overlay algorithm, which can base 1373 the replication strategy on its particular topology. For example, 1374 Chord places replicas on successor peers, which will take over 1375 responsibility should the responsible peer fail [Chord]. 1377 If additional replication is needed, for example if data persistence 1378 is particularly important for a particular usage, then that usage may 1379 specify additional replication, such as implementing random 1380 replications by inserting a different well known constant into the 1381 Resource Name used to store each replicated copy of the resource. 1382 Such replication strategies can be added independent of the 1383 underlying algorithm, and their usage can be determined based on the 1384 needs of the particular usage. 1386 4.2. Usages 1388 By itself, the distributed storage layer just provides infrastructure 1389 on which applications are built. In order to do anything useful, a 1390 usage must be defined. Each Usage needs to specify several things: 1392 o Registers Kind-ID code points for any kinds that the Usage 1393 defines. 1394 o Defines the data structure for each of the kinds. 1395 o Defines access control rules for each of the kinds. 1396 o Defines how the Resource Name is formed that is hashed to form the 1397 Resource-ID where each kind is stored. 1398 o Describes how values will be merged after a network partition. 1399 Unless otherwise specified, the default merging rule is to act as 1400 if all the values that need to be merged were stored and as if the 1401 order they were stored in corresponds to the stored time values 1402 associated with (and carried in) their values. Because the stored 1403 time values are those associated with the peer which did the 1404 writing, clock skew is generally not an issue. If two nodes are 1405 on different partitions, write to the same location, and have 1406 clock skew, this can create merge conflicts. However because 1407 RELOAD deliberately segregates storage so that data from different 1408 users and peers is stored in different locations, and a single 1409 peer will typically only be in a single network partition, this 1410 case will generally not arise. 1412 The kinds defined by a usage may also be applied to other usages. 1413 However, a need for different parameters, such as different size 1414 limits, would imply the need to create a new kind. 1416 4.3. Service Discovery 1418 RELOAD does not currently define a generic service discovery 1419 algorithm as part of the base protocol, although a simplistic TURN- 1420 specific discovery mechanism is provided. A variety of service 1421 discovery algorithms can be implemented as extensions to the base 1422 protocol, such as the service discovery algorithm ReDIR 1423 [opendht-sigcomm05] or [I-D.maenpaa-p2psip-service-discovery]. 1425 4.4. Application Connectivity 1427 There is no requirement that a RELOAD usage must use RELOAD's 1428 primitives for establishing its own communication if it already 1429 possesses its own means of establishing connections. For example, 1430 one could design a RELOAD-based resource discovery protocol which 1431 used HTTP to retrieve the actual data. 1433 For more common situations, however, it is the overlay itself - 1434 rather than an external authority such as DNS - which is used to 1435 establish a connection. RELOAD provides connectivity to applications 1436 using the AppAttach method. For example, if a P2PSIP node wishes to 1437 establish a SIP dialog with another P2PSIP node, it will use 1438 AppAttach to establish a direct connection with the other node. This 1439 new connection is separate from the peer protocol connection. It is 1440 a dedicated UDP or TCP flow used only for the SIP dialog. 1442 5. Overlay Management Protocol 1444 This section defines the basic protocols used to create, maintain, 1445 and use the RELOAD overlay network. We start by defining the basic 1446 concept of how message destinations are interpreted when routing 1447 messages. We then describe the symmetric recursive routing model, 1448 which is RELOAD's default routing algorithm. We then define the 1449 message structure and then finally define the messages used to join 1450 and maintain the overlay. 1452 5.1. Message Receipt and Forwarding 1454 When a peer receives a message, it first examines the overlay, 1455 version, and other header fields to determine whether the message is 1456 one it can process. If any of these are incorrect (e.g., the message 1457 is for an overlay in which the peer does not participate) it is an 1458 error. The peer SHOULD generate an appropriate error but local 1459 policy can override this and cause the messages to be silently 1460 dropped. 1462 Once the peer has determined that the message is correctly formatted, 1463 it examines the first entry on the destination list. There are three 1464 possible cases here: 1466 o The first entry on the destination list is an ID for which the 1467 peer is responsible. 1468 o The first entry on the destination list is an ID for which another 1469 peer is responsible. 1470 o The first entry on the destination list is a private ID that is 1471 being used for destination list compression. This is described 1472 later (note that private IDs can be distinguished from Node-IDs 1473 and Resource-IDs on the wire; see Section 5.3.2.2). 1475 These cases are handled as discussed below. 1477 5.1.1. Responsible ID 1479 If the first entry on the destination list is an ID for which the 1480 node is responsible, there are several sub-cases to consider. 1482 o If the entry is a Resource-ID, then it MUST be the only entry on 1483 the destination list. If there are other entries, the message 1484 MUST be silently dropped. Otherwise, the message is destined for 1485 this node and it passes it up to the upper layers. 1487 o If the entry is a Node-ID which equals this node's Node-ID, then 1488 the message is destined for this node. If this is the only entry 1489 on the destination list, the message is destined for this node and 1490 is passed up to the upper layers. Otherwise the entry is removed 1491 from the destination list and the message is passed to the Message 1492 Transport. If the message is a response and there is state for 1493 the transaction ID, the state is reinserted into the destination 1494 list before the message is further processed. 1495 o If the entry is a Node-ID which is not equal to this node, then 1496 the node MUST drop the message silently unless the Node-ID 1497 corresponds to a node which is directly connected to this node 1498 (i.e., a client). In that case, it MUST forward the message to 1499 the destination node as described in the next section. 1501 Note that this implies that in order to address a message to "the 1502 peer that controls region X", a sender sends to Resource-ID X, not 1503 Node-ID X. 1505 5.1.2. Other ID 1507 If neither of the other three cases applies, then the peer MUST 1508 forward the message towards the first entry on the destination list. 1509 This means that it MUST select one of the peers to which it is 1510 connected and which is likely to be responsible for the first entry 1511 on the destination list. If the first entry on the destination list 1512 is in the peer's connection table, then it SHOULD forward the message 1513 to that peer directly. Otherwise, the peer consults the routing 1514 table to forward the message. 1516 Any intermediate peer which forwards a RELOAD request MUST arrange 1517 that if it receives a response to that message the response can be 1518 routed back through the set of nodes through which the request 1519 passed. This may be arranged in one of two ways: 1521 o The peer MAY add an entry to the via list in the forwarding header 1522 that will enable it to determine the correct node. 1523 o The peer MAY keep per-transaction state which will allow it to 1524 determine the correct node. 1526 As an example of the first strategy, if node D receives a message 1527 from node C with via list (A, B), then D would forward to the next 1528 node (E) with via list (A, B, C). Now, if E wants to respond to the 1529 message, it reverses the via list to produce the destination list, 1530 resulting in (D, C, B, A). When D forwards the response to C, the 1531 destination list will contain (C, B, A). 1533 As an example of the second strategy, if node D receives a message 1534 from node C with transaction ID X and via list (A, B), it could store 1535 (X, C) in its state database and forward the message with the via 1536 list unchanged. When D receives the response, it consults its state 1537 database for transaction id X, determines that the request came from 1538 C, and forwards the response to C. 1540 Intermediate peers which modify the via list are not required to 1541 simply add entries. The only requirement is that the peer be able to 1542 reconstruct the correct destination list on the return route. RELOAD 1543 provides explicit support for this functionality in the form of 1544 private IDs, which can replace any number of via list entries. For 1545 instance, in the above example, Node D might send E a via list 1546 containing only the private ID (I). E would then use the destination 1547 list (D, I) to send its return message. When D processes this 1548 destination list, it would detect that I is a private ID, recover the 1549 via list (A, B, C), and reverse that to produce the correct 1550 destination list (C, B, A) before sending it to C. This feature is 1551 called List Compression. It MAY either be a compressed version of 1552 the original via list or an index into a state database containing 1553 the original via list. 1555 No matter what mechanism for storing via list state is used, if an 1556 intermediate peer exits the overlay, then on the return trip the 1557 message cannot be forwarded and will be dropped. The ordinary 1558 timeout and retransmission mechanisms provide stability over this 1559 type of failure. 1561 Note that if an intermediate peer retains per-transaction state 1562 instead of modifying the via list, it needs some mechanism for timing 1563 out that state, otherwise its state database will grow without bound. 1564 Whatever algorithm is used, unless a FORWARD_CRITICAL forwarding 1565 option or overlay configuration option explicitly indicates this 1566 state is not needed, the state MUST be maintained for at least the 1567 value of the overlay reliability timer (3 seconds) and MAY be kept 1568 longer. Future extension, such as [I-D.jiang-p2psip-relay], may 1569 define mechanisms for determining when this state does not need to be 1570 retained. 1572 None of the above mechanisms are required for responses, since there 1573 is no need to ensure that subsequent requests follow the same path. 1575 5.1.3. Private ID 1577 If the first entry in the destination list is a private id (e.g., a 1578 compressed via list), the peer MUST replace that entry with the 1579 original via list that it replaced and then re-examine the 1580 destination list to determine which of the above cases now applies. 1582 5.2. Symmetric Recursive Routing 1584 This Section defines RELOAD's symmetric recursive routing algorithm, 1585 which is the default algorithm used by nodes to route messages 1586 through the overlay. All implementations MUST implement this routing 1587 algorithm. An overlay may be configured to use alternative routing 1588 algorithms, and alternative routing algorithms may be selected on a 1589 per-message basis. 1591 5.2.1. Request Origination 1593 In order to originate a message to a given Node-ID or Resource-ID, a 1594 node constructs an appropriate destination list. The simplest such 1595 destination list is a single entry containing the Node-ID or 1596 Resource-ID. The resulting message will use the normal overlay 1597 routing mechanisms to forward the message to that destination. The 1598 node can also construct a more complicated destination list for 1599 source routing. 1601 Once the message is constructed, the node sends the message to some 1602 adjacent peer. If the first entry on the destination list is 1603 directly connected, then the message MUST be routed down that 1604 connection. Otherwise, the topology plugin MUST be consulted to 1605 determine the appropriate next hop. 1607 Parallel searches for the resource are a common solution to improve 1608 reliability in the face of churn or of subversive peers. Parallel 1609 searches for usage-specified replicas are managed by the usage layer. 1610 However, a single request can also be routed through multiple 1611 adjacent peers, even when known to be sub-optimal, to improve 1612 reliability [vulnerabilities-acsac04]. Such parallel searches MAY be 1613 specified by the topology plugin. 1615 Because messages may be lost in transit through the overlay, RELOAD 1616 incorporates an end-to-end reliability mechanism. When an 1617 originating node transmits a request it MUST set a 3 second timer. 1618 If a response has not been received when the timer fires, the request 1619 is retransmitted with the same transaction identifier. The request 1620 MAY be retransmitted up to 4 times (for a total of 5 messages). 1621 After the timer for the fifth transmission fires, the message SHALL 1622 be considered to have failed. Note that this retransmission 1623 procedure is not followed by intermediate nodes. They follow the 1624 hop-by-hop reliability procedure described in Section 5.6.3. 1626 The above algorithm can result in multiple requests being delivered 1627 to a node. Receiving nodes MUST generate semantically equivalent 1628 responses to retransmissions of the same request (this can be 1629 determined by transaction id) if the request is received within the 1630 maximum request lifetime (15 seconds). For some requests (e.g., 1631 Fetch) this can be accomplished merely by processing the request 1632 again. For other requests, (e.g., Store) it may be necessary to 1633 maintain state for the duration of the request lifetime. 1635 5.2.2. Response Origination 1637 When a peer sends a response to a request using this routing 1638 algorithm, it MUST construct the destination list by reversing the 1639 order of the entries on the via list. This has the result that the 1640 response traverses the same peers as the request traversed, except in 1641 reverse order (symmetric routing). 1643 5.3. Message Structure 1645 RELOAD is a message-oriented request/response protocol. The messages 1646 are encoded using binary fields. All integers are represented in 1647 network byte order. The general philosophy behind the design was to 1648 use Type, Length, Value fields to allow for extensibility. However, 1649 for the parts of a structure that were required in all messages, we 1650 just define these in a fixed position, as adding a type and length 1651 for them is unnecessary and would simply increase bandwidth and 1652 introduces new potential for interoperability issues. 1654 Each message has three parts, concatenated as shown below: 1656 +-------------------------+ 1657 | Forwarding Header | 1658 +-------------------------+ 1659 | Message Contents | 1660 +-------------------------+ 1661 | Security Block | 1662 +-------------------------+ 1664 The contents of these parts are as follows: 1666 Forwarding Header: Each message has a generic header which is used 1667 to forward the message between peers and to its final destination. 1668 This header is the only information that an intermediate peer 1669 (i.e., one that is not the target of a message) needs to examine. 1671 Message Contents: The message being delivered between the peers. 1672 From the perspective of the forwarding layer, the contents are 1673 opaque, however, they are interpreted by the higher layers. 1675 Security Block: A security block containing certificates and a 1676 digital signature over the "Message Contents" section. Note that 1677 this signature can be computed without parsing the message 1678 contents. All messages MUST be signed by their originator. 1680 The following sections describe the format of each part of the 1681 message. 1683 5.3.1. Presentation Language 1685 The structures defined in this document are defined using a C-like 1686 syntax based on the presentation language used to define TLS. 1687 [RFC5246] Advantages of this style include: 1689 o It familiar enough looking that most readers can grasp it quickly. 1690 o The ability to define nested structures allows a separation 1691 between high-level and low-level message structures. 1692 o It has a straightforward wire encoding that allows quick 1693 implementation, but the structures can be comprehended without 1694 knowing the encoding. 1695 o The ability to mechanically compile encoders and decoders. 1697 Several idiosyncrasies of this language are worth noting. 1699 o All lengths are denoted in bytes, not objects. 1700 o Variable length values are denoted like arrays with angle 1701 brackets. 1702 o "select" is used to indicate variant structures. 1704 For instance, "uint16 array<0..2^8-2>;" represents up to 254 bytes 1705 but only up to 127 values of two bytes (16 bits) each. 1707 5.3.1.1. Common Definitions 1709 The following definitions are used throughout RELOAD and so are 1710 defined here. They also provide a convenient introduction to how to 1711 read the presentation language. 1713 An enum represents an enumerated type. The values associated with 1714 each possibility are represented in parentheses and the maximum value 1715 is represented as a nameless value, for purposes of describing the 1716 width of the containing integral type. For instance, Boolean 1717 represents a true or false: 1719 enum { false (0), true(1), (255)} Boolean; 1721 A boolean value is either a 1 or a 0. The max value of 255 indicates 1722 this is represented as a single byte on the wire. 1724 The NodeId, shown below, represents a single Node-ID. 1726 typedef opaque NodeId[NodeIdLength]; 1728 A NodeId is a fixed-length structure represented as a series of 1729 bytes, with the most significant byte first. The length is set on a 1730 per-overlay basis within the range of 16-20 bytes (128 to 160 bits). 1731 (See Section 10.1 for how NodeIdLength is set.) Note: the use of 1732 "typedef" here is an extension to the TLS language, but its meaning 1733 should be relatively obvious. Note the [ size ] syntax defines a 1734 fixed length element that does not include the length of the element 1735 in the on the wire encoding. 1737 A ResourceId, shown below, represents a single Resource-ID. 1739 typedef opaque ResourceId<0..2^8-1>; 1741 Like a NodeId, a ResourceId is an opaque string of bytes, but unlike 1742 NodeIds, ResourceIds are variable length, up to 254 bytes (2040 bits) 1743 in length. On the wire, each ResourceId is preceded by a single 1744 length byte (allowing lengths up to 255). Thus, the 3-byte value 1745 "FOO" would be encoded as: 03 46 4f 4f. Note the < range > syntax 1746 defines a variable length element that does include the length of the 1747 element in the on the wire encoding. The number of bytes to encode 1748 the length on the wire is derived by range; i.e., it is the minimum 1749 number of bytes which can encode the largest range value. 1751 A more complicated example is IpAddressPort, which represents a 1752 network address and can be used to carry either an IPv6 or IPv4 1753 address: 1755 enum {reservedAddr(0), ipv4_address (1), ipv6_address (2), 1756 (255)} AddressType; 1758 struct { 1759 uint32 addr; 1760 uint16 port; 1761 } IPv4AddrPort; 1763 struct { 1764 uint128 addr; 1765 uint16 port; 1766 } IPv6AddrPort; 1768 struct { 1769 AddressType type; 1770 uint8 length; 1772 select (type) { 1773 case ipv4_address: 1774 IPv4AddrPort v4addr_port; 1776 case ipv6_address: 1777 IPv6AddrPort v6addr_port; 1779 /* This structure can be extended */ 1780 }; 1781 } IpAddressPort; 1783 The first two fields in the structure are the same no matter what 1784 kind of address is being represented: 1786 type: the type of address (v4 or v6). 1787 length: the length of the rest of the structure. 1789 By having the type and the length appear at the beginning of the 1790 structure regardless of the kind of address being represented, an 1791 implementation which does not understand new address type X can still 1792 parse the IpAddressPort field and then discard it if it is not 1793 needed. 1795 The rest of the IpAddressPort structure is either an IPv4AddrPort or 1796 an IPv6AddrPort. Both of these simply consist of an address 1797 represented as an integer and a 16-bit port. As an example, here is 1798 the wire representation of the IPv4 address "192.0.2.1" with port 1799 "6100". 1801 01 ; type = IPv4 1802 06 ; length = 6 1803 c0 00 02 01 ; address = 192.0.2.1 1804 17 d4 ; port = 6100 1806 Unless a given structure that uses a select explicitly allows for 1807 unknown types in the select, any unknown type SHOULD be treated as an 1808 parsing error and the whole message discarded with no response. 1810 5.3.2. Forwarding Header 1812 The forwarding header is defined as a ForwardingHeader structure, as 1813 shown below. 1815 struct { 1816 uint32 relo_token; 1817 uint32 overlay; 1818 uint16 configuration_sequence; 1819 uint8 version; 1820 uint8 ttl; 1821 uint32 fragment; 1822 uint32 length; 1823 uint64 transaction_id; 1824 uint32 max_response_length; 1825 uint16 via_list_length; 1826 uint16 destination_list_length; 1827 uint16 options_length; 1828 Destination via_list[via_list_length]; 1829 Destination destination_list 1830 [destination_list_length]; 1831 ForwardingOptions options[options_length]; 1832 } ForwardingHeader; 1834 The contents of the structure are: 1836 relo_token: The first four bytes identify this message as a RELOAD 1837 message. This field MUST contain the value 0xd2454c4f (the string 1838 'RELO' with the high bit of the first byte set.). 1840 overlay: The 32 bit checksum/hash of the overlay being used. The 1841 variable length string representing the overlay name is hashed 1842 with SHA-1 [RFC3174] and the low order 32 bits are used. The 1843 purpose of this field is to allow nodes to participate in multiple 1844 overlays and to detect accidental misconfiguration. This is not a 1845 security critical function. 1847 configuration_sequence: The sequence number of the configuration 1848 file. 1850 version: The version of the RELOAD protocol being used. This is a 1851 fixed point integer between 0.1 and 25.4. This document describes 1852 version 0.1, with a value of 0x01. [[ Note to RFC Editor: Please 1853 update this to version 1.0 with value of 0x0a and remove this 1854 note. ]] 1856 ttl: An 8 bit field indicating the number of iterations, or hops, a 1857 message can experience before it is discarded. The TTL value MUST 1858 be decremented by one at every hop along the route the message 1859 traverses. If the TTL is 0, the message MUST NOT be propagated 1860 further and MUST be discarded, and a "Error_TTL_Exceeded" error 1861 should be generated. The initial value of the TTL SHOULD be 100 1862 unless defined otherwise by the overlay configuration. 1864 fragment: This field is used to handle fragmentation. The high 1865 order two bits are used to indicate the fragmentation status: If 1866 the high bit (0x80000000) is set, it indicates that the message is 1867 a fragment. If the next bit (0x40000000) is set, it indicates 1868 that this is the last fragment. The next six bits (0x20000000 to 1869 0x01000000) are reserved and SHOULD be set to zero. The remainder 1870 of the field is used to indicate the fragment offset; see 1871 Section 5.7 1873 length: The count in bytes of the size of the message, including the 1874 header. 1876 transaction_id: A unique 64 bit number that identifies this 1877 transaction and also allows receivers to disambiguate transactions 1878 which are otherwise identical. In order to provide a high 1879 probability that transaction IDs are unique, they MUST be randomly 1880 generated. Responses use the same Transaction ID as the request 1881 they correspond to. Transaction IDs are also used for fragment 1882 reassembly. 1884 max_response_length: The maximum size in bytes of a response. Used 1885 by requesting nodes to avoid receiving (unexpected) very large 1886 responses. If this value is non-zero, responding peers MUST check 1887 that any response would not exceed it and if so generate an 1888 Error_Response_Too_Large value. This value SHOULD be set to zero 1889 for responses. 1891 via_list_length: The length of the via list in bytes. Note that in 1892 this field and the following two length fields we depart from the 1893 usual variable-length convention of having the length immediately 1894 precede the value in order to make it easier for hardware decoding 1895 engines to quickly determine the length of the header. 1897 destination_list_length: The length of the destination list in 1898 bytes. 1900 options_length: The length of the header options in bytes. 1902 via_list: The via_list contains the sequence of destinations through 1903 which the message has passed. The via_list starts out empty and 1904 grows as the message traverses each peer. 1906 destination_list: The destination_list contains a sequence of 1907 destinations which the message should pass through. The 1908 destination list is constructed by the message originator. The 1909 first element in the destination list is where the message goes 1910 next. The list shrinks as the message traverses each listed peer. 1912 options: Contains a series of ForwardingOptions entries. See 1913 Section 5.3.2.3. 1915 5.3.2.1. Processing Configuration Sequence Numbers 1917 In order to be part of the overlay, a node MUST have a copy of the 1918 overlay configuration document. In order to allow for configuration 1919 document changes, each version of the configuration document has a 1920 sequence number which is monotonically increasing mod 65536. Because 1921 the sequence number may in principle wrap, greater than or less than 1922 are interpreted by modulo arithmetic as in TCP. 1924 When a destination node receives a request, it MUST check that the 1925 configuration_sequence field is equal to its own configuration 1926 sequence number. If they do not match, it MUST generate an error, 1927 either Error_Config_Too_Old or Error_Config_Too_New. In addition, if 1928 the configuration file in the request is too old, it MUST generate a 1929 ConfigUpdate message to update the requesting node. This allows new 1930 configuration documents to propagate quickly throughout the system. 1931 The one exception to this rule is that if the configuration_sequence 1932 field is equal to 0xffff, and the message type is ConfigUpdate, then 1933 the message MUST be accepted regardless of the receiving node's 1934 configuration sequence number. Since 65535 is a special value, peers 1935 sending a new configuration when the configuration sequence is 1936 currently 65534 MUST set the configuration sequence number to 0 when 1937 they send out a new configuration. 1939 5.3.2.2. Destination and Via Lists 1941 The destination list and via lists are sequences of Destination 1942 values: 1944 enum {reserved(0), node(1), resource(2), compressed(3), 1945 /* 128-255 not allowed */ (255) } 1946 DestinationType; 1948 select (destination_type) { 1949 case node: 1950 NodeId node_id; 1952 case resource: 1953 ResourceId resource_id; 1955 case compressed: 1956 opaque compressed_id<0..2^8-1>; 1958 /* This structure may be extended with new types */ 1959 } DestinationData; 1961 struct { 1962 DestinationType type; 1963 uint8 length; 1964 DestinationData destination_data; 1965 } Destination; 1967 struct { 1968 uint16 compressed_id; /* top bit MUST be 1 */ 1969 } Destination; 1971 If a destination structure has its first bit set to 1, then it is a 1972 16 bit integer. If the first bit is not set, then it is a structure 1973 starting with DestinationType. If it is a 16 bit integer, it is 1974 treated as if it were a full structure with a DestinationType of 1975 compressed and a compressed_id that was 2 bytes long with the value 1976 of the 16 bit integer. When the destination structure is not a 16 1977 bit integer, it is the TLV structure with the following contents: 1979 type 1980 The type of the DestinationData Payload Data Unit (PDU). This may 1981 be one of "node", "resource", or "compressed". 1983 length 1984 The length of the destination_data. 1986 destination_data 1987 The destination value itself, which is an encoded DestinationData 1988 structure, depending on the value of "type". 1990 Note: This structure encodes a type, length, value. The length 1991 field specifies the length of the DestinationData values, which 1992 allows the addition of new DestinationTypes. This allows an 1993 implementation which does not understand a given DestinationType 1994 to skip over it. 1996 A DestinationData can be one of three types: 1998 node 1999 A Node-ID. 2001 compressed 2002 A compressed list of Node-IDs and/or resources. Because this 2003 value was compressed by one of the peers, it is only meaningful to 2004 that peer and cannot be decoded by other peers. Thus, it is 2005 represented as an opaque string. 2007 resource 2008 The Resource-ID of the resource which is desired. This type MUST 2009 only appear in the final location of a destination list and MUST 2010 NOT appear in a via list. It is meaningless to try to route 2011 through a resource. 2013 One possible encoding of the 16 bit integer version as an opaque 2014 identifier is to encode an index into a connection table. To avoid 2015 misrouting responses in the event a response is delayed and the 2016 connection table entry has changed, the identifier SHOULD be split 2017 between an index and a generation counter for that index. At 2018 startup, the generation counters should be initialized to random 2019 values. An implementation could use 12 bits for the connection table 2020 index and 3 bits for the generation counter. (Note that this does 2021 not suggest a 4096 entry connection table for every node, only the 2022 ability to encode for a larger connection table.) When a connection 2023 table slot is used for a new connection, the generation counter is 2024 incremented (with wrapping). Connection table slots are used on a 2025 rotating basis to maximize the time interval between uses of the same 2026 slot for different connections. When routing a message to an entry 2027 in the destination list encoding a connection table entry, the node 2028 confirms that the generation counter matches the current generation 2029 counter of that index before forwarding the message. If it does not 2030 match, the message is silently dropped. 2032 5.3.2.3. Forwarding Options 2034 The Forwarding header can be extended with forwarding header options, 2035 which are a series of ForwardingOptions structures: 2037 enum { reservedForwarding(0), (255) } 2038 ForwardingOptionsType; 2040 struct { 2041 ForwardingOptionsType type; 2042 uint8 flags; 2043 uint16 length; 2044 select (type) { 2045 /* This type may be extended */ 2046 } option; 2047 } ForwardingOption; 2049 Each ForwardingOption consists of the following values: 2051 type 2052 The type of the option. This structure allows for unknown options 2053 types. 2055 length 2056 The length of the rest of the structure. 2058 flags 2059 Three flags are defined FORWARD_CRITICAL(0x01), 2060 DESTINATION_CRITICAL(0x02), and RESPONSE_COPY(0x04). These flags 2061 MUST NOT be set in a response. If the FORWARD_CRITICAL flag is 2062 set, any node that would forward the message but does not 2063 understand this options MUST reject the request with an 2064 Error_Unsupported_Forwarding_Option error response. If the 2065 DESTINATION_CRITICAL flag is set, any node that generates a 2066 response to the message but does not understand the forwarding 2067 option MUST reject the request with an 2068 Error_Unsupported_Forwarding_Option error response. If the 2069 RESPONSE_COPY flag is set, any node generating a response MUST 2070 copy the option from the request to the response except that the 2071 RESPONSE_COPY, FORWARD_CRITICAL and DESTINATION_CRITICAL flags 2072 must be cleared. 2074 option 2075 The option value. 2077 5.3.3. Message Contents Format 2079 The second major part of a RELOAD message is the contents part, which 2080 is defined by MessageContents: 2082 enum { reservedMessagesExtension(0), (2^16-1) } MessageExtensionType; 2084 struct { 2085 MessageExtensionType type; 2086 Boolean critical; 2087 opaque extension_contents<0..2^32-1>; 2088 } MessageExtension; 2090 struct { 2091 uint16 message_code; 2092 opaque message_body<0..2^32-1>; 2093 MessageExtensions extensions<0..2^32-1>; 2094 } MessageContents; 2096 The contents of this structure are as follows: 2098 message_code 2099 This indicates the message that is being sent. The code space is 2100 broken up as follows. 2102 0 Reserved 2104 1 .. 0x7fff Requests and responses. These code points are always 2105 paired, with requests being odd and the corresponding response 2106 being the request code plus 1. Thus, "probe_request" (the 2107 Probe request) has value 1 and "probe_answer" (the Probe 2108 response) has value 2 2110 0xffff Error 2111 The message codes are defined in Section 13.8 2112 message_body 2113 The message body itself, represented as a variable-length string 2114 of bytes. The bytes themselves are dependent on the code value. 2115 See the sections describing the various RELOAD methods (Join, 2116 Update, Attach, Store, Fetch, etc.) for the definitions of the 2117 payload contents. 2119 extensions 2120 Extensions to the message. Currently no extensions are defined, 2121 but new extensions can be defined by the process described in 2122 Section 13.14. 2124 All extensions have the following form: 2126 type 2127 The extension type. 2129 critical 2130 Whether this extension must be understood in order to process the 2131 message. If critical = True and the recipient does not understand 2132 the message, it MUST generate an Error_Unknown_Extension error. 2133 If critical = False, the recipient MAY choose to process the 2134 message even if it does not understand the extension. 2136 extension_contents 2137 The contents of the extension (extension-dependent). 2139 5.3.3.1. Response Codes and Response Errors 2141 A peer processing a request returns its status in the message_code 2142 field. If the request was a success, then the message code is the 2143 response code that matches the request (i.e., the next code up). The 2144 response payload is then as defined in the request/response 2145 descriptions. 2147 If the request has failed, then the message code is set to 0xffff 2148 (error) and the payload MUST be an error_response PDU, as shown 2149 below. 2151 When the message code is 0xffff, the payload MUST be an 2152 ErrorResponse. 2154 public struct { 2155 uint16 error_code; 2156 opaque error_info<0..2^16-1>; 2157 } ErrorResponse; 2159 The contents of this structure are as follows: 2161 error_code 2162 A numeric error code indicating the error that occurred. 2164 error_info 2165 An optional arbitrary byte string. Unless otherwise specified, 2166 this will be a UTF-8 text string providing further information 2167 about what went wrong. 2169 The following error code values are defined. The numeric values for 2170 these are defined in Section 13.9. 2172 Error_Forbidden: The requesting node does not have permission to 2173 make this request. 2175 Error_Not_Found: The resource or peer cannot be found or does not 2176 exist. 2178 Error_Request_Timeout: A response to the request has not been 2179 received in a suitable amount of time. The requesting node MAY 2180 resend the request at a later time. 2182 Error_Data_Too_Old: A store cannot be completed because the 2183 storage_time precedes the existing value. 2185 Error_Data_Too_Old: A store cannot be completed because the 2186 storage_time precedes the existing value. 2188 Error_Data_Too_Large: A store cannot be completed because the 2189 requested object exceeds the size limits for that kind. 2191 Error_Generation_Counter_Too_Low: A store cannot be completed 2192 because the generation counter precedes the existing value. 2194 Error_Incompatible_with_Overlay: A peer receiving the request is 2195 using a different overlay, overlay algorithm, or hash algorithm. 2197 Error_Unsupported_Forwarding_Option: A peer receiving the request 2198 with a forwarding options flagged as critical but the peer does 2199 not support this option. See section Section 5.3.2.3. 2201 Error_TTL_Exceeded: A peer receiving the request where the TTL got 2202 decremented to zero. See section Section 5.3.2. 2204 Error_Message_Too_Large: A peer receiving the request that was too 2205 large. See section Section 5.6. 2207 Error_Response_Too_Large: A peer would have generated a response 2208 that is too large per the max_response_length field. 2210 Error_Config_Too_Old: A destination peer received a request with a 2211 configuration sequence that's too old. See Section 5.3.2.1. 2213 Error_Config_Too_New: A destination node received a request with a 2214 configuration sequence that's too new. See Section 5.3.2.1. 2216 Error_Unknown_Kind: A destination node received a request with an 2217 unknown kind-id. See Section 6.4.1.2. 2219 Error_In_Progress: An Attach is already in progress to this peer. 2220 See Section 5.5.1.2. 2222 Error_Unknown_Extension: A destination node received a request with 2223 an unknown extension. 2225 5.3.4. Security Block 2227 The third part of a RELOAD message is the security block. The 2228 security block is represented by a SecurityBlock structure: 2230 enum { x509(0), (255) } certificate_type; 2232 struct { 2233 certificate_type type; 2234 opaque certificate<0..2^16-1>; 2235 } GenericCertificate; 2237 struct { 2238 GenericCertificate certificates<0..2^16-1>; 2239 Signature signature; 2240 } SecurityBlock; 2242 The contents of this structure are: 2244 certificates 2245 A bucket of certificates. 2247 signature 2248 A signature over the message contents. 2250 The certificates bucket SHOULD contain all the certificates necessary 2251 to verify every signature in both the message and the internal 2252 message objects. This is the only location in the message which 2253 contains certificates, thus allowing for only a single copy of each 2254 certificate to be sent. In systems which have some alternate 2255 certificate distribution mechanism, some certificates MAY be omitted. 2256 However, implementors should note that this creates the possibility 2257 that messages may not be immediately verifiable because certificates 2258 must first be retrieved. 2260 Each certificate is represented by a GenericCertificate structure, 2261 which has the following contents: 2263 type 2264 The type of the certificate. Only one type is defined: x509 2265 representing an X.509 certificate. 2267 certificate 2268 The encoded version of the certificate. For X.509 certificates, 2269 it is the DER form. 2271 The signature is computed over the payload and parts of the 2272 forwarding header. The payload, in case of a Store, may contain an 2273 additional signature computed over a StoreReq structure. All 2274 signatures are formatted using the Signature element. This element 2275 is also used in other contexts where signatures are needed. The 2276 input structure to the signature computation varies depending on the 2277 data element being signed. 2279 enum { reservedSignerIdentity(0), 2280 cert_hash(1), (255)} SignerIdentityType; 2282 struct { 2283 select (identity_type) { 2284 case cert_hash; 2285 HashAlgorithm hash_alg; // From TLS 2286 opaque certificate_hash<0..2^8-1>; 2287 /* This structure may be extended with new types if necessary*/ 2288 }; 2289 } SignerIdentityValue; 2291 struct { 2292 SignerIdentityType identity_type; 2293 uint16 length; 2294 SignerIdentityValue identity[SignerIdentity.length]; 2295 } SignerIdentity; 2297 struct { 2298 SignatureAndHashAlgorithm algorithm; // From TLS 2299 SignerIdentity identity; 2300 opaque signature_value<0..2^16-1>; 2301 } Signature; 2303 The signature construct contains the following values: 2305 algorithm 2306 The signature algorithm in use. The algorithm definitions are 2307 found in the IANA TLS SignatureAlgorithm Registry. All 2308 implementations MUST support RSASSA-PKCS1-v1_5 [RFC3447] 2309 signatures with SHA-256 hashes. 2311 identity 2312 The identity used to form the signature. 2314 signature_value 2315 The value of the signature. 2317 The only currently permitted identity format is a hash of the 2318 signer's certificate. The hash_alg field is used to indicate the 2319 algorithm used to produce the hash. The certificate_hash contains 2320 the hash of the certificate object (i.e., the DER-encoded 2321 certificate). The SignerIdentity structure is typed purely to allow 2322 for future (unanticipated) extensibility. 2324 For signatures over messages the input to the signature is computed 2325 over: 2327 overlay || transaction_id || MessageContents || SignerIdentity 2329 where overlay and transaction_id come from the forwarding header and 2330 || indicates concatenation. 2332 The input to signatures over data values is different, and is 2333 described in Section 6.1. 2335 All RELOAD messages MUST be signed. Upon receipt, the receiving node 2336 MUST verify the signature and the authorizing certificate. This 2337 check provides a minimal level of assurance that the sending node is 2338 a valid part of the overlay as well as cryptographic authentication 2339 of the sending node. In addition, responses MUST be checked as 2340 follows: 2342 1. The response to a message sent to a specific Node-ID MUST have 2343 been sent by that Node-ID. 2344 2. The response to a message sent to a Resource-Id MUST have been 2345 sent by a Node-ID which is as close to or closer to the target 2346 Resource-Id than any node in the requesting node's neighbor 2347 table. 2349 The second condition serves as a primitive check for responses from 2350 wildly wrong nodes but is not a complete check. Note that in periods 2351 of churn, it is possible for the requesting node to obtain a closer 2352 neighbor while the request is outstanding. This will cause the 2353 response to be rejected and the request to be retransmitted. 2355 In addition, some methods (especially Store) have additional 2356 authentication requirements, which are described in the sections 2357 covering those methods. 2359 5.4. Overlay Topology 2361 As discussed in previous sections, RELOAD does not itself implement 2362 any overlay topology. Rather, it relies on Topology Plugins, which 2363 allow a variety of overlay algorithms to be used while maintaining 2364 the same RELOAD core. This section describes the requirements for 2365 new topology plugins and the methods that RELOAD provides for overlay 2366 topology maintenance. 2368 5.4.1. Topology Plugin Requirements 2370 When specifying a new overlay algorithm, at least the following need 2371 to be described: 2373 o Joining procedures, including the contents of the Join message. 2374 o Stabilization procedures, including the contents of the Update 2375 message, the frequency of topology probes and keepalives, and the 2376 mechanism used to detect when peers have disconnected. 2377 o Exit procedures, including the contents of the Leave message. 2378 o The length of the Resource-IDs. For DHTs, the hash algorithm to 2379 compute the hash of an identifier. 2380 o The procedures that peers use to route messages. 2381 o The replication strategy used to ensure data redundancy. 2383 All overlay algorithms MUST specify maintenance procedures that send 2384 Updates to clients and peers that have established connections to the 2385 peer responsible for a particular ID when the responsibility for that 2386 ID changes. Because tracking this information is difficult, overlay 2387 algorithms MAY simply specify that an Update is sent to all members 2388 of the Connection Table whenever the range of IDs for which the peer 2389 is responsible changes. 2391 5.4.2. Methods and types for use by topology plugins 2393 This section describes the methods that topology plugins use to join, 2394 leave, and maintain the overlay. 2396 5.4.2.1. Join 2398 A new peer (but one that already has credentials) uses the JoinReq 2399 message to join the overlay. The JoinReq is sent to the responsible 2400 peer depending on the routing mechanism described in the topology 2401 plugin. This notifies the responsible peer that the new peer is 2402 taking over some of the overlay and it needs to synchronize its 2403 state. 2405 struct { 2406 NodeId joining_peer_id; 2407 opaque overlay_specific_data<0..2^16-1>; 2408 } JoinReq; 2410 The minimal JoinReq contains only the Node-ID which the sending peer 2411 wishes to assume. Overlay algorithms MAY specify other data to 2412 appear in this request. Receivers of the JoinReq MUST verify that 2413 the joining_peer_id field matches the Node-ID used to sign the 2414 message and if not MUST reject the message with an Error_Forbidden 2415 error. 2417 If the request succeeds, the responding peer responds with a JoinAns 2418 message, as defined below: 2420 struct { 2421 opaque overlay_specific_data<0..2^16-1>; 2422 } JoinAns; 2424 If the request succeeds, the responding peer MUST follow up by 2425 executing the right sequence of Stores and Updates to transfer the 2426 appropriate section of the overlay space to the joining peer. In 2427 addition, overlay algorithms MAY define data to appear in the 2428 response payload that provides additional info. 2430 In general, nodes which cannot form connections SHOULD report an 2431 error. However, implementations MUST provide some mechanism whereby 2432 nodes can determine that they are potentially the first node and take 2433 responsibility for the overlay. This specification does not mandate 2434 any particular mechanism, but a configuration flag or setting seems 2435 appropriate. 2437 5.4.2.2. Leave 2439 The LeaveReq message is used to indicate that a node is exiting the 2440 overlay. A node SHOULD send this message to each peer with which it 2441 is directly connected prior to exiting the overlay. 2443 struct { 2444 NodeId leaving_peer_id; 2445 opaque overlay_specific_data<0..2^16-1>; 2446 } LeaveReq; 2448 LeaveReq contains only the Node-ID of the leaving peer. Overlay 2449 algorithms MAY specify other data to appear in this request. 2450 Receivers of the LeaveReq MUST verify that the leaving_peer_id field 2451 matches the Node-ID used to sign the message and if not MUST reject 2452 the message with an Error_Forbidden error. 2454 Upon receiving a Leave request, a peer MUST update its own routing 2455 table, and send the appropriate Store/Update sequences to re- 2456 stabilize the overlay. 2458 5.4.2.3. Update 2460 Update is the primary overlay-specific maintenance message. It is 2461 used by the sender to notify the recipient of the sender's view of 2462 the current state of the overlay (its routing state), and it is up to 2463 the recipient to take whatever actions are appropriate to deal with 2464 the state change. In general, peers send Update messages to all 2465 their adjacencies whenever they detect a topology shift. 2467 When a peer receives an Attach request with the send_update flag set 2468 to "true" Section 5.4.2.4, it MUST send an Update message back to the 2469 sender of the Attach request after the completion of the 2470 corresponding ICE check and TLS connection. Note that the sender of 2471 a such Attach request may not have joined the overlay yet. 2473 When a peer detects through an Update that it is no longer 2474 responsible for any data value it is storing, it MUST attempt to 2475 Store a copy to the correct node unless it knows the newly 2476 responsible node already has a copy of the data. This prevents data 2477 loss during large-scale topology shifts such as the merging of 2478 partitioned overlays. 2480 The contents of the UpdateReq message are completely overlay- 2481 specific. The UpdateAns response is expected to be either success or 2482 an error. 2484 5.4.2.4. RouteQuery 2486 The RouteQuery request allows the sender to ask a peer where they 2487 would route a message directed to a given destination. In other 2488 words, a RouteQuery for a destination X requests the Node-ID for the 2489 node that the receiving peer would next route to in order to get to 2490 X. A RouteQuery can also request that the receiving peer initiate an 2491 Update request to transfer the receiving peer's routing table. 2493 One important use of the RouteQuery request is to support iterative 2494 routing. The sender selects one of the peers in its routing table 2495 and sends it a RouteQuery message with the destination_object set to 2496 the Node-ID or Resource-ID it wishes to route to. The receiving peer 2497 responds with information about the peers to which the request would 2498 be routed. The sending peer MAY then use the Attach method to attach 2499 to that peer(s), and repeat the RouteQuery. Eventually, the sender 2500 gets a response from a peer that is closest to the identifier in the 2501 destination_object as determined by the topology plugin. At that 2502 point, the sender can send messages directly to that peer. 2504 5.4.2.4.1. Request Definition 2506 A RouteQueryReq message indicates the peer or resource that the 2507 requesting node is interested in. It also contains a "send_update" 2508 option allowing the requesting node to request a full copy of the 2509 other peer's routing table. 2511 struct { 2512 Boolean send_update; 2513 Destination destination; 2514 opaque overlay_specific_data<0..2^16-1>; 2516 } RouteQueryReq; 2518 The contents of the RouteQueryReq message are as follows: 2520 send_update 2521 A single byte. This may be set to "true" to indicate that the 2522 requester wishes the responder to initiate an Update request 2523 immediately. Otherwise, this value MUST be set to "false". 2525 destination 2526 The destination which the requester is interested in. This may be 2527 any valid destination object, including a Node-ID, compressed ids, 2528 or Resource-ID. 2530 overlay_specific_data 2531 Other data as appropriate for the overlay. 2533 5.4.2.4.2. Response Definition 2535 A response to a successful RouteQueryReq request is a RouteQueryAns 2536 message. This is completely overlay specific. 2538 5.4.2.5. Probe 2540 Probe provides primitive "exploration" services: it allows node to 2541 determine which resources another node is responsible for; and it 2542 allows some discovery services using multicast, anycast, or 2543 broadcast. A probe can be addressed to a specific Node-ID, or the 2544 peer controlling a given location (by using a Resource-ID). In 2545 either case, the target Node-IDs respond with a simple response 2546 containing some status information. 2548 5.4.2.5.1. Request Definition 2550 The ProbeReq message contains a list (potentially empty) of the 2551 pieces of status information that the requester would like the 2552 responder to provide. 2554 enum { reservedProbeInformation(0), responsible_set(1), 2555 num_resources(2), uptime(3), (255)} 2556 ProbeInformationType; 2558 struct { 2559 ProbeInformationType requested_info<0..2^8-1>; 2560 } ProbeReq 2562 The currently defined values for ProbeInformation are: 2564 responsible_set 2565 indicates that the peer should Respond with the fraction of the 2566 overlay for which the responding peer is responsible. 2568 num_resources 2569 indicates that the peer should Respond with the number of 2570 resources currently being stored by the peer. 2572 uptime 2573 indicates that the peer should Respond with how long the peer has 2574 been up in seconds. 2576 5.4.2.5.2. Response Definition 2578 A successful ProbeAns response contains the information elements 2579 requested by the peer. 2581 struct { 2582 select (type) { 2583 case responsible_set: 2584 uint32 responsible_ppb; 2586 case num_resources: 2587 uint32 num_resources; 2589 case uptime: 2590 uint32 uptime; 2591 /* This type may be extended */ 2593 }; 2594 } ProbeInformationData; 2596 struct { 2597 ProbeInformationType type; 2598 uint8 length; 2599 ProbeInformationData value; 2600 } ProbeInformation; 2602 struct { 2603 ProbeInformation probe_info<0..2^16-1>; 2604 } ProbeAns; 2606 A ProbeAns message contains a sequence of ProbeInformation 2607 structures. Each has a "length" indicating the length of the 2608 following value field. This structure allows for unknown option 2609 types. 2611 Each of the current possible Probe information types is a 32-bit 2612 unsigned integer. For type "responsible_ppb", it is the fraction of 2613 the overlay for which the peer is responsible in parts per billion. 2614 For type "num_resources", it is the number of resources the peer is 2615 storing. For the type "uptime" it is the number of seconds the peer 2616 has been up. 2618 The responding peer SHOULD include any values that the requesting 2619 node requested and that it recognizes. They SHOULD be returned in 2620 the requested order. Any other values MUST NOT be returned. 2622 5.5. Forwarding and Link Management Layer 2624 Each node maintains connections to a set of other nodes defined by 2625 the topology plugin. This section defines the methods RELOAD uses to 2626 form and maintain connections between nodes in the overlay. Three 2627 methods are defined: 2629 Attach: used to form RELOAD connections between nodes. When node 2630 A wants to connect to node B, it sends an Attach message to node B 2631 through the overlay. The Attach contains A's ICE parameters. B 2632 responds with its ICE parameters and the two nodes perform ICE to 2633 form connection. Attach also allows two nodes to connect via No- 2634 ICE instead of full ICE. 2635 AppAttach: used to form application layer connections between 2636 nodes. 2637 Ping: is a simple request/response which is used to verify 2638 connectivity of the target peer. 2640 5.5.1. Attach 2642 A node sends an Attach request when it wishes to establish a direct 2643 TCP or UDP connection to another node for the purpose of sending 2644 RELOAD messages. 2646 As described in Section 5.1, an Attach may be routed to either a 2647 Node-ID or to a Resource-ID. An Attach routed to a specific Node-ID 2648 will fail if that node is not reached. An Attach routed to a 2649 Resource-ID will establish a connection with the peer currently 2650 responsible for that Resource-ID, which may be useful in establishing 2651 a direct connection to the responsible peer for use with frequent or 2652 large resource updates. 2654 An Attach in and of itself does not result in updating the routing 2655 table of either node. That function is performed by Updates. If 2656 node A has Attached to node B, but not received any Updates from B, 2657 it MAY route messages which are directly addressed to B through that 2658 channel but MUST NOT route messages through B to other peers via that 2659 channel. The process of Attaching is separate from the process of 2660 becoming a peer (using Join and Update), to prevent half-open states 2661 where a node has started to form connections but is not really ready 2662 to act as a peer. Thus, clients (unlike peers) can simply Attach 2663 without sending Join or Update. 2665 5.5.1.1. Request Definition 2667 An Attach request message contains the requesting node ICE connection 2668 parameters formatted into a binary structure. 2670 enum { reservedOverlayLink(0), DTLS-UDP-SR(1), 2671 DTLS-UDP-SR-NO-ICE(3), TLS-TCP-FH-NO-ICE(4), 2672 (255) } OverlayLinkType; 2674 enum { reservedCand(0), host(1), srflx(2), prflx(3), relay(4), 2675 (255) } CandType; 2677 struct { 2678 opaque name<0..2^16-1>; 2679 opaque value<0..2^16-1>; 2680 } IceExtension; 2682 struct { 2683 IpAddressPort addr_port; 2684 OverlayLinkType overlay_link; 2685 opaque foundation<0..255>; 2686 uint32 priority; 2687 CandType type; 2688 select (type){ 2689 case host: 2690 ; /* Nothing */ 2691 case srflx: 2692 case prflx: 2693 case relay: 2694 IpAddressPort rel_addr_port; 2695 }; 2696 IceExtension extensions<0..2^16-1>; 2697 } IceCandidate; 2699 struct { 2700 opaque ufrag<0..2^8-1>; 2701 opaque password<0..2^8-1>; 2702 opaque role<0..2^8-1>; 2703 IceCandidate candidates<0..2^16-1>; 2704 Boolean send_update; 2705 } AttachReqAns; 2707 The values contained in AttachReqAns are: 2709 ufrag 2710 The username fragment (from ICE). 2712 password 2713 The ICE password. 2715 role 2716 An active/passive/actpass attribute from RFC 4145 [RFC4145]. This 2717 value MUST be 'passive' for the offerer (the peer sending the 2718 Attach request) and 'active' for the answerer (the peer sending 2719 the Attach response). 2721 candidates 2722 One or more ICE candidate values, as described below. 2723 send_update 2724 Has the same meaning as the send_update field in RouteQueryReq. 2726 Each ICE candidate is represented as an IceCandidate structure, which 2727 is a direct translation of the information from the ICE string 2728 structures, with the exception of the component ID. Since there is 2729 only one component, it is always 1, and thus left out of the PDU. 2730 The remaining values are specified as follows: 2732 addr_port 2733 corresponds to the connection-address and port productions. 2735 overlay_link 2736 corresponds to the OverlayLinkType production, Overlay Link 2737 protocols used with No-ICE MUST specify "No-ICE" in their 2738 description. Future overlay link values can be added be defining 2739 new OverlayLinkType values in the IANA registry in Section 13.10. 2740 Future extensions to the encapsulation or framing that provide for 2741 backward compatibility with that specified by a previously defined 2742 OverlayLinkType values MUST use that previous value. 2743 OverlayLinkType protocols are defined in Section 5.6 2744 A single AttachReqAns MUST NOT include both candidates whose 2745 OverlayLinkType protocols use ICE (the default) and candidates 2746 that specify "No-ICE". 2748 foundation 2749 corresponds to the foundation production. 2751 priority 2752 corresponds to the priority production. 2754 type 2755 corresponds to the cand-type production. 2757 rel_addr_port 2758 corresponds to the rel-addr and rel-port productions. Only 2759 present for type "relay". 2761 extensions 2762 ICE extensions. The name and value fields correspond to binary 2763 translations of the equivalent fields in the ICE extensions. 2765 These values should be generated using the procedures described in 2766 Section 5.5.1.3. 2768 5.5.1.2. Response Definition 2770 If a peer receives an Attach request, it MUST determine how to 2771 process the request as follows: 2773 o If it has not initiated an Attach request to the originating peer 2774 of this Attach request, it MUST process this request and SHOULD 2775 generate its own response with an AttachReqAns. It should then 2776 begin ICE checks. 2777 o If it has already sent an Attach request to and received the 2778 response from the originating peer of this Attach request, and as 2779 a as a result, an ICE check and TLS connection is in progress, 2780 then it SHOULD generate an Error_In_Progress error instead of an 2781 AttachReqAns. 2782 o If it has already sent an Attach request to but not yet received 2783 the response from the originating peer of this Attach request, it 2784 SHOULD apply the following tie-breaker heuristic to determine how 2785 to handle this Attach request and the incomplete Attach request it 2786 has sent out: 2787 * If the peer's own Node-ID is smaller, it MUST cancel its own 2788 incomplete Attach request. It MUST then process this Attach 2789 request, generate an AttachReqAns response, and proceed with 2790 the corresponding ICE check. 2791 * If the peer's own Node-ID is larger, it MUST generate an 2792 Error_In_Progress error to this Attach request, then proceed to 2793 wait for and complete the Attach and the corresponding ICE 2794 check it has originated. 2795 If the peer is overloaded or detects some other kind of error, it 2796 MAY generate an error instead of an AttachReqAns. 2798 When a peer receives an Attach response, it SHOULD parse the response 2799 and begin its own ICE checks. 2801 5.5.1.3. Using ICE With RELOAD 2803 This section describes the profile of ICE that is used with RELOAD. 2804 RELOAD implementations MUST implement full ICE. 2806 In ICE as defined by [RFC5245], SDP is used to carry the ICE 2807 parameters. In RELOAD, this function is performed by a binary 2808 encoding in the Attach method. This encoding is more restricted than 2809 the SDP encoding because the RELOAD environment is simpler: 2811 o Only a single media stream is supported. 2812 o In this case, the "stream" refers not to RTP or other types of 2813 media, but rather to a connection for RELOAD itself or for SIP 2814 signaling. 2815 o RELOAD only allows for a single offer/answer exchange. Unlike the 2816 usage of ICE within SIP, there is never a need to send a 2817 subsequent offer to update the default candidates to match the 2818 ones selected by ICE. 2820 An agent follows the ICE specification as described in [RFC5245] with 2821 the changes and additional procedures described in the subsections 2822 below. 2824 5.5.1.4. Collecting STUN Servers 2826 ICE relies on the node having one or more STUN servers to use. In 2827 conventional ICE, it is assumed that nodes are configured with one or 2828 more STUN servers through some out of band mechanism. This is still 2829 possible in RELOAD but RELOAD also learns STUN servers as it connects 2830 to other peers. Because all RELOAD peers implement ICE and use STUN 2831 keepalives, every peer is a capable of responding to STUN Binding 2832 requests [RFC5389]. Accordingly, any peer that a node knows about 2833 can be used like a STUN server -- though of course it may be behind a 2834 NAT. 2836 A peer on a well-provisioned wide-area overlay will be configured 2837 with one or more bootstrap nodes. These nodes make an initial list 2838 of STUN servers. However, as the peer forms connections with 2839 additional peers, it builds more peers it can use like STUN servers. 2841 Because complicated NAT topologies are possible, a peer may need more 2842 than one STUN server. Specifically, a peer that is behind a single 2843 NAT will typically observe only two IP addresses in its STUN checks: 2844 its local address and its server reflexive address from a STUN server 2845 outside its NAT. However, if there are more NATs involved, it may 2846 learn additional server reflexive addresses (which vary based on 2847 where in the topology the STUN server is). To maximize the chance of 2848 achieving a direct connection, a peer SHOULD group other peers by the 2849 peer-reflexive addresses it discovers through them. It SHOULD then 2850 select one peer from each group to use as a STUN server for future 2851 connections. 2853 Only peers to which the peer currently has connections may be used. 2855 If the connection to that host is lost, it MUST be removed from the 2856 list of stun servers and a new server from the same group MUST be 2857 selected unless there are no others servers in the group in which 2858 case some other peer MAY be used. 2860 5.5.1.5. Gathering Candidates 2862 When a node wishes to establish a connection for the purposes of 2863 RELOAD signaling or application signaling, it follows the process of 2864 gathering candidates as described in Section 4 of ICE [RFC5245]. 2865 RELOAD utilizes a single component. Consequently, gathering for 2866 these "streams" requires a single component. In the case where a 2867 node has not yet found a TURN server, the agent would not include a 2868 relayed candidate. 2870 The ICE specification assumes that an ICE agent is configured with, 2871 or somehow knows of, TURN and STUN servers. RELOAD provides a way 2872 for an agent to learn these by querying the overlay, as described in 2873 Section 5.5.1.4 and Section 8. 2875 The default candidate selection described in Section 4.1.4 of ICE is 2876 ignored; defaults are not signaled or utilized by RELOAD. 2878 An alternative to using the full ICE supported by the Attach request 2879 is to use No-ICE mechanism by providing candidates with "No-ICE" 2880 Overlay Link protocols. Configuration for the overlay indicates 2881 whether or not these Overlay Link protocols can be used. An overlay 2882 MUST be either all ICE or all No-ICE. 2884 No-ICE will not work in all of the scenarios where ICE would work, 2885 but in some cases, particularly those with no NATs or firewalls, it 2886 will work. Therefore it is RECOMMENDED that full ICE be used even 2887 for a node that has a public, unfiltered IP address, to take 2888 advantage of STUN connectivity checks, etc. 2890 5.5.1.6. Prioritizing Candidates 2892 At the time of writing, UDP is the only transport protocol specified 2893 to work with ICE. However, standardization of additional protocols 2894 for use with ICE is expected, including TCP and datagram-oriented 2895 protocols such as SCTP and DCCP. In particular, UDP encapsulations 2896 for SCTP and DCCP are expected to be standardized in the near future, 2897 greatly expanding the available Overlay Link protocols available for 2898 RELOAD. When additional protocols are available, the following 2899 prioritization is RECOMMENDED: 2901 o Highest priority is assigned to message-oriented protocols that 2902 offer well-understood congestion and flow control without head of 2903 line blocking. For example, SCTP without message ordering, DCCP, 2904 or those protocols encapsulated using UDP. 2905 o Second highest priority is assigned to stream-oriented protocols, 2906 e.g. TCP. 2907 o Lowest priority is assigned to protocols encapsulated over UDP 2908 that do not implement well-established congestion control 2909 algorithms. The DTLS/UDP with SR overlay link protocol is an 2910 example of such a protocol. 2912 5.5.1.7. Encoding the Attach Message 2914 Section 4.3 of ICE describes procedures for encoding the SDP for 2915 conveying RELOAD candidates. Instead of actually encoding an SDP, 2916 the candidate information (IP address and port and transport 2917 protocol, priority, foundation, type and related address) is carried 2918 within the attributes of the Attach request or its response. 2919 Similarly, the username fragment and password are carried in the 2920 Attach message or its response. Section 5.5.1 describes the detailed 2921 attribute encoding for Attach. The Attach request and its response 2922 do not contain any default candidates or the ice-lite attribute, as 2923 these features of ICE are not used by RELOAD. 2925 Since the Attach request contains the candidate information and short 2926 term credentials, it is considered as an offer for a single media 2927 stream that happens to be encoded in a format different than SDP, but 2928 is otherwise considered a valid offer for the purposes of following 2929 the ICE specification. Similarly, the Attach response is considered 2930 a valid answer for the purposes of following the ICE specification. 2932 5.5.1.8. Verifying ICE Support 2934 An agent MUST skip the verification procedures in Section 5.1 and 6.1 2935 of ICE. Since RELOAD requires full ICE from all agents, this check 2936 is not required. 2938 5.5.1.9. Role Determination 2940 The roles of controlling and controlled as described in Section 5.2 2941 of ICE are still utilized with RELOAD. However, the offerer (the 2942 entity sending the Attach request) will always be controlling, and 2943 the answerer (the entity sending the Attach response) will always be 2944 controlled. The connectivity checks MUST still contain the ICE- 2945 CONTROLLED and ICE-CONTROLLING attributes, however, even though the 2946 role reversal capability for which they are defined will never be 2947 needed with RELOAD. This is to allow for a common codebase between 2948 ICE for RELOAD and ICE for SDP. 2950 5.5.1.10. Full ICE 2952 When neither side has provided an No-ICE candidate, connectivity 2953 checks and nominations are used as in regular ICE. 2955 5.5.1.10.1. Connectivity Checks 2957 The processes of forming check lists in Section 5.7 of ICE, 2958 scheduling checks in Section 5.8, and checking connectivity checks in 2959 Section 7 are used with RELOAD without change. 2961 5.5.1.10.2. Concluding ICE 2963 The procedures in Section 8 of ICE are followed to conclude ICE, with 2964 the following exceptions: 2966 o The controlling agent MUST NOT attempt to send an updated offer 2967 once the state of its single media stream reaches Completed. 2968 o Once the state of ICE reaches Completed, the agent can immediately 2969 free all unused candidates. This is because RELOAD does not have 2970 the concept of forking, and thus the three second delay in Section 2971 8.3 of ICE does not apply. 2973 5.5.1.10.3. Media Keepalives 2975 STUN MUST be utilized for the keepalives described in Section 10 of 2976 ICE. 2978 5.5.1.11. No-ICE 2980 No-ICE is selected when either side has provided "no ICE" Overlay 2981 Link candidates. STUN is not used for connectivity checks when doing 2982 No-ICE; instead the DTLS or TLS handshake (or similar security layer 2983 of future overlay link protocols) forms the connectivity check. The 2984 certificate exchanged during the (D)TLS handshake must match the node 2985 that sent the AttachReqAns and if it does not, the connection MUST be 2986 closed. 2988 5.5.1.12. Subsequent Offers and Answers 2990 An agent MUST NOT send a subsequent offer or answer. Thus, the 2991 procedures in Section 9 of ICE MUST be ignored. 2993 5.5.1.13. Sending Media 2995 The procedures of Section 11 of ICE apply to RELOAD as well. 2996 However, in this case, the "media" takes the form of application 2997 layer protocols (RELOAD) over TLS or DTLS. Consequently, once ICE 2998 processing completes, the agent will begin TLS or DTLS procedures to 2999 establish a secure connection. The node which sent the Attach 3000 request MUST be the TLS server. The other node MUST be the TLS 3001 client. The server MUST request TLS client authentication. The 3002 nodes MUST verify that the certificate presented in the handshake 3003 matches the identity of the other peer as found in the Attach 3004 message. Once the TLS or DTLS signaling is complete, the application 3005 protocol is free to use the connection. 3007 The concept of a previous selected pair for a component does not 3008 apply to RELOAD, since ICE restarts are not possible with RELOAD. 3010 5.5.1.14. Receiving Media 3012 An agent MUST be prepared to receive packets for the application 3013 protocol (TLS or DTLS carrying RELOAD, SIP or anything else) at any 3014 time. The jitter and RTP considerations in Section 11 of ICE do not 3015 apply to RELOAD. 3017 5.5.2. AppAttach 3019 A node sends an AppAttach request when it wishes to establish a 3020 direct connection to another node for the purposes of sending 3021 application layer messages. AppAttach is nearly identical to Attach, 3022 except for the purpose of the connection: it is used to transport 3023 non-RELOAD "media". A separate request is used to avoid implementor 3024 confusion between the two methods (this was found to be a real 3025 problem with initial implementations). The AppAttach request and its 3026 response contain an application attribute, which indicates what 3027 protocol is to be run over the connection. 3029 5.5.2.1. Request Definition 3031 An AppAttachReq message contains the requesting node's ICE connection 3032 parameters formatted into a binary structure. 3034 struct { 3035 opaque ufrag<0..2^8-1>; 3036 opaque password<0..2^8-1>; 3037 uint16 application; 3038 opaque role<0..2^8-1>; 3039 IceCandidate candidates<0..2^16-1>; 3040 } AppAttachReq; 3042 The values contained in AppAttachReq and AppAttachAns are: 3044 ufrag 3045 The username fragment (from ICE) 3047 password 3048 The ICE password. 3050 application 3051 A 16-bit application-id as defined in the Section 13.5. This 3052 number represents the IANA registered application that is going to 3053 send data on this connection. For SIP, this is 5060 or 5061. 3055 role 3056 An active/passive/actpass attribute from RFC 4145 [RFC4145]. 3058 candidates 3059 One or more ICE candidate values 3061 The application using connection set up with this request is 3062 responsible for providing sufficiently frequent keep traffic for NAT 3063 and Firewall keep alive and for deciding when to close the 3064 connection. 3066 5.5.2.2. Response Definition 3068 If a peer receives an AppAttach request, it SHOULD process the 3069 request and generate its own response with a AppAttachAns. It should 3070 then begin ICE checks. When a peer receives an AppAttach response, 3071 it SHOULD parse the response and begin its own ICE checks. If the 3072 application ID is not supported, the peer MUST reply with an 3073 Error_Not_Found error. 3075 struct { 3076 opaque ufrag<0..2^8-1>; 3077 opaque password<0..2^8-1>; 3078 uint16 application; 3079 opaque role<0..2^8-1>; 3080 IceCandidate candidates<0..2^16-1>; 3081 } AppAttachAns; 3083 The meaning of the fields is the same as in the AppAttachReq. 3085 5.5.3. Ping 3087 Ping is used to test connectivity along a path. A ping can be 3088 addressed to a specific Node-ID, to the peer controlling a given 3089 location (by using a resource ID), or to the broadcast Node-ID 3090 (2^128-1). 3092 5.5.3.1. Request Definition 3094 struct { 3095 opaque<0..2^16-1> padding; 3096 } PingReq 3098 The Ping request is empty of meaningful contents. However, it may 3099 contain up to 65535 bytes of padding to facilitate the discovery of 3100 overlay maximum packet sizes. 3102 5.5.3.2. Response Definition 3104 A successful PingAns response contains the information elements 3105 requested by the peer. 3107 struct { 3108 uint64 response_id; 3109 uint64 time; 3110 } PingAns; 3112 A PingAns message contains the following elements: 3114 response_id 3115 A randomly generated 64-bit response ID. This is used to 3116 distinguish Ping responses. 3118 time 3119 The time when the Ping response was created represented in the 3120 same way as storage_time defined in Section 6. 3122 5.5.4. ConfigUpdate 3124 The ConfigUpdate method is used to push updated configuration data 3125 across the overlay. Whenever a node detects that another node has 3126 old configuration data, it MUST generate a ConfigUpdate request. The 3127 ConfigUpdate request allows updating of two kinds of data: the 3128 configuration data (Section 5.3.2.1) and kind information 3129 (Section 6.4.1.1). 3131 5.5.4.1. Request Definition 3133 enum { reservedConfigUpdate(0), config(1), kind(2), (255) } 3134 ConfigUpdateType; 3136 typedef uint32 KindId; 3137 typedef opaque KindDescription<0..2^16-1>; 3139 struct { 3140 ConfigUpdateType type; 3141 uint32 length; 3143 select (type) { 3144 case config: 3145 opaque config_data<0..2^24-1>; 3147 case kind: 3148 KindDescription kinds<0..2^24-1>; 3150 /* This structure may be extended with new types*/ 3151 }; 3152 } ConfigUpdateReq; 3154 The ConfigUpdateReq message contains the following elements: 3156 type 3157 The type of the contents of the message. This structure allows 3158 for unknown content types. 3159 length 3160 The length of the remainder of the message. This is included to 3161 preserve backward compatibility and is 32 bits instead of 24 to 3162 facilitate easy conversion between network and host byte order. 3163 config_data (type==config) 3164 The contents of the configuration document. 3165 kinds (type==kind) 3166 One or more XML kind-block productions (see Section 10.1). These 3167 MUST be encoded with UTF-8 and assume a default namespace of 3168 "urn:ietf:params:xml:ns:p2p:config-base". 3170 5.5.4.2. Response Definition 3172 struct { 3173 } ConfigUpdateAns 3175 If the ConfigUpdateReq is of type "config" it MUST only be processed 3176 if all the following are true: 3178 o The sequence number in the document is greater than the current 3179 configuration sequence number. 3180 o The configuration document is correctly digitally signed (see 3181 Section 10 for details on signatures. 3182 Otherwise appropriate errors MUST be generated. 3184 If the ConfigUpdateReq is of type "kind" it MUST only be processed if 3185 it is correctly digitally signed by an acceptable kind signer as 3186 specified in the configuration file. Details on kind-signer field in 3187 the configuration file is described in Section 10.1. In addition, if 3188 the kind update conflicts with an existing known kind (i.e., it is 3189 signed by a different signer), then it should be rejected with 3190 "Error_Forbidden". This should not happen in correctly functioning 3191 overlays. 3193 If the update is acceptable, then the node MUST reconfigure itself to 3194 match the new information. This may include adding permissions for 3195 new kinds, deleting old kinds, or even, in extreme circumstances, 3196 exiting and reentering the overlay, if, for instance, the DHT 3197 algorithm has changed. 3199 The response for ConfigUpdate is empty. 3201 5.6. Overlay Link Layer 3203 RELOAD can use multiple Overlay Link protocols to send its messages. 3204 Because ICE is used to establish connections (see Section 5.5.1.3), 3205 RELOAD nodes are able to detect which Overlay Link protocols are 3206 offered by other nodes and establish connections between them. Any 3207 link protocol needs to be able to establish a secure, authenticated 3208 connection and to provide data origin authentication and message 3209 integrity for individual data elements. RELOAD currently supports 3210 three Overlay Link protocols: 3212 o DTLS [RFC4347] over UDP with Simple Reliability (SR) 3213 o TLS [RFC5246] over TCP with Framing Header, No-ICE 3214 o DTLS [RFC4347] over UDP with SR, No-ICE 3216 Note that although UDP does not properly have "connections", both TLS 3217 and DTLS have a handshake which establishes a similar, stateful 3218 association, and we simply refer to these as "connections" for the 3219 purposes of this document. 3221 If a peer receives a message that is larger than value of max- 3222 message-size defined in the overlay configuration, the peer SHOULD 3223 send an Error_Message_Too_Large error and then close the TLS or DTLS 3224 session from which the message was received. Note that this error 3225 can be sent and the session closed before receiving the complete 3226 message. If the forwarding header is larger than the max-message- 3227 size, the receiver SHOULD close the TLS or DTLS session without 3228 sending an error. 3230 The Framing Header (FH) is used to frame messages and provide timing 3231 when used on a reliable stream-based transport protocol. Simple 3232 Reliability (SR) makes use of the FH to provide congestion control 3233 and semi-reliability when using unreliable message-oriented transport 3234 protocols. We will first define each of these algorithms, then 3235 define overlay link protocols that use them. 3237 Note: We expect future Overlay Link protocols to define replacements 3238 for all components of these protocols, including the framing header. 3239 These protocols have been chosen for simplicity of implementation and 3240 reasonable performance. 3242 Note to implementers: There are inherent tradeoffs in utilizing 3243 short timeouts to determine when a link has failed. To balance the 3244 tradeoffs, an implementation should be able to quickly act to remove 3245 entries from the routing table when there is reason to suspect the 3246 link has failed. For example, in a Chord-derived overlay algorithm, 3247 a closer finger table entry could be substituted for an entry in the 3248 finger table that has experienced a timeout. That entry can be 3249 restored if it proves to resume functioning, or replaced at some 3250 point in the future if necessary. End-to-end retransmissions will 3251 handle any lost messages, but only if the failing entries do not 3252 remain in the finger table for subsequent retransmissions. 3254 5.6.1. Future Overlay Link Protocols 3256 The only currently defined overlay link protocols are TLS and DTLS. 3257 It is possible to define new link-layer protocols and apply them to a 3258 new overlay using the "overlay-link-protocol" configuration directive 3259 (see Section 10.1.). However, any new protocols MUST meet the 3260 following requirements. 3262 Endpoint authentication When a node forms an association with 3263 another endpoint, it MUST be possible to cryptographically verify 3264 that the endpoint has a given Node-Id. 3266 Traffic origin authentication and integrity When a node receives 3267 traffic from another endpoint, it MUST be possible to 3268 cryptographically verify that the traffic came from a given 3269 association and that it has not been modified in transit from the 3270 other endpoint in the association. The overlay link protocol MUST 3271 also provide replay prevention/detection. 3273 Traffic confidentiality When a node sends traffic to another 3274 endpoint, it MUST NOT be possible for a third party not involved 3275 in the association to determine the contents of that traffic. 3277 Any new overlay protocol MUST be defined via RFC 5226 Standards 3278 Action; see Section 13.11. 3280 5.6.1.1. HIP 3282 In a Host Identity Protocol Based Overlay Networking Environment (HIP 3283 BONE) [I-D.ietf-hip-bone] HIP [RFC5201] provides connection 3284 management (e.g., NAT traversal and mobility) and security for the 3285 overlay network. The P2PSIP Working Group has expressed interest in 3286 supporting a HIP-based link protocol. Such support would require 3287 specifying such details as: 3289 o How to issue certificates which provided identities meaningful to 3290 the HIP base exchange. We anticipate that this would require a 3291 mapping between ORCHIDs and NodeIds. 3292 o How to carry the HIP I1 and I2 messages. 3293 o How to carry RELOAD messages over HIP. 3295 [I-D.ietf-hip-reload-instance] documents work in progress on using 3296 RELOAD with the HIP BONE. 3298 5.6.1.2. ICE-TCP 3300 The ICE-TCP draft [I-D.ietf-mmusic-ice-tcp] should allow TCP to be 3301 supported as an Overlay Link protocol that can be added using ICE. 3303 5.6.1.3. Message-oriented Transports 3305 Modern message-oriented transports offer high performance, good 3306 congestion control, and avoid head of line blocking in case of lost 3307 data. These characteristics make them preferable as underlying 3308 transport protocols for RELOAD links. SCTP without message ordering 3309 and DCCP are two examples of such protocols. However, currently they 3310 are not well-supported by commonly available NATs, and specifications 3311 for ICE session establishment are not available. 3313 5.6.1.4. Tunneled Transports 3315 As of the time of this writing, there is significant interest in the 3316 IETF community in tunneling other transports over UDP, motivated by 3317 the situation that UDP is well-supported by modern NAT hardware, and 3318 similar performance can be achieved to native implementation. 3319 Currently SCTP, DCCP, and a generic tunneling extension are being 3320 proposed for message-oriented protocols. Baset et al. have proposed 3321 tunneling TCP over UDP for similar reasons 3322 [I-D.baset-tsvwg-tcp-over-udp]. Once ICE traversal has been 3323 specified for these tunneled protocols, they should be 3324 straightforward to support as overlay link protocols. 3326 5.6.2. Framing Header 3328 In order to support unreliable links and to allow for quick detection 3329 of link failures when using reliable end-to-end transports, each 3330 message is wrapped in a very simple framing layer (FramedMessage) 3331 which is only used for each hop. This layer contains a sequence 3332 number which can then be used for ACKs. The same header is used for 3333 both reliable and unreliable transports for simplicity of 3334 implementation. 3336 The definition of FramedMessage is: 3338 enum { data(128), ack(129), (255)} FramedMessageType; 3340 struct { 3341 FramedMessageType type; 3343 select (type) { 3344 case data: 3345 uint32 sequence; 3346 opaque message<0..2^24-1>; 3348 case ack: 3349 uint32 ack_sequence; 3350 uint32 received; 3351 }; 3352 } FramedMessage; 3354 The type field of the PDU is set to indicate whether the message is 3355 data or an acknowledgement. 3357 If the message is of type "data", then the remainder of the PDU is as 3358 follows: 3360 sequence 3361 the sequence number. This increments by 1 for each framed message 3362 sent over this transport session. 3364 message 3365 the message that is being transmitted. 3367 Each connection has it own sequence number space. Initially the 3368 value is zero and it increments by exactly one for each message sent 3369 over that connection. 3371 When the receiver receives a message, it SHOULD immediately send an 3372 ACK message. The receiver MUST keep track of the 32 most recent 3373 sequence numbers received on this association in order to generate 3374 the appropriate ack. 3376 If the PDU is of type "ack", the contents are as follows: 3378 ack_sequence 3379 The sequence number of the message being acknowledged. 3381 received 3382 A bitmask indicating if each of the previous 32 sequence numbers 3383 before this packet has been among the 32 packets most recently 3384 received on this connection. When a packet is received with a 3385 sequence number N, the receiver looks at the sequence number of 3386 the previously 32 packets received on this connection. Call the 3387 previously received packet number M. For each of the previous 32 3388 packets, if the sequence number M is less than N but greater than 3389 N-32, the N-M bit of the received bitmask is set to one; otherwise 3390 it is zero. Note that a bit being set to one indicates positively 3391 that a particular packet was received, but a bit being set to zero 3392 means only that it is unknown whether or not the packet has been 3393 received, because it might have been received before the 32 most 3394 recently received packets. 3396 The received field bits in the ACK provide a high degree of 3397 redundancy so that the sender can figure out which packets the 3398 receiver has received and can then estimate packet loss rates. If 3399 the sender also keeps track of the time at which recent sequence 3400 numbers have been sent, the RTT can be estimated. 3402 5.6.3. Simple Reliability 3404 When RELOAD is carried over DTLS or another unreliable link protocol, 3405 it needs to be used with a reliability and congestion control 3406 mechanism, which is provided on a hop-by-hop basis. The basic 3407 principle is that each message, regardless of whether or not it 3408 carries a request or response, will get an ACK and be reliably 3409 retransmitted. The receiver's job is very simple, limited to just 3410 sending ACKs. All the complexity is at the sender side. This allows 3411 the sending implementation to trade off performance versus 3412 implementation complexity without affecting the wire protocol. 3414 5.6.3.1. Retransmission and Flow Control 3416 Because the receiver's role is limited to providing packet 3417 acknowledgements, a wide variety of congestion control algorithms can 3418 be implemented on the sender side while using the same basic wire 3419 protocol. In general, senders MAY implement any rate control scheme 3420 of their choice, provided that it is REQUIRED to be no more 3421 aggressive then TFRC[RFC5348]. 3423 The following section describes a simple, inefficient, scheme that 3424 complies with this requirement. Another alternative would be TFRC-SP 3425 [RFC4828] and use the received bitmask to allow the sender to compute 3426 packet loss event rates. 3428 5.6.3.1.1. Trivial Retransmission 3430 A node SHOULD retransmit a message if it has not received an ACK 3431 after an interval of RTO ("Retransmission TimeOut"). The node MUST 3432 double the time to wait after each retransmission. In each 3433 retransmission, the sequence number is incremented. 3435 The RTO is an estimate of the round-trip time (RTT). Implementations 3436 can use a static value for RTO or a dynamic estimate which will 3437 result in better performance. For implementations that use a static 3438 value, the default value for RTO is 500 ms. Nodes MAY use smaller 3439 values of RTO if it is known that all nodes are within the local 3440 network. The default RTO MAY be chosen larger, and this is 3441 RECOMMENDED if it is known in advance (such as on high latency access 3442 links) that the round-trip time is larger. 3444 Implementations that use a dynamic estimate to compute the RTO MUST 3445 use the algorithm described in RFC 2988[RFC2988], with the exception 3446 that the value of RTO SHOULD NOT be rounded up to the nearest second 3447 but instead rounded up to the nearest millisecond. The RTT of a 3448 successful STUN transaction from the ICE stage is used as the initial 3449 measurement for formula 2.2 of RFC 2988. The sender keeps track of 3450 the time each message was sent for all recently sent messages. Any 3451 time an ACK is received, the sender can compute the RTT for that 3452 message by looking at the time the ACK was received and the time when 3453 the message was sent. This is used as a subsequent RTT measurement 3454 for formula 2.3 of RFC 2988 to update the RTO estimate. (Note that 3455 because retransmissions receive new sequence numbers, all received 3456 ACKs are used.) 3457 The value for RTO is calculated separately for each DTLS session. 3459 Retransmissions continue until a response is received, or until a 3460 total of 5 requests have been sent or there has been a hard ICMP 3461 error [RFC1122] or a TLS alert. The sender knows a response was 3462 received when it receives an ACK with a sequence number that 3463 indicates it is a response to one of the transmissions of this 3464 messages. For example, assuming an RTO of 500 ms, requests would be 3465 sent at times 0 ms, 500 ms, 1500 ms, 3500 ms, and 7500 ms. If all 3466 retransmissions for a message fail, then the sending node SHOULD 3467 close the connection routing the message. 3469 To determine when a link may be failing without waiting for the final 3470 timeout, observe when no ACKs have been received for an entire RTO 3471 interval, and then wait for three retransmissions to occur beyond 3472 that point. If no ACKs have been received by the time the third 3473 retransmission occurs, it is RECOMMENDED that the link be removed 3474 from the routing table. The link MAY be restored to the routing 3475 table if ACKs resume before the connection is closed, as described 3476 above. 3478 Once an ACK has been received for a message, the next message can be 3479 sent, but the peer SHOULD ensure that there is at least 10 ms between 3480 sending any two messages. The only time a value less than 10 ms can 3481 be used is when it is known that all nodes are on a network that can 3482 support retransmissions faster than 10 ms with no congestion issues. 3484 5.6.4. DTLS/UDP with SR 3486 This overlay link protocol consists of DTLS over UDP while 3487 implementing the Simple Reliability protocol. STUN Connectivity 3488 checks and keepalives are used. 3490 5.6.5. TLS/TCP with FH, No-ICE 3492 This overlay link protocol consists of TLS over TCP with the framing 3493 header. Because ICE is not used, STUN connectivity checks are not 3494 used upon establishing the TCP connection, nor are they used for 3495 keepalives. 3497 Because the TCP layer's application-level timeout is too slow to be 3498 useful for overlay routing, the Overlay Link implementation MUST use 3499 the framing header to measure the RTT of the connection and calculate 3500 an RTO as specified in Section 2 of [RFC2988]. The resulting RTO is 3501 not used for retransmissions, but as a timeout to indicate when the 3502 link SHOULD be removed from the routing table. It is RECOMMENDED 3503 that such a connection be retained for 30s to determine if the 3504 failure was transient before concluding the link has failed 3505 permanently. 3507 When sending candidates for TLS/TCP with FH, No-ICE, a passive 3508 candidate MUST be provided. 3510 5.6.6. DTLS/UDP with SR, No-ICE 3512 This overlay link protocol consists of DTLS over UDP while 3513 implementing the Simple Reliability protocol. Because ICE is not 3514 used, no STUN connectivity checks or keepalives are used. 3516 5.7. Fragmentation and Reassembly 3518 In order to allow transmission over datagram protocols such as DTLS, 3519 RELOAD messages may be fragmented. 3521 Any node along the path can fragment the message but only the final 3522 destination reassembles the fragments. When a node takes a packet 3523 and fragments it, each fragment has a full copy of the Forwarding 3524 Header but the data after the Forwarding Header is broken up in 3525 appropriate sized chunks. The size of the payload chunks needs to 3526 take into account space to allow the via and destination lists to 3527 grow. Each fragment MUST contain a full copy of the via and 3528 destination list and MUST contain at least 256 bytes of the message 3529 body. If the via and destination list are so large that this is not 3530 possible, RELOAD fragmentation is not performed and IP-layer 3531 fragmentation is allowed to occur. When a message must be 3532 fragmented, it SHOULD be split into equal-sized fragments that are no 3533 larger than the PMTU of the next overlay link minus 32 bytes. This 3534 is to allow the via list to grow before further fragmentation is 3535 required. 3537 Note that this fragmentation is not optimal for the end-to-end path - 3538 a message may be refragmented multiple times as it traverses the 3539 overlay but is only assembled at the final destination. This option 3540 has been chosen as it is far easier to implement than e2e PMTU 3541 discovery across an ever-changing overlay, and it effectively 3542 addresses the reliability issues of relying on IP-layer 3543 fragmentation. However, PING can be used to allow e2e PMTU to be 3544 implemented if desired. 3546 Upon receipt of a fragmented message by the intended peer, the peer 3547 holds the fragments in a holding buffer until the entire message has 3548 been received. The message is then reassembled into a single message 3549 and processed. In order to mitigate denial of service attacks, 3550 receivers SHOULD time out incomplete fragments after maximum request 3551 lifetime (15 seconds). Note this time was derived from looking at 3552 the end to end retransmission time and saving fragments long enough 3553 for the full end to end retransmissions to take place. Ideally the 3554 receiver would have enough buffer space to deal with as many 3555 fragments as can arrive in the maximum request lifetime. However, if 3556 the receiver runs out of buffer space to reassemble the messages it 3557 MUST drop the message. 3559 When a message is fragmented, the fragment offset value is stored in 3560 the lower 24 bits of the fragment field of the forwarding header. 3561 The offset is the number of bytes between the end of the forwarding 3562 header and the start of the data. The first fragment therefore has 3563 an offset of 0. The first and last bit indicators MUST be 3564 appropriately set. If the message is not fragmented, then both the 3565 first and last fragment bits are set to 1 and the offset is 0 3566 resulting in a fragment value of 0xC0000000. Note that this means 3567 that the first fragment bit is always 1, so isn't actually that 3568 useful. 3570 6. Data Storage Protocol 3572 RELOAD provides a set of generic mechanisms for storing and 3573 retrieving data in the Overlay Instance. These mechanisms can be 3574 used for new applications simply by defining new code points and a 3575 small set of rules. No new protocol mechanisms are required. 3577 The basic unit of stored data is a single StoredData structure: 3579 struct { 3580 uint32 length; 3581 uint64 storage_time; 3582 uint32 lifetime; 3583 StoredDataValue value; 3584 Signature signature; 3585 } StoredData; 3587 The contents of this structure are as follows: 3589 length 3590 The size of the StoredData structure in octets excluding the size 3591 of length itself. 3593 storage_time 3594 The time when the data was stored represented as the number of 3595 milliseconds elapsed since midnight Jan 1, 1970 UTC not counting 3596 leap seconds. This will have the same values for seconds as 3597 standard UNIX time or POSIX time. More information can be found 3598 at [UnixTime]. Any attempt to store a data value with a storage 3599 time before that of a value already stored at this location MUST 3600 generate a Error_Data_Too_Old error. This prevents rollback 3601 attacks. Note that this does not require synchronized clocks: 3602 the receiving peer uses the storage time in the previous store, 3603 not its own clock. 3604 A node that is attempting to store new data in response to a user 3605 request (rather than as an overlay maintenance operation such as 3606 occurs during unpartitioning) is rejected with an 3607 Error_Data_Too_Old error, the node MAY elect to perform its store 3608 using a storage_time that increments the value used with the 3609 previous store. This situation may occur when the clocks of nodes 3610 storing to this location are not properly synchronized. 3612 lifetime 3613 The validity period for the data, in seconds, starting from the 3614 time of store. 3616 value 3617 The data value itself, as described in Section 6.2. 3619 signature 3620 A signature as defined in Section 6.1. 3622 Each Resource-ID specifies a single location in the Overlay Instance. 3623 However, each location may contain multiple StoredData values 3624 distinguished by Kind-ID. The definition of a kind describes both 3625 the data values which may be stored and the data model of the data. 3626 Some data models allow multiple values to be stored under the same 3627 Kind-ID. Section Section 6.2 describes the available data models. 3628 Thus, for instance, a given Resource-ID might contain a single-value 3629 element stored under Kind-ID X and an array containing multiple 3630 values stored under Kind-ID Y. 3632 6.1. Data Signature Computation 3634 Each StoredData element is individually signed. However, the 3635 signature also must be self-contained and cover the Kind-ID and 3636 Resource-ID even though they are not present in the StoredData 3637 structure. The input to the signature algorithm is: 3639 resource_id || kind || storage_time || StoredDataValue || 3640 SignerIdentity 3642 Where || indicates concatenation. 3644 Where these values are: 3646 resource_id 3647 The resource ID where this data is stored. 3649 kind 3650 The Kind-ID for this data. 3652 storage_time 3654 The contents of the storage_time data value. 3655 StoredDataValue 3656 The contents of the stored data value, as described in the 3657 previous sections. 3659 SignerIdentity 3660 The signer identity as defined in Section 5.3.4. 3662 Once the signature has been computed, the signature is represented 3663 using a signature element, as described in Section 5.3.4. 3665 6.2. Data Models 3667 The protocol currently defines the following data models: 3669 o single value 3670 o array 3671 o dictionary 3673 These are represented with the StoredDataValue structure. The actual 3674 dataModel is known from the kind being stored. 3676 struct { 3677 Boolean exists; 3678 opaque value<0..2^32-1>; 3679 } DataValue; 3681 struct { 3682 select (dataModel) { 3683 case single_value: 3684 DataValue single_value_entry; 3686 case array: 3687 ArrayEntry array_entry; 3689 case dictionary: 3690 DictionaryEntry dictionary_entry; 3692 /* This structure may be extended */ 3693 }; 3694 } StoredDataValue; 3696 We now discuss the properties of each data model in turn: 3698 6.2.1. Single Value 3700 A single-value element is a simple sequence of bytes. There may be 3701 only one single-value element for each Resource-ID, Kind-ID pair. 3703 A single value element is represented as a DataValue, which contains 3704 the following two elements: 3706 exists 3707 This value indicates whether the value exists at all. If it is 3708 set to False, it means that no value is present. If it is True, 3709 that means that a value is present. This gives the protocol a 3710 mechanism for indicating nonexistence as opposed to emptiness. 3712 value 3713 The stored data. 3715 6.2.2. Array 3717 An array is a set of opaque values addressed by an integer index. 3718 Arrays are zero based. Note that arrays can be sparse. For 3719 instance, a Store of "X" at index 2 in an empty array produces an 3720 array with the values [ NA, NA, "X"]. Future attempts to fetch 3721 elements at index 0 or 1 will return values with "exists" set to 3722 False. 3724 A array element is represented as an ArrayEntry: 3726 struct { 3727 uint32 index; 3728 DataValue value; 3729 } ArrayEntry; 3731 The contents of this structure are: 3733 index 3734 The index of the data element in the array. 3736 value 3737 The stored data. 3739 6.2.3. Dictionary 3741 A dictionary is a set of opaque values indexed by an opaque key with 3742 one value for each key. A single dictionary entry is represented as 3743 follows: 3745 A dictionary element is represented as a DictionaryEntry: 3747 typedef opaque DictionaryKey<0..2^16-1>; 3749 struct { 3750 DictionaryKey key; 3751 DataValue value; 3752 } DictionaryEntry; 3754 The contents of this structure are: 3756 key 3757 The dictionary key for this value. 3759 value 3760 The stored data. 3762 6.3. Access Control Policies 3764 Every kind which is storable in an overlay MUST be associated with an 3765 access control policy. This policy defines whether a request from a 3766 given node to operate on a given value should succeed or fail. It is 3767 anticipated that only a small number of generic access control 3768 policies are required. To that end, this section describes a small 3769 set of such policies and Section 13.4 establishes a registry for new 3770 policies if required. Each policy has a short string identifier 3771 which is used to reference it in the configuration document. 3773 6.3.1. USER-MATCH 3775 In the USER-MATCH policy, a given value MUST be written (or 3776 overwritten) if and only if the request is signed with a key 3777 associated with a certificate whose user name hashes (using the hash 3778 function for the overlay) to the Resource-ID for the resource. 3779 Recall that the certificate may, depending on the overlay 3780 configuration, be self-signed. 3782 6.3.2. NODE-MATCH 3784 In the NODE-MATCH policy, a given value MUST be written (or 3785 overwritten) if and only if the request is signed with a key 3786 associated with a certificate whose Node-ID hashes (using the hash 3787 function for the overlay) to the Resource-ID for the resource. 3789 6.3.3. USER-NODE-MATCH 3791 The USER-NODE-MATCH policy may only be used with dictionary types. 3792 In the USER-NODE-MATCH policy, a given value MUST be written (or 3793 overwritten) if and only if the request is signed with a key 3794 associated with a certificate whose user name hashes (using the hash 3795 function for the overlay) to the Resource-ID for the resource. In 3796 addition, the dictionary key MUST be equal to the Node-ID in the 3797 certificate. 3799 6.3.4. NODE-MULTIPLE 3801 In the NODE-MULTIPLE policy, a given value MUST be written (or 3802 overwritten) if and only if the request is signed with a key 3803 associated with a certificate containing a Node-ID such that 3804 H(Node-ID || i) is equal to the Resource-ID for some small integer 3805 value of i. When this policy is in use, the maximum value of i MUST 3806 be specified in the kind definition. 3808 Note that as i is not carried on the wire, the verifier MUST iterate 3809 through potential i values up to the maximum value in order to 3810 determine whether a store is acceptable. 3812 6.4. Data Storage Methods 3814 RELOAD provides several methods for storing and retrieving data: 3816 o Store values in the overlay 3817 o Fetch values from the overlay 3818 o Stat: get metadata about values in the overlay 3819 o Find the values stored at an individual peer 3821 These methods are each described in the following sections. 3823 6.4.1. Store 3825 The Store method is used to store data in the overlay. The format of 3826 the Store request depends on the data model which is determined by 3827 the kind. 3829 6.4.1.1. Request Definition 3831 A StoreReq message is a sequence of StoreKindData values, each of 3832 which represents a sequence of stored values for a given kind. The 3833 same Kind-ID MUST NOT be used twice in a given store request. Each 3834 value is then processed in turn. These operations MUST be atomic. 3835 If any operation fails, the state MUST be rolled back to before the 3836 request was received. 3838 The store request is defined by the StoreReq structure: 3840 struct { 3841 KindId kind; 3842 uint64 generation_counter; 3843 StoredData values<0..2^32-1>; 3844 } StoreKindData; 3846 struct { 3847 ResourceId resource; 3848 uint8 replica_number; 3849 StoreKindData kind_data<0..2^32-1>; 3850 } StoreReq; 3852 A single Store request stores data of a number of kinds to a single 3853 resource location. The contents of the structure are: 3855 resource 3856 The resource to store at. 3858 replica_number 3859 The number of this replica. When a storing peer saves replicas to 3860 other peers each peer is assigned a replica number starting from 1 3861 and sent in the Store message. This field is set to 0 when a node 3862 is storing its own data. This allows peers to distinguish replica 3863 writes from original writes. 3865 kind_data 3866 A series of elements, one for each kind of data to be stored. 3868 If the replica number is zero, then the peer MUST check that it is 3869 responsible for the resource and, if not, reject the request. If the 3870 replica number is nonzero, then the peer MUST check that it expects 3871 to be a replica for the resource and that the request sender is 3872 consistent with being the responsible node (i.e., that the receiving 3873 peer does not know of a better node) and, if not, reject the request. 3875 Each StoreKindData element represents the data to be stored for a 3876 single Kind-ID. The contents of the element are: 3878 kind 3879 The Kind-ID. Implementations MUST reject requests corresponding 3880 to unknown kinds. 3882 generation_counter 3883 The expected current state of the generation counter 3884 (approximately the number of times this object has been written; 3885 see below for details). 3887 values 3888 The value or values to be stored. This may contain one or more 3889 stored_data values depending on the data model associated with 3890 each kind. 3892 The peer MUST perform the following checks: 3894 o The Kind-ID is known and supported. 3895 o The signatures over each individual data element (if any) are 3896 valid. If this check fails, the request MUST be rejected with an 3897 Error_Forbidden error. 3898 o Each element is signed by a credential which is authorized to 3899 write this kind at this Resource-ID. If this check fails, the 3900 request MUST be rejected with an Error_Forbidden error. 3902 o For original (non-replica) stores, the peer MUST check that if the 3903 generation counter is non-zero, it equals the current value of the 3904 generation counter for this kind. This feature allows the 3905 generation counter to be used in a way similar to the HTTP Etag 3906 feature. 3907 o For replica Stores, the peer MUST set the generation counter to 3908 match the generation counter in the message, and MUST NOT check 3909 the generation counter against the current value. Replica Stores 3910 MUST NOT use a generation counter of 0. 3911 o The storage time values are greater than that of any value which 3912 would be replaced by this Store. 3913 o The size and number of the stored values is consistent with the 3914 limits specified in the overlay configuration. 3916 If all these checks succeed, the peer MUST attempt to store the data 3917 values. For non-replica stores, if the store succeeds and the data 3918 is changed, then the peer must increase the generation counter by at 3919 least one. If there are multiple stored values in a single 3920 StoreKindData, it is permissible for the peer to increase the 3921 generation counter by only 1 for the entire Kind-ID, or by 1 or more 3922 than one for each value. Accordingly, all stored data values must 3923 have a generation counter of 1 or greater. 0 is used in the Store 3924 request to indicate that the generation counter should be ignored for 3925 processing this request; however the responsible peer should increase 3926 the stored generation counter and should return the correct 3927 generation counter in the response. 3929 When a peer stores data previously stored by another node (e.g., for 3930 replicas or topology shifts) it MUST adjust the lifetime value 3931 downward to reflect the amount of time the value was stored at the 3932 peer. The adjustment SHOULD be implemented by an algorithm 3933 equivalent to the following: at the time the peer initially receives 3934 the StoreReq it notes the local time T. When it then attempts to do a 3935 StoreReq to another node it should decrement the lifetime value by 3936 the difference between the current local time and T. 3938 Unless otherwise specified by the usage, if a peer attempts to store 3939 data previously stored by another node (e.g., for replicas or 3940 topology shifts) and that store fails with either an 3941 Error_Generation_Counter_Too_Low or an Error_Data_Too old error, the 3942 peer MUST fetch the newer data from the peer generating the error and 3943 use that to replace its own copy. This rule allows resynchronization 3944 after partitions heal. 3946 The properties of stores for each data model are as follows: 3948 Single-value: 3949 A store of a new single-value element creates the element if it 3950 does not exist and overwrites any existing value with the new 3951 value. 3953 Array: 3954 A store of an array entry replaces (or inserts) the given value at 3955 the location specified by the index. Because arrays are sparse, a 3956 store past the end of the array extends it with nonexistent values 3957 (exists=False) as required. A store at index 0xffffffff places 3958 the new value at the end of the array regardless of the length of 3959 the array. The resulting StoredData has the correct index value 3960 when it is subsequently fetched. 3962 Dictionary: 3963 A store of a dictionary entry replaces (or inserts) the given 3964 value at the location specified by the dictionary key. 3966 The following figure shows the relationship between these structures 3967 for an example store which stores the following values at resource 3968 "1234" 3970 o The value "abc" in the single value location for kind X 3971 o The value "foo" at index 0 in the array for kind Y 3972 o The value "bar" at index 1 in the array for kind Y 3973 Store 3974 resource=1234 3975 replica_number = 0 3976 / \ 3977 / \ 3978 StoreKindData StoreKindData 3979 kind=X (Single-Value) kind=Y (Array) 3980 generation_counter = 99 generation_counter = 107 3981 | /\ 3982 | / \ 3983 StoredData / \ 3984 storage_time = xxxxxxx / \ 3985 lifetime = 86400 / \ 3986 signature = XXXX / \ 3987 | | | 3988 | StoredData StoredData 3989 | storage_time = storage_time = 3990 | yyyyyyyy zzzzzzz 3991 | lifetime = 86400 lifetime = 33200 3992 | signature = YYYY signature = ZZZZ 3993 | | | 3994 StoredDataValue | | 3995 value="abc" | | 3996 | | 3997 StoredDataValue StoredDataValue 3998 index=0 index=1 3999 value="foo" value="bar" 4001 6.4.1.2. Response Definition 4003 In response to a successful Store request the peer MUST return a 4004 StoreAns message containing a series of StoreKindResponse elements 4005 containing the current value of the generation counter for each 4006 Kind-ID, as well as a list of the peers where the data will be 4007 replicated by the node processing the request. 4009 struct { 4010 KindId kind; 4011 uint64 generation_counter; 4012 NodeId replicas<0..2^16-1>; 4013 } StoreKindResponse; 4015 struct { 4016 StoreKindResponse kind_responses<0..2^16-1>; 4017 } StoreAns; 4019 The contents of each StoreKindResponse are: 4021 kind 4022 The Kind-ID being represented. 4024 generation_counter 4025 The current value of the generation counter for that Kind-ID. 4027 replicas 4028 The list of other peers at which the data was/will be replicated. 4029 In overlays and applications where the responsible peer is 4030 intended to store redundant copies, this allows the storing peer 4031 to independently verify that the replicas have in fact been 4032 stored. It does this verification by using the Stat method. Note 4033 that the storing peer is not required to perform this 4034 verification. 4036 The response itself is just StoreKindResponse values packed end-to- 4037 end. 4039 If any of the generation counters in the request precede the 4040 corresponding stored generation counter, then the peer MUST fail the 4041 entire request and respond with an Error_Generation_Counter_Too_Low 4042 error. The error_info in the ErrorResponse MUST be a StoreAns 4043 response containing the correct generation counter for each kind and 4044 the replica list, which will be empty. For original (non-replica) 4045 stores, a node which receives such an error SHOULD attempt to fetch 4046 the data and, if the storage_time value is newer, replace its own 4047 data with that newer data. This rule improves data consistency in 4048 the case of partitions and merges. 4050 If the data being stored is too large for the allowed limit by the 4051 given usage, then the peer MUST fail the request and generate an 4052 Error_Data_Too_Large error. 4054 If any type of request tries to access a data kind that the node does 4055 not know about, an Error_Unknown_Kind MUST be generated. The 4056 error_info in the Error_Response is: 4058 KindId unknown_kinds<0..2^8-1>; 4060 which lists all the kinds that were unrecognized. A node which 4061 receives this error MUST generate a ConfigUpdate message which 4062 contains the appropriate kind definition (assuming that in fact a 4063 kind was used which was defined in the configuration document). 4065 6.4.1.3. Removing Values 4067 This version of RELOAD (unlike previous versions) does not have an 4068 explicit Remove operation. Rather, values are Removed by storing 4069 "nonexistent" values in their place. Each DataValue contains a 4070 boolean value called "exists" which indicates whether a value is 4071 present at that location. In order to effectively remove a value, 4072 the owner stores a new DataValue with: 4074 exists = false 4075 value = {} (0 length) 4077 Storing nodes MUST treat these nonexistent values the same way they 4078 treat any other stored value, including overwriting the existing 4079 value, replicating them, and aging them out as necessary when 4080 lifetime expires. When a stored nonexistent value's lifetime 4081 expires, it is simply removed from the storing node like any other 4082 stored value expiration. Note that in the case of arrays and 4083 dictionaries, this may create an implicit, unsigned "nonexistent" 4084 value to represent a gap in the data structure. However, this value 4085 isn't persistent nor is it replicated. It is simply synthesized by 4086 the storing node. 4088 6.4.2. Fetch 4090 The Fetch request retrieves one or more data elements stored at a 4091 given Resource-ID. A single Fetch request can retrieve multiple 4092 different kinds. 4094 6.4.2.1. Request Definition 4096 struct { 4097 int32 first; 4098 int32 last; 4099 } ArrayRange; 4101 struct { 4102 KindId kind; 4103 uint64 generation; 4104 uint16 length; 4106 select (dataModel) { 4107 case single_value: ; /* Empty */ 4109 case array: 4110 ArrayRange indices<0..2^16-1>; 4112 case dictionary: 4113 DictionaryKey keys<0..2^16-1>; 4115 /* This structure may be extended */ 4117 } model_specifier; 4118 } StoredDataSpecifier; 4120 struct { 4121 ResourceId resource; 4122 StoredDataSpecifier specifiers<0..2^16-1>; 4123 } FetchReq; 4125 The contents of the Fetch requests are as follows: 4127 resource 4128 The Resource-ID to fetch from. 4130 specifiers 4131 A sequence of StoredDataSpecifier values, each specifying some of 4132 the data values to retrieve. 4134 Each StoredDataSpecifier specifies a single kind of data to retrieve 4135 and (if appropriate) the subset of values that are to be retrieved. 4136 The contents of the StoredDataSpecifier structure are as follows: 4138 kind 4139 The Kind-ID of the data being fetched. Implementations SHOULD 4140 reject requests corresponding to unknown kinds unless specifically 4141 configured otherwise. 4143 dataModel 4144 The data model of the data. This is not transmitted on the wire 4145 but comes from the definition of the kind. 4147 generation 4148 The last generation counter that the requesting node saw. This 4149 may be used to avoid unnecessary fetches or it may be set to zero. 4151 length 4152 The length of the rest of the structure, thus allowing 4153 extensibility. 4155 model_specifier 4156 A reference to the data value being requested within the data 4157 model specified for the kind. For instance, if the data model is 4158 "array", it might specify some subset of the values. 4160 The model_specifier is as follows: 4162 o If the data model is single value, the specifier is empty. 4163 o If the data model is array, the specifier contains a list of 4164 ArrayRange elements, each of which contains two integers. The 4165 first integer is the beginning of the range and the second is the 4166 end of the range. 0 is used to indicate the first element and 4167 0xffffffff is used to indicate the final element. The first 4168 integer must be less than the second. While multiple ranges MAY 4169 be specified, they MUST NOT overlap. 4170 o If the data model is dictionary then the specifier contains a list 4171 of the dictionary keys being requested. If no keys are specified, 4172 than this is a wildcard fetch and all key-value pairs are 4173 returned. 4175 The generation counter is used to indicate the requester's expected 4176 state of the storing peer. If the generation counter in the request 4177 matches the stored counter, then the storing peer returns a response 4178 with no StoredData values. 4180 Note that because the certificate for a user is typically stored at 4181 the same location as any data stored for that user, a requesting node 4182 that does not already have the user's certificate should request the 4183 certificate in the Fetch as an optimization. 4185 6.4.2.2. Response Definition 4187 The response to a successful Fetch request is a FetchAns message 4188 containing the data requested by the requester. 4190 struct { 4191 KindId kind; 4192 uint64 generation; 4193 StoredData values<0..2^32-1>; 4194 } FetchKindResponse; 4196 struct { 4197 FetchKindResponse kind_responses<0..2^32-1>; 4198 } FetchAns; 4200 The FetchAns structure contains a series of FetchKindResponse 4201 structures. There MUST be one FetchKindResponse element for each 4202 Kind-ID in the request. 4204 The contents of the FetchKindResponse structure are as follows: 4206 kind 4207 the kind that this structure is for. 4209 generation 4210 the generation counter for this kind. 4212 values 4213 the relevant values. If the generation counter in the request 4214 matches the generation counter in the stored data, then no 4215 StoredData values are returned. Otherwise, all relevant data 4216 values MUST be returned. A nonexistent value is represented with 4217 "exists" set to False. 4219 There is one subtle point about signature computation on arrays. If 4220 the storing node uses the append feature (where the 4221 index=0xffffffff), then the index in the StoredData that is returned 4222 will not match that used by the storing node, which would break the 4223 signature. In order to avoid this issue, the index value in the 4224 array is set to zero before the signature is computed. This implies 4225 that malicious storing nodes can reorder array entries without being 4226 detected. 4228 6.4.3. Stat 4230 The Stat request is used to get metadata (length, generation counter, 4231 digest, etc.) for a stored element without retrieving the element 4232 itself. The name is from the UNIX stat(2) system call which performs 4233 a similar function for files in a file system. It also allows the 4234 requesting node to get a list of matching elements without requesting 4235 the entire element. 4237 6.4.3.1. Request Definition 4239 The Stat request is identical to the Fetch request. It simply 4240 specifies the elements to get metadata about. 4242 struct { 4243 ResourceId resource; 4244 StoredDataSpecifier specifiers<0..2^16-1>; 4245 } StatReq; 4247 6.4.3.2. Response Definition 4249 The Stat response contains the same sort of entries that a Fetch 4250 response would contain; however, instead of containing the element 4251 data it contains metadata. 4253 struct { 4254 Boolean exists; 4255 uint32 value_length; 4256 HashAlgorithm hash_algorithm; 4257 opaque hash_value<0..255>; 4258 } MetaData; 4260 struct { 4261 uint32 index; 4262 MetaData value; 4263 } ArrayEntryMeta; 4265 struct { 4266 DictionaryKey key; 4267 MetaData value; 4268 } DictionaryEntryMeta; 4270 struct { 4271 select (model) { 4272 case single_value: 4273 MetaData single_value_entry; 4275 case array: 4276 ArrayEntryMeta array_entry; 4278 case dictionary: 4279 DictionaryEntryMeta dictionary_entry; 4281 /* This structure may be extended */ 4282 }; 4283 } MetaDataValue; 4285 struct { 4286 uint32 value_length; 4287 uint64 storage_time; 4288 uint32 lifetime; 4289 MetaDataValue metadata; 4290 } StoredMetaData; 4292 struct { 4293 KindId kind; 4294 uint64 generation; 4295 StoredMetaData values<0..2^32-1>; 4296 } StatKindResponse; 4298 struct { 4299 StatKindResponse kind_responses<0..2^32-1>; 4300 } StatAns; 4302 The structures used in StatAns parallel those used in FetchAns: a 4303 response consists of multiple StatKindResponse values, one for each 4304 kind that was in the request. The contents of the StatKindResponse 4305 are the same as those in the FetchKindResponse, except that the 4306 values list contains StoredMetaData entries instead of StoredData 4307 entries. 4309 The contents of the StoredMetaData structure are the same as the 4310 corresponding fields in StoredData except that there is no signature 4311 field and the value is a MetaDataValue rather than a StoredDataValue. 4313 A MetaDataValue is a variant structure, like a StoredDataValue, 4314 except for the types of each arm, which replace DataValue with 4315 MetaData. 4317 The only really new structure is MetaData, which has the following 4318 contents: 4320 exists 4321 Same as in DataValue 4323 value_length 4324 The length of the stored value. 4326 hash_algorithm 4327 The hash algorithm used to perform the digest of the value. 4329 hash_value 4330 A digest of the value using hash_algorithm. 4332 6.4.4. Find 4334 The Find request can be used to explore the Overlay Instance. A Find 4335 request for a Resource-ID R and a Kind-ID T retrieves the Resource-ID 4336 (if any) of the resource of kind T known to the target peer which is 4337 closest to R. This method can be used to walk the Overlay Instance by 4338 iteratively fetching R_n+1=nearest(1 + R_n). 4340 6.4.4.1. Request Definition 4342 The FindReq message contains a Resource-ID and a series of Kind-IDs 4343 identifying the resource the peer is interested in. 4345 struct { 4346 ResourceId resource; 4347 KindId kinds<0..2^8-1>; 4348 } FindReq; 4350 The request contains a list of Kind-IDs which the Find is for, as 4351 indicated below: 4353 resource 4354 The desired Resource-ID 4356 kinds 4357 The desired Kind-IDs. Each value MUST only appear once, and if 4358 not the request MUST be rejected with an error. 4360 6.4.4.2. Response Definition 4362 A response to a successful Find request is a FindAns message 4363 containing the closest Resource-ID on the peer for each kind 4364 specified in the request. 4366 struct { 4367 KindId kind; 4368 ResourceId closest; 4369 } FindKindData; 4371 struct { 4372 FindKindData results<0..2^16-1>; 4373 } FindAns; 4375 If the processing peer is not responsible for the specified 4376 Resource-ID, it SHOULD return an Error_Not_Found error code. 4378 For each Kind-ID in the request the response MUST contain a 4379 FindKindData indicating the closest Resource-ID for that Kind-ID, 4380 unless the kind is not allowed to be used with Find in which case a 4381 FindKindData for that Kind-ID MUST NOT be included in the response. 4382 If a Kind-ID is not known, then the corresponding Resource-ID MUST be 4383 0. Note that different Kind-IDs may have different closest Resource- 4384 IDs. 4386 The response is simply a series of FindKindData elements, one per 4387 kind, concatenated end-to-end. The contents of each element are: 4389 kind 4390 The Kind-ID. 4392 closest 4393 The closest resource ID to the specified resource ID. This is 0 4394 if no resource ID is known. 4396 Note that the response does not contain the contents of the data 4397 stored at these Resource-IDs. If the requester wants this, it must 4398 retrieve it using Fetch. 4400 6.4.5. Defining New Kinds 4402 There are two ways to define a new kind. The first is by writing a 4403 document and registering the kind-id with IANA. This is the 4404 preferred method for kinds which may be widely used and reused. The 4405 second method is to simply define the kind and its parameters in the 4406 configuration document using the section of kind-id space set aside 4407 for private use. This method MAY be used to define ad hoc kinds in 4408 new overlays. 4410 However a kind is defined, the definition must include: 4412 o The meaning of the data to be stored (in some textual form). 4413 o The Kind-ID. 4414 o The data model (single value, array, dictionary, etc). 4415 o The access control model. 4417 In addition, when kinds are registered with IANA, each kind is 4418 assigned a short string name which is used to refer to it in 4419 configuration documents. 4421 While each kind needs to define what data model is used for its data, 4422 that does not mean that it must define new data models. Where 4423 practical, kinds should use the existing data models. The intention 4424 is that the basic data model set be sufficient for most applications/ 4425 usages. 4427 7. Certificate Store Usage 4429 The Certificate Store usage allows a peer to store its certificate in 4430 the overlay, thus avoiding the need to send a certificate in each 4431 message - a reference may be sent instead. 4433 A user/peer MUST store its certificate at Resource-IDs derived from 4434 two Resource Names: 4436 o The user name in the certificate. 4438 o The Node-ID in the certificate. 4440 Note that in the second case the certificate is not stored at the 4441 peer's Node-ID but rather at a hash of the peer's Node-ID. The 4442 intention here (as is common throughout RELOAD) is to avoid making a 4443 peer responsible for its own data. 4445 A peer MUST ensure that the user's certificates are stored in the 4446 Overlay Instance. New certificates are stored at the end of the 4447 list. This structure allows users to store an old and a new 4448 certificate that both have the same Node-ID, which allows for 4449 migration of certificates when they are renewed. 4451 This usage defines the following kinds: 4453 Name: CERTIFICATE_BY_NODE 4455 Data Model: The data model for CERTIFICATE_BY_NODE data is array. 4457 Access Control: NODE-MATCH. 4459 Name: CERTIFICATE_BY_USER 4461 Data Model: The data model for CERTIFICATE_BY_USER data is array. 4463 Access Control: USER-MATCH. 4465 8. TURN Server Usage 4467 The TURN server usage allows a RELOAD peer to advertise that it is 4468 prepared to be a TURN server as defined in [RFC5766]. When a node 4469 starts up, it joins the overlay network and forms several connections 4470 in the process. If the ICE stage in any of these connections returns 4471 a reflexive address that is not the same as the peer's perceived 4472 address, then the peer is behind a NAT and not a candidate for a TURN 4473 server. Additionally, if the peer's IP address is in the private 4474 address space range, then it is also not a candidate for a TURN 4475 server. Otherwise, the peer SHOULD assume it is a potential TURN 4476 server and follow the procedures below. 4478 If the node is a candidate for a TURN server it will insert some 4479 pointers in the overlay so that other peers can find it. The overlay 4480 configuration file specifies a turn-density parameter that indicates 4481 how many times each TURN server should record itself in the overlay. 4482 Typically this should be set to the reciprocal of the estimate of 4483 what percentage of peers will act as TURN servers. If the turn- 4484 density is not set to zero, for each value, called d, between 1 and 4485 turn-density, the peer forms a Resource Name by concatenating its 4486 Node-ID and the value d. This Resource Name is hashed to form a 4487 Resource-ID. The address of the peer is stored at that Resource-ID 4488 using type TURN-SERVICE and the TurnServer object: 4490 struct { 4491 uint8 iteration; 4492 IpAddressAndPort server_address; 4493 } TurnServer; 4495 The contents of this structure are as follows: 4497 iteration 4498 the d value 4500 server_address 4501 the address at which the TURN server can be contacted. 4503 Note: Correct functioning of this algorithm depends on having turn- 4504 density be an reasonable estimate of the reciprocal of the 4505 proportion of nodes in the overlay that can act as TURN servers. 4506 If the turn-density value in the configuration file is too low, 4507 then the process of finding TURN servers becomes more expensive as 4508 multiple candidate Resource-IDs must be probed to find a TURN 4509 server. 4511 Peers that provide this service need to support the TURN extensions 4512 to STUN for media relay as defined in [RFC5766]. 4514 This usage defines the following kind to indicate that a peer is 4515 willing to act as a TURN server: 4517 Name TURN-SERVICE 4518 Data Model The TURN-SERVICE kind stores a single value for each 4519 Resource-ID. 4520 Access Control NODE-MULTIPLE, with maximum iteration counter 20. 4522 Peers can find other servers by selecting a random Resource-ID and 4523 then doing a Find request for the appropriate Kind-ID with that 4524 Resource-ID. The Find request gets routed to a random peer based on 4525 the Resource-ID. If that peer knows of any servers, they will be 4526 returned. The returned response may be empty if the peer does not 4527 know of any servers, in which case the process gets repeated with 4528 some other random Resource-ID. As long as the ratio of servers 4529 relative to peers is not too low, this approach will result in 4530 finding a server relatively quickly. 4532 9. Chord Algorithm 4534 This algorithm is assigned the name chord-reload to indicate it is an 4535 adaptation of the basic Chord DHT algorithm. 4537 This algorithm differs from the originally presented Chord algorithm 4538 [Chord]. It has been updated based on more recent research results 4539 and implementation experiences, and to adapt it to the RELOAD 4540 protocol. A short list of differences: 4542 o The original Chord algorithm specified that a single predecessor 4543 and a successor list be stored. The chord-reload algorithm 4544 attempts to have more than one predecessor and successor. The 4545 predecessor sets help other neighbors learn their successor list. 4546 o The original Chord specification and analysis called for iterative 4547 routing. RELOAD specifies recursive routing. In addition to the 4548 performance implications, the cost of NAT traversal dictates 4549 recursive routing. 4550 o Finger table entries are indexed in opposite order. Original 4551 Chord specifies finger[0] as the immediate successor of the peer. 4552 chord-reload specifies finger[0] as the peer 180 degrees around 4553 the ring from the peer. This change was made to simplify 4554 discussion and implementation of variable sized finger tables. 4555 However, with either approach no more than O(log N) entries should 4556 typically be stored in a finger table. 4557 o The stabilize() and fix_fingers() algorithms in the original Chord 4558 algorithm are merged into a single periodic process. 4559 Stabilization is implemented slightly differently because of the 4560 larger neighborhood, and fix_fingers is not as aggressive to 4561 reduce load, nor does it search for optimal matches of the finger 4562 table entries. 4563 o RELOAD uses a 128 bit hash instead of a 160 bit hash, as RELOAD is 4564 not designed to be used in networks with close to or more than 4565 2^128 nodes (and it is hard to see how one would assemble such a 4566 network). 4567 o RELOAD uses randomized finger entries as described in 4568 Section 9.7.4.2. 4569 o This algorithm allows the use of either reactive or periodic 4570 recovery. The original Chord paper used periodic recovery. 4571 Reactive recovery provides better performance in small overlays, 4572 but is believed to be unstable in large (>1000) overlays with high 4573 levels of churn [handling-churn-usenix04]. The overlay 4574 configuration file specifies a "chord-reactive" element that 4575 indicates whether reactive recovery should be used. 4577 9.1. Overview 4579 The algorithm described here is a modified version of the Chord 4580 algorithm. Each peer keeps track of a finger table and a neighbor 4581 table. The neighbor table contains at least the three peers before 4582 and after this peer in the DHT ring. There may not be three entries 4583 in all cases such as small rings or while the ring topology is 4584 changing. The first entry in the finger table contains the peer 4585 half-way around the ring from this peer; the second entry contains 4586 the peer that is 1/4 of the way around; the third entry contains the 4587 peer that is 1/8th of the way around, and so on. Fundamentally, the 4588 chord data structure can be thought of a doubly-linked list formed by 4589 knowing the successors and predecessor peers in the neighbor table, 4590 sorted by the Node-ID. As long as the successor peers are correct, 4591 the DHT will return the correct result. The pointers to the prior 4592 peers are kept to enable the insertion of new peers into the list 4593 structure. Keeping multiple predecessor and successor pointers makes 4594 it possible to maintain the integrity of the data structure even when 4595 consecutive peers simultaneously fail. The finger table forms a skip 4596 list, so that entries in the linked list can be found in O(log(N)) 4597 time instead of the typical O(N) time that a linked list would 4598 provide. 4600 A peer, n, is responsible for a particular Resource-ID k if k is less 4601 than or equal to n and k is greater than p, where p is the Node-ID of 4602 the previous peer in the neighbor table. Care must be taken when 4603 computing to note that all math is modulo 2^128. 4605 9.2. Hash Function 4607 For this Chord topology plugin, the size of the Resource-ID is 128 4608 bits. The hash of a Resource-ID is computed using SHA-1 4609 [RFC3174]then truncating the SHA-1 result to the most significant 128 4610 bits. 4612 9.3. Routing 4614 The routing table is the union of the neighbor table and the finger 4615 table. 4617 If a peer is not responsible for a Resource-ID k, but is directly 4618 connected to a node with Node-ID k, then it routes the message to 4619 that node. Otherwise, it routes the request to the peer in the 4620 routing table that has the largest Node-ID that is in the interval 4621 between the peer and k. If no such node is found, it finds the 4622 smallest Node-Id that is greater than k and routes the message to 4623 that node. 4625 9.4. Redundancy 4627 When a peer receives a Store request for Resource-ID k, and it is 4628 responsible for Resource-ID k, it stores the data and returns a 4629 success response. It then sends a Store request to its successor in 4630 the neighbor table and to that peer's successor. Note that these 4631 Store requests are addressed to those specific peers, even though the 4632 Resource-ID they are being asked to store is outside the range that 4633 they are responsible for. The peers receiving these check they came 4634 from an appropriate predecessor in their neighbor table and that they 4635 are in a range that this predecessor is responsible for, and then 4636 they store the data. They do not themselves perform further Stores 4637 because they can determine that they are not responsible for the 4638 Resource-ID. 4640 Managing replicas as the overlay changes is described in 4641 Section 9.7.3. 4643 The sequential replicas used in this overlay algorithm protect 4644 against peer failure but not against malicious peers. Additional 4645 replication from the Usage is required to protect resources from such 4646 attacks, as discussed in Section 12.5.4. 4648 9.5. Joining 4650 The join process for a joining party (JP) with Node-ID n is as 4651 follows. 4653 1. JP MUST connect to its chosen bootstrap node. 4654 2. JP SHOULD send an Attach request to the admitting peer (AP) for 4655 Node-ID n. The "send_update" flag should be used to acquire the 4656 routing table for AP. 4657 3. JP SHOULD send Attach requests to initiate connections to each of 4658 the peers in the neighbor table as well as to the desired finger 4659 table entries. Note that this does not populate their routing 4660 tables, but only their connection tables, so JP will not get 4661 messages that it is expected to route to other nodes. 4662 4. JP MUST enter all the peers it has contacted into its routing 4663 table. 4664 5. JP MUST send a Join to AP. The AP sends the response to the 4665 Join. 4666 6. AP MUST do a series of Store requests to JP to store the data 4667 that JP will be responsible for. 4668 7. AP MUST send JP an Update explicitly labeling JP as its 4669 predecessor. At this point, JP is part of the ring and 4670 responsible for a section of the overlay. AP can now forget any 4671 data which is assigned to JP and not AP. 4673 8. The AP MUST send an Update to all of its neighbors with the new 4674 values of its neighbor set (including JP). 4675 9. The JP MUST send Updates to all the peers in its neighbor table. 4677 If JP sends an Attach to AP with send_update, it immediately knows 4678 most of its expected neighbors from AP's routing table update and can 4679 directly connect to them. This is the RECOMMENDED procedure. 4681 If for some reason JP does not get AP's routing table, it can still 4682 populate its neighbor table incrementally. It sends a Ping directed 4683 at Resource-ID n+1 (directly after its own Resource-ID). This allows 4684 it to discover its own successor. Call that node p0. It then sends 4685 a ping to p0+1 to discover its successor (p1). This process can be 4686 repeated to discover as many successors as desired. The values for 4687 the two peers before p will be found at a later stage when n receives 4688 an Update. An alternate procedure is to send Attaches to those nodes 4689 rather than pings, which forms the connections immediately but may be 4690 slower if the nodes need to collect ICE candidates, thus reducing 4691 parallelism. 4693 In order to set up its finger table entry for peer i, JP simply sends 4694 an Attach to peer (n+2^(128-i). This will be routed to a peer in 4695 approximately the right location around the ring. 4697 The joining peer MUST NOT send any Update message placing itself in 4698 the overlay until it has successfully completed an Attach with each 4699 peer that should be in its neighbor table. 4701 9.6. Routing Attaches 4703 When a peer needs to Attach to a new peer in its neighbor table, it 4704 MUST source-route the Attach request through the peer from which it 4705 learned the new peer's Node-ID. Source-routing these requests allows 4706 the overlay to recover from instability. 4708 All other Attach requests, such as those for new finger table 4709 entries, are routed conventionally through the overlay. 4711 9.7. Updates 4713 A chord Update is defined as 4714 enum { reserved (0), 4715 peer_ready(1), neighbors(2), full(3), (255) } 4716 ChordUpdateType; 4718 struct { 4719 uint32 uptime; 4720 ChordUpdateType type; 4721 select(type){ 4722 case peer_ready: /* Empty */ 4723 ; 4725 case neighbors: 4726 NodeId predecessors<0..2^16-1>; 4727 NodeId successors<0..2^16-1>; 4729 case full: 4730 NodeId predecessors<0..2^16-1>; 4731 NodeId successors<0..2^16-1>; 4732 NodeId fingers<0..2^16-1>; 4733 }; 4734 } ChordUpdate; 4736 The "uptime" field contains the time this peer has been up in 4737 seconds. 4739 The "type" field contains the type of the update, which depends on 4740 the reason the update was sent. 4742 peer_ready: this peer is ready to receive messages. This message 4743 is used to indicate that a node which has Attached is a peer and 4744 can be routed through. It is also used as a connectivity check to 4745 non-neighbor peers. 4747 neighbors: this version is sent to members of the Chord neighbor 4748 table. 4750 full: this version is sent to peers which request an Update with a 4751 RouteQueryReq. 4753 If the message is of type "neighbors", then the contents of the 4754 message will be: 4756 predecessors 4757 The predecessor set of the Updating peer. 4759 successors 4760 The successor set of the Updating peer. 4762 If the message is of type "full", then the contents of the message 4763 will be: 4765 predecessors 4766 The predecessor set of the Updating peer. 4768 successors 4769 The successor set of the Updating peer. 4771 fingers 4772 The finger table of the Updating peer, in numerically ascending 4773 order. 4775 A peer MUST maintain an association (via Attach) to every member of 4776 its neighbor set. A peer MUST attempt to maintain at least three 4777 predecessors and three successors, even though this will not be 4778 possible if the ring is very small. It is RECOMMENDED that O(log(N)) 4779 predecessors and successors be maintained in the neighbor set. 4781 9.7.1. Handling Neighbor Failures 4783 Every time a connection to a peer in the neighbor table is lost (as 4784 determined by connectivity pings or the failure of some request), the 4785 peer MUST remove the entry from its neighbor table and replace it 4786 with the best match it has from the other peers in its routing table. 4787 If using reactive recovery, it then sends an immediate Update to all 4788 nodes in its Neighbor Table. The update will contain all the Node- 4789 IDs of the current entries of the table (after the failed one has 4790 been removed). Note that when replacing a successor the peer SHOULD 4791 delay the creation of new replicas for successor replacement hold- 4792 down time (30 seconds) after removing the failed entry from its 4793 neighbor table in order to allow a triggered update to inform it of a 4794 better match for its neighbor table. 4796 If the neighbor failure effects the peer's range of responsible IDs, 4797 then the Update MUST be sent to all nodes in its Connection Table. 4799 A peer MAY attempt to reestablish connectivity with a lost neighbor 4800 either by waiting additional time to see if connectivity returns or 4801 by actively routing a new Attach to the lost peer. Details for these 4802 procedures are beyond the scope of this document. In no event does 4803 an attempt to reestablish connectivity with a lost neighbor allow the 4804 peer to remain in the neighbor table. Such a peer is returned to the 4805 neighbor table once connectivity is reestablished. 4807 If connectivity is lost to all successor peers in the neighbor table, 4808 then this peer should behave as if it is joining the network and use 4809 Pings to find a peer and send it a Join. If connectivity is lost to 4810 all the peers in the finger table, this peer should assume that it 4811 has been disconnected from the rest of the network, and it should 4812 periodically try to join the DHT. 4814 9.7.2. Handling Finger Table Entry Failure 4816 If a finger table entry is found to have failed, all references to 4817 the failed peer are removed from the finger table and replaced with 4818 the closest preceding peer from the finger table or neighbor table. 4820 If using reactive recovery, the peer initiates a search for a new 4821 finger table entry as described below. 4823 9.7.3. Receiving Updates 4825 When a peer, N, receives an Update request, it examines the Node-IDs 4826 in the UpdateReq and at its neighbor table and decides if this 4827 UpdateReq would change its neighbor table. This is done by taking 4828 the set of peers currently in the neighbor table and comparing them 4829 to the peers in the update request. There are two major cases: 4831 o The UpdateReq contains peers that match N's neighbor table, so no 4832 change is needed to the neighbor set. 4833 o The UpdateReq contains peers N does not know about that should be 4834 in N's neighbor table, i.e. they are closer than entries in the 4835 neighbor table. 4837 In the first case, no change is needed. 4839 In the second case, N MUST attempt to Attach to the new peers and if 4840 it is successful it MUST adjust its neighbor set accordingly. Note 4841 that it can maintain the now inferior peers as neighbors, but it MUST 4842 remember the closer ones. 4844 After any Pings and Attaches are done, if the neighbor table changes 4845 and the peer is using reactive recovery, the peer sends an Update 4846 request to each member of its Connection Table. These Update 4847 requests are what end up filling in the predecessor/successor tables 4848 of peers that this peer is a neighbor to. A peer MUST NOT enter 4849 itself in its successor or predecessor table and instead should leave 4850 the entries empty. 4852 If peer N is responsible for a Resource-ID R, and N discovers that 4853 the replica set for R (the next two nodes in its successor set) has 4854 changed, it MUST send a Store for any data associated with R to any 4855 new node in the replica set. It SHOULD NOT delete data from peers 4856 which have left the replica set. 4858 When a peer N detects that it is no longer in the replica set for a 4859 resource R (i.e., there are three predecessors between N and R), it 4860 SHOULD delete all data associated with R from its local store. 4862 When a peer discovers that its range of responsible IDs have changed, 4863 it MUST send an Update to all entries in its connection table. 4865 9.7.4. Stabilization 4867 There are four components to stabilization: 4868 1. exchange Updates with all peers in its neighbor table to exchange 4869 state. 4870 2. search for better peers to place in its finger table. 4871 3. search to determine if the current finger table size is 4872 sufficiently large. 4873 4. search to determine if the overlay has partitioned and needs to 4874 recover. 4876 9.7.4.1. Updating neighbor table 4878 A peer MUST periodically send an Update request to every peer in its 4879 Connection Table. The purpose of this is to keep the predecessor and 4880 successor lists up to date and to detect failed peers. The default 4881 time is about every ten minutes, but the configuration server SHOULD 4882 set this in the configuration document using the "chord-update- 4883 interval" element (denominated in seconds.) A peer SHOULD randomly 4884 offset these Update requests so they do not occur all at once. 4886 9.7.4.2. Refreshing finger table 4888 A peer MUST periodically search for new peers to replace invalid 4889 (repeated) entries in the finger table. A finger table entry i is 4890 valid if it is in the range [n+2^(128-i), 4891 n+2^(128-(i-1))-2^(128-(i+1))]. Invalid entries occur in the finger 4892 table when a previous finger table entry has failed or when no peer 4893 has been found in that range. 4895 A peer SHOULD NOT send Ping requests looking for new finger table 4896 entries more often than the configuration element "chord-ping- 4897 interval", which defaults to 3600 seconds (one per hour). 4899 Two possible methods for searching for new peers for the finger table 4900 entries are presented: 4902 Alternative 1: A peer selects one entry in the finger table from 4903 among the invalid entries. It pings for a new peer for that finger 4904 table entry. The selection SHOULD be exponentially weighted to 4905 attempt to replace earlier (lower i) entries in the finger table. A 4906 simple way to implement this selection is to search through the 4907 finger table entries from i=0 and each time an invalid entry is 4908 encountered, send a Ping to replace that entry with probability 0.5. 4910 Alternative 2: A peer monitors the Update messages received from its 4911 connections to observe when an Update indicates a peer that would be 4912 used to replace in invalid finger table entry, i, and flags that 4913 entry in the finger table. Every "chord-ping-interval" seconds, the 4914 peer selects from among those flagged candidates using an 4915 exponentially weighted probability as above. 4917 When searching for a better entry, the peer SHOULD send the Ping to a 4918 Node-ID selected randomly from that range. Random selection is 4919 preferred over a search for strictly spaced entries to minimize the 4920 effect of churn on overlay routing [minimizing-churn-sigcomm06]. An 4921 implementation or subsequent specification MAY choose a method for 4922 selecting finger table entries other than choosing randomly within 4923 the range. Any such alternate methods SHOULD be employed only on 4924 finger table stabilization and not for the selection of initial 4925 finger table entries unless the alternative method is faster and 4926 imposes less overhead on the overlay. 4928 A peer MAY choose to keep connections to multiple peers that can act 4929 for a given finger table entry. 4931 9.7.4.3. Adjusting finger table size 4933 If the finger table has less than 16 entries, the node SHOULD attempt 4934 to discover more fingers to grow the size of the table to 16. The 4935 value 16 was chosen to ensure high odds of a node maintaining 4936 connectivity to the overlay even with strange network partitions. 4938 For many overlays, 16 finger table entries will be enough, but as an 4939 overlay grows very large, more than 16 entries may be required in the 4940 finger table for efficient routing. An implementation SHOULD be 4941 capable of increasing the number of entries in the finger table to 4942 128 entries. 4944 Note to implementers: Although log(N) entries are all that are 4945 required for optimal performance, careful implementation of 4946 stabilization will result in no additional traffic being generated 4947 when maintaining a finger table larger than log(N) entries. 4948 Implementers are encouraged to make use of RouteQuery and algorithms 4949 for determining where new finger table entries may be found. 4950 Complete details of possible implementations are outside the scope of 4951 this specification. 4953 A simple approach to sizing the finger table is to ensure the finger 4954 table is large enough to contain at least the final successor in the 4955 peer's neighbor table. 4957 9.7.4.4. Detecting partitioning 4959 To detect that a partitioning has occurred and to heal the overlay, a 4960 peer P MUST periodically repeat the discovery process used in the 4961 initial join for the overlay to locate an appropriate bootstrap node, 4962 B. P should then send a Ping for its own Node-ID routed through B. If 4963 a response is received from a peer S', which is not P's successor, 4964 then the overlay is partitioned and P should send an Attach to S' 4965 routed through B, followed by an Update sent to S'. (Note that S' 4966 may not be in P's neighbor table once the overlay is healed, but the 4967 connection will allow S' to discover appropriate neighbor entries for 4968 itself via its own stabilization.) 4970 Future specifications may describe alternative mechanisms for 4971 determining when to repeat the discovery process. 4973 9.8. Route query 4975 For this topology plugin, the RouteQueryReq contains no additional 4976 information. The RouteQueryAns contains the single node ID of the 4977 next peer to which the responding peer would have routed the request 4978 message in recursive routing: 4980 struct { 4981 NodeId next_peer; 4982 } ChordRouteQueryAns; 4984 The contents of this structure are as follows: 4986 next_peer 4987 The peer to which the responding peer would route the message in 4988 order to deliver it to the destination listed in the request. 4990 If the requester has set the send_update flag, the responder SHOULD 4991 initiate an Update immediately after sending the RouteQueryAns. 4993 9.9. Leaving 4995 To support extensions, such as [I-D.maenpaa-p2psip-self-tuning], 4996 Peers SHOULD send a Leave request to all members of their neighbor 4997 table prior to exiting the Overlay Instance. The 4998 overlay_specific_data field MUST contain the ChordLeaveData structure 4999 defined below: 5001 enum { reserved (0), 5002 from_succ(1), from_pred(2), (255) } 5003 ChordLeaveType; 5005 struct { 5006 ChordLeaveType type; 5008 select(type) { 5009 case from_succ: 5010 NodeId successors<0..2^16-1>; 5011 case from_pred: 5012 NodeId predecessors<0..2^16-1>; 5013 }; 5014 } ChordLeaveData; 5016 The 'type' field indicates whether the Leave request was sent by a 5017 predecessor or a successor of the recipient: 5019 from_succ 5020 The Leave request was sent by a successor. 5022 from_pred 5023 The Leave request was sent by a predecessor. 5025 If the type of the request is 'from_succ', the contents will be: 5027 successors 5028 The sender's successor list. 5030 If the type of the request is 'from_pred', the contents will be: 5032 predecessors 5033 The sender's predecessor list. 5035 Any peer which receives a Leave for a peer n in its neighbor set 5036 follows procedures as if it had detected a peer failure as described 5037 in Section 9.7.1. 5039 10. Enrollment and Bootstrap 5041 The section defines the format of the configuration data as well the 5042 process to join a new overlay. 5044 10.1. Overlay Configuration 5046 This specification defines a new content type "application/ 5047 p2p-overlay+xml" for an MIME entity that contains overlay 5048 information. An example document is shown below. 5050 5051 5054 5056 CHORD-RELOAD 5057 16 5058 5059 MIIDJDCCAo2gAwIBAgIBADANBgkqhkiG9w0BAQUFADBwMQswCQYDVQQGEwJVUzET 5060 MBEGA1UECBMKQ2FsaWZvcm5pYTERMA8GA1UEBxMIU2FuIEpvc2UxDjAMBgNVBAoT 5061 BXNpcGl0MSkwJwYDVQQLEyBTaXBpdCBUZXN0IENlcnRpZmljYXRlIEF1dGhvcml0 5062 eTAeFw0wMzA3MTgxMjIxNTJaFw0xMzA3MTUxMjIxNTJaMHAxCzAJBgNVBAYTAlVT 5063 MRMwEQYDVQQIEwpDYWxpZm9ybmlhMREwDwYDVQQHEwhTYW4gSm9zZTEOMAwGA1UE 5064 ChMFc2lwaXQxKTAnBgNVBAsTIFNpcGl0IFRlc3QgQ2VydGlmaWNhdGUgQXV0aG9y 5065 aXR5MIGfMA0GCSqGSIb3DQEBAQUAA4GNADCBiQKBgQDDIh6DkcUDLDyK9BEUxkud 5066 +nJ4xrCVGKfgjHm6XaSuHiEtnfELHM+9WymzkBNzZpJu30yzsxwfKoIKugdNUrD4 5067 N3viCicwcN35LgP/KnbN34cavXHr4ZlqxH+OdKB3hQTpQa38A7YXdaoz6goW2ft5 5068 Mi74z03GNKP/G9BoKOGd5QIDAQABo4HNMIHKMB0GA1UdDgQWBBRrRhcU6pR2JYBU 5069 bhNU2qHjVBShtjCBmgYDVR0jBIGSMIGPgBRrRhcU6pR2JYBUbhNU2qHjVBShtqF0 5070 pHIwcDELMAkGA1UEBhMCVVMxEzARBgNVBAgTCkNhbGlmb3JuaWExETAPBgNVBAcT 5071 CFNhbiBKb3NlMQ4wDAYDVQQKEwVzaXBpdDEpMCcGA1UECxMgU2lwaXQgVGVzdCBD 5072 ZXJ0aWZpY2F0ZSBBdXRob3JpdHmCAQAwDAYDVR0TBAUwAwEB/zANBgkqhkiG9w0B 5073 AQUFAAOBgQCWbRvv1ZGTRXxbH8/EqkdSCzSoUPrs+rQqR0xdQac9wNY/nlZbkR3O 5074 qAezG6Sfmklvf+DOg5RxQq/+Y6I03LRepc7KeVDpaplMFGnpfKsibETMipwzayNQ 5075 QgUf4cKBiF+65Ue7hZuDJa2EMv8qW4twEhGDYclpFU9YozyS1OhvUg== 5076 5077 YmFkIGNlcnQK 5078 https://example.org 5079 https://example.net 5080 false 5082 5083 5084 5085 20 5086 5087 5088 false 5089 false 5090 5091 400 5092 30 5093 true 5094 password 5095 4000 5096 30 5097 TLS 5098 47112162e84c69ba 5099 6eba45d31a900c06 5100 6ebc45d31a900c06 5101 6ebc45d31a900ca6 5103 foo 5105 5106 5107 5108 SINGLE 5109 USER-MATCH 5110 1 5111 100 5112 5113 5114 VGhpcyBpcyBub3QgcmlnaHQhCg== 5115 5116 5117 5118 5119 ARRAY 5120 NODE-MULTIPLE 5121 3 5122 22 5123 4 5124 1 5125 5126 5127 5128 VGhpcyBpcyBub3QgcmlnaHQhCg== 5129 5130 5131 5132 5133 VGhpcyBpcyBub3QgcmlnaHQhCg== 5134 5135 5136 VGhpcyBpcyBub3QgcmlnaHQhCg== 5138 5140 The file MUST be a well formed XML document and it SHOULD contain an 5141 encoding declaration in the XML declaration. If the charset 5142 parameter of the MIME content type declaration is present and it is 5143 different from the encoding declaration, the charset parameter takes 5144 precedence. Every application conforming to this specification MUST 5145 accept the UTF-8 character encoding to ensure minimal 5146 interoperability. The namespace for the elements defined in this 5147 specification is urn:ietf:params:xml:ns:p2p:config-base and 5148 urn:ietf:params:xml:ns:p2p:config-chord". 5150 The file can contain multiple "configuration" elements where each one 5151 contains the configuration information for a different overlay. Each 5152 "configuration" has the following attributes: 5154 instance-name: name of the overlay 5155 expiration: time in the future at which this overlay configuration 5156 is no longer valid. The peer SHOULD retrieve a new copy of the 5157 configuration at a randomly selected time that is before the 5158 expiration time. 5159 sequence: a monotonically increasing sequence number between 1 and 5160 2^16-2 5162 Inside each overlay element, the following elements can occur: 5164 topology-plugin This element defines the overlay algorithm being 5165 used. If missing the default is "CHORD-RELOAD". 5166 node-id-length This element contains the length of a NodeId 5167 (NodeIdLength) in bytes. This value MUST be between 16 (128 bits) 5168 and 20 (160 bits). If this element is not present, the default of 5169 16 is used. 5170 root-cert This element contains a base-64 encoded X.509v3 5171 certificate that is a root trust anchor used to sign all 5172 certificates in this overlay. There can be more than one root- 5173 cert element. 5174 enrollment-server This element contains the URL at which the 5175 enrollment server can be reached in a "url" element. This URL 5176 MUST be of type "https:". More than one enrollment-server element 5177 may be present. 5179 self-signed-permitted This element indicates whether self-signed 5180 certificates are permitted. If it is set to "true", then self- 5181 signed certificates are allowed, in which case the enrollment- 5182 server and root-cert elements may be absent. Otherwise, it SHOULD 5183 be absent, but MAY be set to "false". This element also contains 5184 an attribute "digest" which indicates the digest to be used to 5185 compute the Node-ID. Valid values for this parameter are "sha1" 5186 and "sha256" representing SHA-1 [RFC3174] and SHA-256 [RFC4634] 5187 respectively. Implementations MUST support both of these 5188 algorithms. 5189 bootstrap-node This element represents the address of one of the 5190 bootstrap nodes. It has an attribute called "address" that 5191 represents the IP address (either IPv4 or IPv6, since they can be 5192 distinguished) and an optional attribute called "port" that 5193 represents the port and defaults to 6084. The IP address is in 5194 typical hexadecimal form using standard period and colon 5195 separators as specified in [RFC5952]. More than one bootstrap- 5196 peer element may be present. 5197 turn-density This element is a positive integer that represents the 5198 approximate reciprocal of density of nodes that can act as TURN 5199 servers. For example, if 5% of the nodes can act as TURN servers, 5200 this would be set to 20. If it is not present, the default value 5201 is 1. If there are no TURN servers in the overlay, it is set to 5202 zero. 5203 multicast-bootstrap This element represents the address of a 5204 multicast, broadcast, or anycast address and port that may be used 5205 for bootstrap. Nodes SHOULD listen on the address. It has an 5206 attributed called "address" that represents the IP address and an 5207 optional attribute called "port" that represents the port and 5208 defaults to 6084. More than one "multicast-bootstrap" element may 5209 be present. 5210 clients-permitted This element represents whether clients are 5211 permitted or whether all nodes must be peers. If it is set to 5212 "true" or absent, this indicates that clients are permitted. If 5213 it is set to "false" then nodes are not allowed to remain clients 5214 after the initial join. There is currently no way for the overlay 5215 to enforce this. 5216 no-ice This element represents whether nodes are required to use 5217 the "No-ICE" Overlay Link protocols in this overlay. If it is 5218 absent, it is treated as if it were set to "false". 5219 chord-update-interval The update frequency for the Chord-reload 5220 topology plugin (see Section 9). 5221 chord-ping-interval The ping frequency for the Chord-reload 5222 topology plugin (see Section 9). 5224 chord-reactive Whether reactive recovery should be used for this 5225 overlay. Set to "true" or "false". Default if missing is "true". 5226 (see Section 9). 5227 shared-secret If shared secret mode is used, this contains the 5228 shared secret. 5229 max-message-size Maximum size in bytes of any message in the 5230 overlay. If this value is not present, the default is 5000. 5231 initial-ttl Initial default TTL (time to live, see Section 5.3.2) 5232 for messages. If this value is not present, the default is 100. 5233 overlay-link-protocol Indicates a permissible overlay link protocol 5234 (see Section 5.6.1 for requirements for such protocols). An 5235 arbitrary number of these elements may appear. If none appear, 5236 then this implies the default value, "TLS", which refers to the 5237 use of TLS and DTLS. If one or more elements appear, then no 5238 default value applies. 5239 kind-signer This contains a single Node-ID in hexadecimal and 5240 indicates that the certificate with this Node-ID is allowed to 5241 sign kinds. Identifying kind-signer by Node-ID instead of 5242 certificate allows the use of short lived certificates without 5243 constantly having to provide an updated configuration file. 5244 bad-node This contains a single Node-ID in hexadecimal and 5245 indicates that the certificate with this Node-ID MUST NOT be 5246 considered valid. This allows certificate revocation. An 5247 arbitrary number of these elements can be provided. Note that 5248 because certificates may expire, bad-node entries need only be 5249 present for the lifetime of the certificate. Technically 5250 speaking, bad node-ids may be reused once their certificates have 5251 expired, the requirement for node-ids to be pseudo randomly 5252 generated gives this event a vanishing probability. 5254 Inside each overlay element, the required-kinds elements can also 5255 occur. This element indicates the kinds that members must support 5256 and contains multiple kind-block elements that each define a single 5257 kind that MUST be supported by nodes in the overlay. Each kind-block 5258 consists of a single kind element and a kind-signature. The kind 5259 element defines the kind. The kind-signature is the signature 5260 computed over the kind element. 5262 Each kind has either an id attribute or a name attribute. The name 5263 attribute is a string representing the kind (the name registered to 5264 IANA) while the id is an integer kind-id allocated out of private 5265 space. 5267 In addition, the kind element contains the following elements: 5269 max-count: the maximum number of values which members of the overlay 5270 must support. 5271 data-model: the data model to be used. 5272 max-size: the maximum size of individual values. 5273 access-control: the access control model to be used. 5274 max-node-multiple: This is optional and only used when the access 5275 control is NODE-MULTIPLE. This indicates the maximum value for 5276 the i counter. This is an integer greater than 0. 5278 All of the non optional values MUST be provided. If the kind is 5279 registered with IANA, the data-model and access-control elements MUST 5280 match those in the kind registration, and clients MUST ignore them in 5281 favor of the IANA versions. Multiple required-kinds elements MAY be 5282 present. 5284 The kind-block element also MUST contain a "kind-signature" element. 5285 This signature is computed across the kind from the beginning of the 5286 first < of the kind to the end of the last > of the kind in the same 5287 way as the signature element described later in this section. 5289 The configuration file is a binary file and cannot be changed - 5290 including whitespace changes - or the signature will break. The 5291 signature is computed by taking each configuration element and 5292 starting from, and including, the first < at the start of 5293 up to and including the > in and 5294 treating this as a binary blob that is signed using the standard 5295 SecurityBlock defined in Section 5.3.4. The SecurityBlock is base 64 5296 encoded using the base64 alphabet from RFC[RFC4648] and put in the 5297 signature element following the configuration object in the 5298 configuration file. 5300 When a node receives a new configuration file, it MUST change its 5301 configuration to meet the new requirements. This may require the 5302 node to exit the DHT and re-join. If a node is not capable of 5303 supporting the new requirements, it MUST exit the overlay. If some 5304 information about a particular kind changes from what the node 5305 previously knew about the kind (for example the max size), the new 5306 information in the configuration files overrides any previously 5307 learned information. If any kind data was signed by a node that is 5308 no longer allowed to sign kinds, that kind MUST be discarded along 5309 with any stored information of that kind. Note that forcing an 5310 avalanche restart of the overlay with a configuration change that 5311 requires re-joining the overlay may result in serious performance 5312 problems, including total collapse of the network if configuration 5313 parameters are not properly considered. Such an event may be 5314 necessary in case of a compromised CA or similar problem, but for 5315 large overlays should be avoided in almost all circumstances. 5317 10.1.1. Relax NG Grammar 5319 The grammar for the configuration data is: 5321 namespace chord = "urn:ietf:params:xml:ns:p2p:config-chord" 5322 namespace local = "" 5323 default namespace p2pcf = "urn:ietf:params:xml:ns:p2p:config-base" 5324 namespace rng = "http://relaxng.org/ns/structure/1.0" 5326 anything = 5327 (element * { anything } 5328 | attribute * { text } 5329 | text)* 5331 foreign-elements = element * - (p2pcf:* | local:* | chord:*) 5332 { anything }* 5333 foreign-attributes = attribute * - (p2pcf:*|local:*|chord:*) 5334 { text }* 5335 foreign-nodes = (foreign-attributes | foreign-elements)* 5337 start = element p2pcf:overlay { 5338 overlay-element 5339 } 5341 overlay-element &= element configuration { 5342 attribute instance-name { xsd:string }, 5343 attribute expiration { xsd:dateTime }?, 5344 attribute sequence { xsd:long }?, 5345 foreign-attributes*, 5346 parameter 5347 }+ 5348 overlay-element &= element signature { 5349 attribute algorithm { signature-algorithm-type }?, 5350 xsd:base64Binary 5351 }* 5353 signature-algorithm-type |= "rsa-sha1" 5354 signature-algorithm-type |= xsd:string # signature alg extensions 5356 parameter &= element topology-plugin { topology-plugin-type }? 5357 topology-plugin-type |= xsd:string # topo plugin extensions 5358 parameter &= element max-message-size { xsd:unsignedInt }? 5359 parameter &= element initial-ttl { xsd:int }? 5360 parameter &= element root-cert { xsd:base64Binary }* 5361 parameter &= element required-kinds { kind-block* }? 5362 parameter &= element enrollment-server { xsd:anyURI }* 5363 parameter &= element kind-signer { xsd:string }* 5364 parameter &= element bad-node { xsd:string }* 5365 parameter &= element no-ice { xsd:boolean }? 5366 parameter &= element shared-secret { xsd:string }? 5367 parameter &= element overlay-link-protocol { xsd:string }* 5368 parameter &= element clients-permitted { xsd:boolean }? 5369 parameter &= element turn-density { xsd:int }? 5370 parameter &= element node-id-length { xsd:int }? 5371 parameter &= foreign-elements* 5373 parameter &= 5374 element self-signed-permitted { 5375 attribute digest { self-signed-digest-type }, 5376 xsd:boolean 5377 }? 5378 self-signed-digest-type |= "sha1" 5379 self-signed-digest-type |= xsd:string # signature digest extensions 5381 parameter &= element bootstrap-node { 5382 attribute address { xsd:string }, 5383 attribute port { xsd:int }? 5384 }* 5386 parameter &= element multicast-bootstrap { 5387 attribute address { xsd:string }, 5388 attribute port { xsd:int }? 5389 }* 5391 kind-block = element kind-block { 5392 element kind { 5393 ( attribute name { kind-names } 5394 | attribute id { xsd:int } ), 5395 kind-parameter 5396 } & 5397 element kind-signature { 5398 attribute algorithm { signature-algorithm-type }?, 5399 xsd:base64Binary 5400 }? 5401 } 5403 kind-parameter &= element max-count { xsd:int } 5404 kind-parameter &= element max-size { xsd:int } 5405 kind-parameter &= element max-node-multiple { xsd:int }? 5407 kind-parameter &= element data-model { data-model-type } 5408 data-model-type |= "SINGLE" 5409 data-model-type |= "ARRAY" 5410 data-model-type |= "DICTIONARY" 5411 data-model-type |= xsd:string # data model extensions 5412 kind-parameter &= element access-control { access-control-type } 5413 access-control-type |= "USER-MATCH" 5414 access-control-type |= "NODE-MATCH" 5415 access-control-type |= "USER-NODE-MATCH" 5416 access-control-type |= "NODE-MULTIPLE" 5417 access-control-type |= xsd:string # access control extensions 5419 kind-parameter &= foreign-elements* 5421 kind-names |= "TURN-SERVICE" 5422 kind-names |= "CERTIFICATE_BY_NODE" 5423 kind-names |= "CERTIFICATE_BY_USER" 5424 kind-names |= xsd:string # kind extensions 5426 # Chord specific parameters 5427 topology-plugin-type |= "CHORD-RELOAD" 5428 parameter &= element chord:chord-ping-interval { xsd:int }? 5429 parameter &= element chord:chord-update-interval { xsd:int }? 5430 parameter &= element chord:chord-reactive { xsd:boolean }? 5432 10.2. Discovery Through Configuration Server 5434 When a node first enrolls in a new overlay, it starts with a 5435 discovery process to find a configuration server. 5437 The node first determines the overlay name. This value is provided 5438 by the user or some other out of band provisioning mechanism. The 5439 out of band mechanisms may also provide an optional URL for the 5440 configuration server. If a URL for the configuration server is not 5441 provided, the node MUST do a DNS SRV query using a Service name of 5442 "p2psip-enroll" and a protocol of TCP to find a configuration server 5443 and form the URL by appending a path of "/.well-known/p2psip-enroll" 5444 to the overlay name. This uses the "well known URI" framework 5445 defined in [RFC5785]. For example, if the overlay name was 5446 example.com, the URL would be 5447 "https://example.com//.well-known/p2psip-enroll". 5449 Once an address and URL for the configuration server is determined, 5450 the peer forms an HTTPS connection to that IP address. The 5451 certificate MUST match the overlay name as described in [RFC2818]. 5452 Then the node MUST fetch a new copy of the configuration file. To do 5453 this, the peer performs a GET to the URL. The result of the HTTP GET 5454 is an XML configuration file described above, which replaces any 5455 previously learned configuration file for this overlay. 5457 For overlays that do not use a configuration server, nodes obtain the 5458 configuration information needed to join the overlay through some out 5459 of band approach such an XML configuration file sent over email. 5461 10.3. Credentials 5463 If the configuration document contains a enrollment-server element, 5464 credentials are required to join the Overlay Instance. A peer which 5465 does not yet have credentials MUST contact the enrollment server to 5466 acquire them. 5468 RELOAD defines its own trivial certificate request protocol. We 5469 would have liked to have used an existing protocol but were concerned 5470 about the implementation burden of even the simplest of those 5471 protocols, such as [RFC5272] and [RFC5273]. Our objective was to 5472 have a protocol which could be easily implemented in a Web server 5473 which the operator did not control (e.g., in a hosted service) and 5474 was compatible with the existing certificate handling tooling as used 5475 with the Web certificate infrastructure. This means accepting bare 5476 PKCS#10 requests and returning a single bare X.509 certificate. 5477 Although the MIME types for these objects are defined, none of the 5478 existing protocols support exactly this model. 5480 The certificate request protocol is performed over HTTPS. The 5481 request is an HTTP POST with the following properties: 5483 o If authentication is required, there is a URL parameter of 5484 "password" and "username" containing the user's name and password 5485 in the clear (hence the need for HTTPS) 5486 o The body is of content type "application/pkcs10", as defined in 5487 [RFC2311]. 5488 o The Accept header contains the type "application/pkix-cert", 5489 indicating the type that is expected in the response. 5491 The enrollment server MUST authenticate the request using the 5492 provided user name and password. If the authentication succeeds and 5493 the requested user name is acceptable, the server generates and 5494 returns a certificate. The SubjectAltName field in the certificate 5495 contains the following values: 5497 o One or more Node-IDs which MUST be cryptographically random 5498 [RFC4086]. Each MUST be chosen by the enrollment server in such a 5499 way that they are unpredictable to the requesting user. E.g., the 5500 user MUST NOT be informed of potential (random) Node-IDs prior to 5501 authenticating. Each is placed in the subjectAltName using the 5502 uniformResourceIdentifier type and MUST contain RELOAD URIs as 5503 described in Section 13.15 and MUST contain a Destination list 5504 with a single entry of type "node_id". 5506 o A single name this user is allowed to use in the overlay, using 5507 type rfc822Name. 5509 The certificate is returned as type "application/pkix-cert" as 5510 defined in [RFC2585], with an HTTP status code of 200 OK. 5511 Certificate processing errors should be treated as HTTP errors and 5512 have appropriate HTTP status codes. 5514 The client MUST check that the certificate returned was signed by one 5515 of the certificates received in the "root-cert" list of the overlay 5516 configuration data. The node then reads the certificate to find the 5517 Node-IDs it can use. 5519 10.3.1. Self-Generated Credentials 5521 If the "self-signed-permitted" element is present in the 5522 configuration and set to "true", then a node MUST generate its own 5523 self-signed certificate to join the overlay. The self-signed 5524 certificate MAY contain any user name of the users choice. 5526 The Node-ID MUST be computed by applying the digest specified in the 5527 self-signed-permitted element to the DER representation of the user's 5528 public key (more specifically the subjectPublicKeyInfo) and taking 5529 the high order bits. When accepting a self-signed certificate, nodes 5530 MUST check that the Node-ID and public keys match. This prevents 5531 Node-ID theft. 5533 Once the node has constructed a self-signed certificate, it MAY join 5534 the overlay. Before storing its certificate in the overlay 5535 (Section 7) it SHOULD look to see if the user name is already taken 5536 and if so choose another user name. Note that this only provides 5537 protection against accidental name collisions. Name theft is still 5538 possible. If protection against name theft is desired, then the 5539 enrollment service must be used. 5541 10.4. Searching for a Bootstrap Node 5543 If no cached bootstrap nodes are available and the configuration file 5544 has an multicast-bootstrap element, then the node SHOULD send a Ping 5545 request over UDP to the address and port found to each multicast- 5546 bootstrap element found in the configuration document. This MAY be a 5547 multicast, broadcast, or anycast address. The Ping should use the 5548 wildcard Node-ID as the destination Node-ID. 5550 The responder node that receives the Ping request SHOULD check that 5551 the overlay name is correct and that the requester peer sending the 5552 request has appropriate credentials for the overlay before responding 5553 to the Ping request even if the response is only an error. 5555 10.5. Contacting a Bootstrap Node 5557 In order to join the overlay, the joining node MUST contact a node in 5558 the overlay. Typically this means contacting the bootstrap nodes, 5559 since they are reachable by the local peer or have public IP 5560 addresses. If the joining node has cached a list of peers it has 5561 previously been connected with in this overlay, as an optimization it 5562 MAY attempt to use one or more of them as bootstrap nodes before 5563 falling back to the bootstrap nodes listed in the configuration file. 5565 When contacting a bootstrap node, the joining node first forms the 5566 DTLS or TLS connection to the bootstrap node and then sends an Attach 5567 request over this connection with the destination Node-ID set to the 5568 joining node's Node-ID. 5570 When the requester node finally does receive a response from some 5571 responding node, it can note the Node-ID in the response and use this 5572 Node-ID to start sending requests to join the Overlay Instance as 5573 described in Section 5.4. 5575 After a node has successfully joined the overlay network, it will 5576 have direct connections to several peers. Some MAY be added to the 5577 cached bootstrap nodes list and used in future boots. Peers that are 5578 not directly connected MUST NOT be cached. The suggested number of 5579 peers to cache is 10. Algorithms for determining which peers to 5580 cache are beyond the scope of this specification. 5582 11. Message Flow Example 5584 The following abbreviation are used in the message flow diagrams: JP 5585 = joining peer, AP = admitting peer, NP = next peer after the AP, NNP 5586 = next next peer which is the peer after NP, PP = previous peer 5587 before the AP, PPP = previous previous peer which is the peer before 5588 the PP, BP = bootstrap peer. 5590 In the following example, we assume that JP has formed a connection 5591 to one of the bootstrap nodes. JP then sends an Attach through that 5592 peer to a resource ID of itself (JP). It gets routed to the 5593 admitting peer (AP) because JP is not yet part of the overlay. When 5594 AP responds, JP and AP use ICE to set up a connection and then set up 5595 TLS. Once AP has connected to JP, AP sends to JP an Update to 5596 populate its Routing Table. The following example shows the Update 5597 happening after the TLS connection is formed but it could also happen 5598 before in which case the Update would often be routed through other 5599 nodes. 5601 JP PPP PP AP NP NNP BP 5602 | | | | | | | 5603 | | | | | | | 5604 | | | | | | | 5605 |Attach Dest=JP | | | | | 5606 |---------------------------------------------------------->| 5607 | | | | | | | 5608 | | | | | | | 5609 | | |Attach Dest=JP | | | 5610 | | |<--------------------------------------| 5611 | | | | | | | 5612 | | | | | | | 5613 | | |Attach Dest=JP | | | 5614 | | |-------->| | | | 5615 | | | | | | | 5616 | | | | | | | 5617 | | |AttachAns | | | 5618 | | |<--------| | | | 5619 | | | | | | | 5620 | | | | | | | 5621 | | |AttachAns | | | 5622 | | |-------------------------------------->| 5623 | | | | | | | 5624 | | | | | | | 5625 |AttachAns | | | | | 5626 |<----------------------------------------------------------| 5627 | | | | | | | 5628 | | | | | | | 5629 |TLS | | | | | | 5630 |.............................| | | | 5631 | | | | | | | 5632 | | | | | | | 5633 | | | | | | | 5634 |Update | | | | | | 5635 |<----------------------------| | | | 5636 | | | | | | | 5637 | | | | | | | 5638 |UpdateAns| | | | | | 5639 |---------------------------->| | | | 5640 | | | | | | | 5641 | | | | | | | 5642 | | | | | | | 5644 The JP then forms connections to the appropriate neighbors, such as 5645 NP, by sending an Attach which gets routed via other nodes. When NP 5646 responds, JP and NP use ICE and TLS to set up a connection. 5648 JP PPP PP AP NP NNP BP 5649 | | | | | | | 5650 | | | | | | | 5651 | | | | | | | 5652 |Attach NP | | | | | 5653 |---------------------------->| | | | 5654 | | | | | | | 5655 | | | | | | | 5656 | | | |Attach NP| | | 5657 | | | |-------->| | | 5658 | | | | | | | 5659 | | | | | | | 5660 | | | |AttachAns| | | 5661 | | | |<--------| | | 5662 | | | | | | | 5663 | | | | | | | 5664 |AttachAns | | | | | 5665 |<----------------------------| | | | 5666 | | | | | | | 5667 | | | | | | | 5668 |Attach | | | | | | 5669 |-------------------------------------->| | | 5670 | | | | | | | 5671 | | | | | | | 5672 |TLS | | | | | | 5673 |.......................................| | | 5674 | | | | | | | 5675 | | | | | | | 5676 | | | | | | | 5677 | | | | | | | 5679 JP also needs to populate its finger table (for Chord). It issues an 5680 Attach to a variety of locations around the overlay. The diagram 5681 below shows it sending an Attach halfway around the Chord ring to the 5682 JP + 2^127. 5684 JP NP XX TP 5685 | | | | 5686 | | | | 5687 | | | | 5688 |Attach JP+2<<126 | | 5689 |-------->| | | 5690 | | | | 5691 | | | | 5692 | |Attach JP+2<<126 | 5693 | |-------->| | 5694 | | | | 5695 | | | | 5696 | | |Attach JP+2<<126 5697 | | |-------->| 5698 | | | | 5699 | | | | 5700 | | |AttachAns| 5701 | | |<--------| 5702 | | | | 5703 | | | | 5704 | |AttachAns| | 5705 | |<--------| | 5706 | | | | 5707 | | | | 5708 |AttachAns| | | 5709 |<--------| | | 5710 | | | | 5711 | | | | 5712 |TLS | | | 5713 |.............................| 5714 | | | | 5715 | | | | 5716 | | | | 5717 | | | | 5719 Once JP has a reasonable set of connections, it is ready to take its 5720 place in the DHT. It does this by sending a Join to AP. AP does a 5721 series of Store requests to JP to store the data that JP will be 5722 responsible for. AP then sends JP an Update explicitly labeling JP 5723 as its predecessor. At this point, JP is part of the ring and 5724 responsible for a section of the overlay. AP can now forget any data 5725 which is assigned to JP and not AP. 5727 JP PPP PP AP NP NNP BP 5728 | | | | | | | 5729 | | | | | | | 5730 | | | | | | | 5731 |JoinReq | | | | | | 5732 |---------------------------->| | | | 5733 | | | | | | | 5734 | | | | | | | 5735 |JoinAns | | | | | | 5736 |<----------------------------| | | | 5737 | | | | | | | 5738 | | | | | | | 5739 |StoreReq Data A | | | | | 5740 |<----------------------------| | | | 5741 | | | | | | | 5742 | | | | | | | 5743 |StoreAns | | | | | | 5744 |---------------------------->| | | | 5745 | | | | | | | 5746 | | | | | | | 5747 |StoreReq Data B | | | | | 5748 |<----------------------------| | | | 5749 | | | | | | | 5750 | | | | | | | 5751 |StoreAns | | | | | | 5752 |---------------------------->| | | | 5753 | | | | | | | 5754 | | | | | | | 5755 |UpdateReq| | | | | | 5756 |<----------------------------| | | | 5757 | | | | | | | 5758 | | | | | | | 5759 |UpdateAns| | | | | | 5760 |---------------------------->| | | | 5761 | | | | | | | 5762 | | | | | | | 5763 | | | | | | | 5764 | | | | | | | 5766 In Chord, JP's neighbor table needs to contain its own predecessors. 5767 It couldn't connect to them previously because it did not yet know 5768 their addresses. However, now that it has received an Update from 5769 AP, it has AP's predecessors, which are also its own, so it sends 5770 Attaches to them. Below it is shown connecting to AP's closest 5771 predecessor, PP. 5773 JP PPP PP AP NP NNP BP 5774 | | | | | | | 5775 | | | | | | | 5776 | | | | | | | 5777 |Attach Dest=PP | | | | | 5778 |---------------------------->| | | | 5779 | | | | | | | 5780 | | | | | | | 5781 | | |Attach Dest=PP | | | 5782 | | |<--------| | | | 5783 | | | | | | | 5784 | | | | | | | 5785 | | |AttachAns| | | | 5786 | | |-------->| | | | 5787 | | | | | | | 5788 | | | | | | | 5789 |AttachAns| | | | | | 5790 |<----------------------------| | | | 5791 | | | | | | | 5792 | | | | | | | 5793 |TLS | | | | | | 5794 |...................| | | | | 5795 | | | | | | | 5796 | | | | | | | 5797 |UpdateReq| | | | | | 5798 |------------------>| | | | | 5799 | | | | | | | 5800 | | | | | | | 5801 |UpdateAns| | | | | | 5802 |<------------------| | | | | 5803 | | | | | | | 5804 | | | | | | | 5805 |UpdateReq| | | | | | 5806 |---------------------------->| | | | 5807 | | | | | | | 5808 | | | | | | | 5809 |UpdateAns| | | | | | 5810 |<----------------------------| | | | 5811 | | | | | | | 5812 | | | | | | | 5813 |UpdateReq| | | | | | 5814 |-------------------------------------->| | | 5815 | | | | | | | 5816 | | | | | | | 5817 |UpdateAns| | | | | | 5818 |<--------------------------------------| | | 5819 | | | | | | | 5820 | | | | | | | 5822 Finally, now that JP has a copy of all the data and is ready to route 5823 messages and receive requests, it sends Updates to everyone in its 5824 Routing Table to tell them it is ready to go. Below, it is shown 5825 sending such an update to TP. 5827 JP NP XX TP 5828 | | | | 5829 | | | | 5830 | | | | 5831 |Update | | | 5832 |---------------------------->| 5833 | | | | 5834 | | | | 5835 |UpdateAns| | | 5836 |<----------------------------| 5837 | | | | 5838 | | | | 5839 | | | | 5840 | | | | 5842 12. Security Considerations 5844 12.1. Overview 5846 RELOAD provides a generic storage service, albeit one designed to be 5847 useful for P2PSIP. In this section we discuss security issues that 5848 are likely to be relevant to any usage of RELOAD. More background 5849 information can be found in [RFC5765]. 5851 In any Overlay Instance, any given user depends on a number of peers 5852 with which they have no well-defined relationship except that they 5853 are fellow members of the Overlay Instance. In practice, these other 5854 nodes may be friendly, lazy, curious, or outright malicious. No 5855 security system can provide complete protection in an environment 5856 where most nodes are malicious. The goal of security in RELOAD is to 5857 provide strong security guarantees of some properties even in the 5858 face of a large number of malicious nodes and to allow the overlay to 5859 function correctly in the face of a modest number of malicious nodes. 5861 P2PSIP deployments require the ability to authenticate both peers and 5862 resources (users) without the active presence of a trusted entity in 5863 the system. We describe two mechanisms. The first mechanism is 5864 based on public key certificates and is suitable for general 5865 deployments. The second is an admission control mechanism based on 5866 an overlay-wide shared symmetric key. 5868 12.2. Attacks on P2P Overlays 5870 The two basic functions provided by overlay nodes are storage and 5871 routing: some node is responsible for storing a peer's data and for 5872 allowing a third peer to fetch this stored data. Other nodes are 5873 responsible for routing messages to and from the storing nodes. Each 5874 of these issues is covered in the following sections. 5876 P2P overlays are subject to attacks by subversive nodes that may 5877 attempt to disrupt routing, corrupt or remove user registrations, or 5878 eavesdrop on signaling. The certificate-based security algorithms we 5879 describe in this specification are intended to protect overlay 5880 routing and user registration information in RELOAD messages. 5882 To protect the signaling from attackers pretending to be valid peers 5883 (or peers other than themselves), the first requirement is to ensure 5884 that all messages are received from authorized members of the 5885 overlay. For this reason, RELOAD transports all messages over a 5886 secure channel (TLS and DTLS are defined in this document) which 5887 provides message integrity and authentication of the directly 5888 communicating peer. In addition, messages and data are digitally 5889 signed with the sender's private key, providing end-to-end security 5890 for communications. 5892 12.3. Certificate-based Security 5894 This specification stores users' registrations and possibly other 5895 data in an overlay network. This requires a solution to securing 5896 this data as well as securing, as well as possible, the routing in 5897 the overlay. Both types of security are based on requiring that 5898 every entity in the system (whether user or peer) authenticate 5899 cryptographically using an asymmetric key pair tied to a certificate. 5901 When a user enrolls in the Overlay Instance, they request or are 5902 assigned a unique name, such as "alice@dht.example.net". These names 5903 are unique and are meant to be chosen and used by humans much like a 5904 SIP Address of Record (AOR) or an email address. The user is also 5905 assigned one or more Node-IDs by the central enrollment authority. 5906 Both the name and the Node-ID are placed in the certificate, along 5907 with the user's public key. 5909 Each certificate enables an entity to act in two sorts of roles: 5911 o As a user, storing data at specific Resource-IDs in the Overlay 5912 Instance corresponding to the user name. 5913 o As a overlay peer with the Node-ID(s) listed in the certificate. 5915 Note that since only users of this Overlay Instance need to validate 5916 a certificate, this usage does not require a global PKI. Instead, 5917 certificates are signed by a central enrollment authority which acts 5918 as the certificate authority for the Overlay Instance. This 5919 authority signs each peer's certificate. Because each peer possesses 5920 the CA's certificate (which they receive on enrollment) they can 5921 verify the certificates of the other entities in the overlay without 5922 further communication. Because the certificates contain the user/ 5923 peer's public key, communications from the user/peer can be verified 5924 in turn. 5926 If self-signed certificates are used, then the security provided is 5927 significantly decreased, since attackers can mount Sybil attacks. In 5928 addition, attackers cannot trust the user names in certificates 5929 (though they can trust the Node-IDs because they are 5930 cryptographically verifiable). This scheme may be appropriate for 5931 some small deployments, such as a small office or an ad hoc overlay 5932 set up among participants in a meeting where all hosts on the network 5933 are trusted. Some additional security can be provided by using the 5934 shared secret admission control scheme as well. 5936 Because all stored data is signed by the owner of the data the 5937 storing peer can verify that the storer is authorized to perform a 5938 store at that Resource-ID and also allow any consumer of the data to 5939 verify the provenance and integrity of the data when it retrieves it. 5941 Note that RELOAD does not itself provide a revocation/status 5942 mechanism (though certificates may of course include OCSP responder 5943 information). Thus, certificate lifetimes should be chosen to 5944 balance the compromise window versus the cost of certificate renewal. 5945 Because RELOAD is already designed to operate in the face of some 5946 fraction of malicious peers, this form of compromise is not fatal. 5948 All implementations MUST implement certificate-based security. 5950 12.4. Shared-Secret Security 5952 RELOAD also supports a shared secret admission control scheme that 5953 relies on a single key that is shared among all members of the 5954 overlay. It is appropriate for small groups that wish to form a 5955 private network without complexity. In shared secret mode, all the 5956 peers share a single symmetric key which is used to key TLS-PSK 5957 [RFC4279] or TLS-SRP [RFC5054] mode. A peer which does not know the 5958 key cannot form TLS connections with any other peer and therefore 5959 cannot join the overlay. 5961 One natural approach to a shared-secret scheme is to use a user- 5962 entered password as the key. The difficulty with this is that in 5963 TLS-PSK mode, such keys are very susceptible to dictionary attacks. 5965 If passwords are used as the source of shared-keys, then TLS-SRP is a 5966 superior choice because it is not subject to dictionary attacks. 5968 12.5. Storage Security 5970 When certificate-based security is used in RELOAD, any given 5971 Resource-ID/Kind-ID pair is bound to some small set of certificates. 5972 In order to write data, the writer must prove possession of the 5973 private key for one of those certificates. Moreover, all data is 5974 stored, signed with the same private key that was used to authorize 5975 the storage. This set of rules makes questions of authorization and 5976 data integrity - which have historically been thorny for overlays - 5977 relatively simple. 5979 12.5.1. Authorization 5981 When a client wants to store some value, it first digitally signs the 5982 value with its own private key. It then sends a Store request that 5983 contains both the value and the signature towards the storing peer 5984 (which is defined by the Resource Name construction algorithm for 5985 that particular kind of value). 5987 When the storing peer receives the request, it must determine whether 5988 the storing client is authorized to store at this Resource-ID/Kind-ID 5989 pair. Determining this requires comparing the user's identity to the 5990 requirements of the access control model (see Section 6.3). If it 5991 satisfies those requirements the user is authorized to write, pending 5992 quota checks as described in the next section. 5994 For example, consider the certificate with the following properties: 5996 User name: alice@dht.example.com 5997 Node-ID: 013456789abcdef 5998 Serial: 1234 6000 If Alice wishes to Store a value of the "SIP Location" kind, the 6001 Resource Name will be the SIP AOR "sip:alice@dht.example.com". The 6002 Resource-ID will be determined by hashing the Resource Name. Because 6003 SIP Location uses the USER-NODE-MATCH policy, it first verifies that 6004 the user name in the certificate hashes to the requested Resource-ID. 6005 It then verifies that the Node-Id in the certificate matches the 6006 dictionary key being used for the store. If both of these checks 6007 succeed, the Store is authorized. Note that because the access 6008 control model is different for different kinds, the exact set of 6009 checks will vary. 6011 12.5.2. Distributed Quota 6013 Being a peer in an Overlay Instance carries with it the 6014 responsibility to store data for a given region of the Overlay 6015 Instance. However, allowing clients to store unlimited amounts of 6016 data would create unacceptable burdens on peers and would also enable 6017 trivial denial of service attacks. RELOAD addresses this issue by 6018 requiring configurations to define maximum sizes for each kind of 6019 stored data. Attempts to store values exceeding this size MUST be 6020 rejected (if peers are inconsistent about this, then strange 6021 artifacts will happen when the zone of responsibility shifts and a 6022 different peer becomes responsible for overlarge data). Because each 6023 Resource-ID/Kind-ID pair is bound to a small set of certificates, 6024 these size restrictions also create a distributed quota mechanism, 6025 with the quotas administered by the central configuration server. 6027 Allowing different kinds of data to have different size restrictions 6028 allows new usages the flexibility to define limits that fit their 6029 needs without requiring all usages to have expansive limits. 6031 12.5.3. Correctness 6033 Because each stored value is signed, it is trivial for any retrieving 6034 peer to verify the integrity of the stored value. Some more care 6035 needs to be taken to prevent version rollback attacks. Rollback 6036 attacks on storage are prevented by the use of store times and 6037 lifetime values in each store. A lifetime represents the latest time 6038 at which the data is valid and thus limits (though does not 6039 completely prevent) the ability of the storing node to perform a 6040 rollback attack on retrievers. In order to prevent a rollback attack 6041 at the time of the Store request, we require that storage times be 6042 monotonically increasing. Storing peers MUST reject Store requests 6043 with storage times smaller than or equal to those they are currently 6044 storing. In addition, a fetching node which receives a data value 6045 with a storage time older than the result of the previous fetch knows 6046 a rollback has occurred. 6048 12.5.4. Residual Attacks 6050 The mechanisms described here provides a high degree of security, but 6051 some attacks remain possible. Most simply, it is possible for 6052 storing nodes to refuse to store a value (i.e., reject any request). 6053 In addition, a storing node can deny knowledge of values which it has 6054 previously accepted. To some extent these attacks can be ameliorated 6055 by attempting to store to/retrieve from replicas, but a retrieving 6056 client does not know whether it should try this or not, since there 6057 is a cost to doing so. 6059 The certificate-based authentication scheme prevents a single peer 6060 from being able to forge data owned by other peers. Furthermore, 6061 although a subversive peer can refuse to return data resources for 6062 which it is responsible, it cannot return forged data because it 6063 cannot provide authentication for such registrations. Therefore 6064 parallel searches for redundant registrations can mitigate most of 6065 the effects of a compromised peer. The ultimate reliability of such 6066 an overlay is a statistical question based on the replication factor 6067 and the percentage of compromised peers. 6069 In addition, when a kind is multivalued (e.g., an array data model), 6070 the storing node can return only some subset of the values, thus 6071 biasing its responses. This can be countered by using single values 6072 rather than sets, but that makes coordination between multiple 6073 storing agents much more difficult. This is a trade off that must be 6074 made when designing any usage. 6076 12.6. Routing Security 6078 Because the storage security system guarantees (within limits) the 6079 integrity of the stored data, routing security focuses on stopping 6080 the attacker from performing a DOS attack that misroutes requests in 6081 the overlay. There are a few obvious observations to make about 6082 this. First, it is easy to ensure that an attacker is at least a 6083 valid peer in the Overlay Instance. Second, this is a DOS attack 6084 only. Third, if a large percentage of the peers on the Overlay 6085 Instance are controlled by the attacker, it is probably impossible to 6086 perfectly secure against this. 6088 12.6.1. Background 6090 In general, attacks on DHT routing are mounted by the attacker 6091 arranging to route traffic through one or two nodes it controls. In 6092 the Eclipse attack [Eclipse] the attacker tampers with messages to 6093 and from nodes for which it is on-path with respect to a given victim 6094 node. This allows it to pretend to be all the nodes that are 6095 reachable through it. In the Sybil attack [Sybil], the attacker 6096 registers a large number of nodes and is therefore able to capture a 6097 large amount of the traffic through the DHT. 6099 Both the Eclipse and Sybil attacks require the attacker to be able to 6100 exercise control over her Node-IDs. The Sybil attack requires the 6101 creation of a large number of peers. The Eclipse attack requires 6102 that the attacker be able to impersonate specific peers. In both 6103 cases, these attacks are limited by the use of centralized, 6104 certificate-based admission control. 6106 12.6.2. Admissions Control 6108 Admission to a RELOAD Overlay Instance is controlled by requiring 6109 that each peer have a certificate containing its Node-Id. The 6110 requirement to have a certificate is enforced by using certificate- 6111 based mutual authentication on each connection. (Note: the 6112 following only applies when self-signed certificates are not used.) 6113 Whenever a peer connects to another peer, each side automatically 6114 checks that the other has a suitable certificate. These Node-Ids are 6115 randomly assigned by the central enrollment server. This has two 6116 benefits: 6118 o It allows the enrollment server to limit the number of Node-IDs 6119 issued to any individual user. 6120 o It prevents the attacker from choosing specific Node-Ids. 6122 The first property allows protection against Sybil attacks (provided 6123 the enrollment server uses strict rate limiting policies). The 6124 second property deters but does not completely prevent Eclipse 6125 attacks. Because an Eclipse attacker must impersonate peers on the 6126 other side of the attacker, he must have a certificate for suitable 6127 Node-Ids, which requires him to repeatedly query the enrollment 6128 server for new certificates, which will match only by chance. From 6129 the attacker's perspective, the difficulty is that if he only has a 6130 small number of certificates, the region of the Overlay Instance he 6131 is impersonating appears to be very sparsely populated by comparison 6132 to the victim's local region. 6134 12.6.3. Peer Identification and Authentication 6136 In general, whenever a peer engages in overlay activity that might 6137 affect the routing table it must establish its identity. This 6138 happens in two ways. First, whenever a peer establishes a direct 6139 connection to another peer it authenticates via certificate-based 6140 mutual authentication. All messages between peers are sent over this 6141 protected channel and therefore the peers can verify the data origin 6142 of the last hop peer for requests and responses without further 6143 cryptography. 6145 In some situations, however, it is desirable to be able to establish 6146 the identity of a peer with whom one is not directly connected. The 6147 most natural case is when a peer Updates its state. At this point, 6148 other peers may need to update their view of the overlay structure, 6149 but they need to verify that the Update message came from the actual 6150 peer rather than from an attacker. To prevent this, all overlay 6151 routing messages are signed by the peer that generated them. 6153 Replay is typically prevented for messages that impact the topology 6154 of the overlay by having the information come directly, or be 6155 verified by, the nodes that claimed to have generated the update. 6156 Data storage replay detection is done by signing time of the node 6157 that generated the signature on the store request thus providing a 6158 time based replay protection but the time synchronization is only 6159 needed between peers that can write to the same location. 6161 12.6.4. Protecting the Signaling 6163 The goal here is to stop an attacker from knowing who is signaling 6164 what to whom. An attacker is unlikely to be able to observe the 6165 activities of a specific individual given the randomization of IDs 6166 and routing based on the present peers discussed above. Furthermore, 6167 because messages can be routed using only the header information, the 6168 actual body of the RELOAD message can be encrypted during 6169 transmission. 6171 There are two lines of defense here. The first is the use of TLS or 6172 DTLS for each communications link between peers. This provides 6173 protection against attackers who are not members of the overlay. The 6174 second line of defense is to digitally sign each message. This 6175 prevents adversarial peers from modifying messages in flight, even if 6176 they are on the routing path. 6178 12.6.5. Residual Attacks 6180 The routing security mechanisms in RELOAD are designed to contain 6181 rather than eliminate attacks on routing. It is still possible for 6182 an attacker to mount a variety of attacks. In particular, if an 6183 attacker is able to take up a position on the overlay routing between 6184 A and B it can make it appear as if B does not exist or is 6185 disconnected. It can also advertise false network metrics in an 6186 attempt to reroute traffic. However, these are primarily DOS 6187 attacks. 6189 The certificate-based security scheme secures the namespace, but if 6190 an individual peer is compromised or if an attacker obtains a 6191 certificate from the CA, then a number of subversive peers can still 6192 appear in the overlay. While these peers cannot falsify responses to 6193 resource queries, they can respond with error messages, effecting a 6194 DoS attack on the resource registration. They can also subvert 6195 routing to other compromised peers. To defend against such attacks, 6196 a resource search must still consist of parallel searches for 6197 replicated registrations. 6199 13. IANA Considerations 6201 This section contains the new code points registered by this 6202 document. [NOTE TO IANA/RFC-EDITOR: Please replace RFC-AAAA with 6203 the RFC number for this specification in the following list.] 6205 13.1. Well-Known URI Registration 6207 IANA will make the following "Well Known URI" registration as 6208 described in [RFC5785]: 6210 [[Note to RFC Editor - this paragraph can be removed before 6211 publication. ]] A review request was sent to 6212 wellknown-uri-review@ietf.org on October 12, 2010. 6214 +----------------------------+----------------------+ 6215 | URI suffix: | p2psip-enroll | 6216 | Change controller: | IETF | 6217 | Specification document(s): | [RFC-AAAA] | 6218 | Related information: | None | 6219 +----------------------------+----------------------+ 6221 13.2. Port Registrations 6223 [[Note to RFC Editor - this paragraph can be removed before 6224 publication. ]] IANA has already allocated a TCP port for the main 6225 peer to peer protocol. This port has the name p2p-sip and the port 6226 number of 6084. IANA needs to update this registration to be defined 6227 for UDP as well as TCP. 6229 IANA will make the following port registration: 6231 +------------------------------+------------------------------------+ 6232 | Registration Technical | Cullen Jennings | 6233 | Contact | | 6234 | Registration Owner | IETF | 6235 | Transport Protocol | TCP & UDP | 6236 | Port Number | 6084 | 6237 | Service Name | p2psip-enroll | 6238 | Description | Peer to Peer Infrastructure | 6239 | | Enrollment | 6240 | Reference | [RFC-AAAA] | 6241 +------------------------------+------------------------------------+ 6243 13.3. Overlay Algorithm Types 6245 IANA SHALL create a "RELOAD Overlay Algorithm Type" Registry. 6246 Entries in this registry are strings denoting the names of overlay 6247 algorithms. The registration policy for this registry is RFC 5226 6248 IETF Review. The initial contents of this registry are: 6250 +----------------+----------+ 6251 | Algorithm Name | RFC | 6252 +----------------+----------+ 6253 | CHORD-RELOAD | RFC-AAAA | 6254 +----------------+----------+ 6256 13.4. Access Control Policies 6258 IANA SHALL create a "RELOAD Access Control Policy" Registry. Entries 6259 in this registry are strings denoting access control policies, as 6260 described in Section 6.3. New entries in this registry SHALL be 6261 registered via RFC 5226 Standards Action. The initial contents of 6262 this registry are: 6264 +-----------------+----------+ 6265 | Access Policy | RFC | 6266 +-----------------+----------+ 6267 | USER-MATCH | RFC-AAAA | 6268 | NODE-MATCH | RFC-AAAA | 6269 | USER-NODE-MATCH | RFC-AAAA | 6270 | NODE-MULTIPLE | RFC-AAAA | 6271 +-----------------+----------+ 6273 13.5. Application-ID 6275 IANA SHALL create a "RELOAD Application-ID" Registry. Entries in 6276 this registry are 16-bit integers denoting application kinds. Code 6277 points in the range 0x0001 to 0x7fff SHALL be registered via RFC 5226 6278 Standards Action. Code points in the range 0x8000 to 0xf000 SHALL be 6279 registered via RFC 5226 Expert Review. Code points in the range 6280 0xf001 to 0xfffe are reserved for private use. The initial contents 6281 of this registry are: 6283 +-------------+----------------+-------------------------------+ 6284 | Application | Application-ID | Specification | 6285 +-------------+----------------+-------------------------------+ 6286 | INVALID | 0 | RFC-AAAA | 6287 | SIP | 5060 | Reserved for use by SIP Usage | 6288 | SIP | 5061 | Reserved for use by SIP Usage | 6289 | Reserved | 0xffff | RFC-AAAA | 6290 +-------------+----------------+-------------------------------+ 6292 13.6. Data Kind-ID 6294 IANA SHALL create a "RELOAD Data Kind-ID" Registry. Entries in this 6295 registry are 32-bit integers denoting data kinds, as described in 6296 Section 4.2. Code points in the range 0x00000001 to 0x7fffffff SHALL 6297 be registered via RFC 5226 Standards Action. Code points in the 6298 range 0x8000000 to 0xf0000000 SHALL be registered via RFC 5226 Expert 6299 Review. Code points in the range 0xf0000001 to 0xfffffffe are 6300 reserved for private use via the kind description mechanism described 6301 in Section 10. The initial contents of this registry are: 6303 +---------------------+------------+----------+ 6304 | Kind | Kind-ID | RFC | 6305 +---------------------+------------+----------+ 6306 | INVALID | 0 | RFC-AAAA | 6307 | TURN_SERVICE | 2 | RFC-AAAA | 6308 | CERTIFICATE_BY_NODE | 3 | RFC-AAAA | 6309 | CERTIFICATE_BY_USER | 16 | RFC-AAAA | 6310 | Reserved | 0x7fffffff | RFC-AAAA | 6311 | Reserved | 0xfffffffe | RFC-AAAA | 6312 +---------------------+------------+----------+ 6314 13.7. Data Model 6316 IANA SHALL create a "RELOAD Data Model" Registry. Entries in this 6317 registry denoting data models, as described in Section 6.2. Code 6318 points in this registry SHALL be registered via RFC 5226 Standards 6319 Action. The initial contents of this registry are: 6321 +------------+----------+ 6322 | Data Model | RFC | 6323 +------------+----------+ 6324 | INVALID | RFC-AAAA | 6325 | SINGLE | RFC-AAAA | 6326 | ARRAY | RFC-AAAA | 6327 | DICTIONARY | RFC-AAAA | 6328 | RESERVED | RFC-AAAA | 6329 +------------+----------+ 6331 13.8. Message Codes 6333 IANA SHALL create a "RELOAD Message Code" Registry. Entries in this 6334 registry are 16-bit integers denoting method codes as described in 6335 Section 5.3.3. These codes SHALL be registered via RFC 5226 6336 Standards Action. The initial contents of this registry are: 6338 +---------------------------------+----------------+----------+ 6339 | Message Code Name | Code Value | RFC | 6340 +---------------------------------+----------------+----------+ 6341 | invalid | 0 | RFC-AAAA | 6342 | probe_req | 1 | RFC-AAAA | 6343 | probe_ans | 2 | RFC-AAAA | 6344 | attach_req | 3 | RFC-AAAA | 6345 | attach_ans | 4 | RFC-AAAA | 6346 | unused | 5 | | 6347 | unused | 6 | | 6348 | store_req | 7 | RFC-AAAA | 6349 | store_ans | 8 | RFC-AAAA | 6350 | fetch_req | 9 | RFC-AAAA | 6351 | fetch_ans | 10 | RFC-AAAA | 6352 | unused (was remove_req) | 11 | RFC-AAAA | 6353 | unused (was remove_ans) | 12 | RFC-AAAA | 6354 | find_req | 13 | RFC-AAAA | 6355 | find_ans | 14 | RFC-AAAA | 6356 | join_req | 15 | RFC-AAAA | 6357 | join_ans | 16 | RFC-AAAA | 6358 | leave_req | 17 | RFC-AAAA | 6359 | leave_ans | 18 | RFC-AAAA | 6360 | update_req | 19 | RFC-AAAA | 6361 | update_ans | 20 | RFC-AAAA | 6362 | route_query_req | 21 | RFC-AAAA | 6363 | route_query_ans | 22 | RFC-AAAA | 6364 | ping_req | 23 | RFC-AAAA | 6365 | ping_ans | 24 | RFC-AAAA | 6366 | stat_req | 25 | RFC-AAAA | 6367 | stat_ans | 26 | RFC-AAAA | 6368 | unused (was attachlite_req) | 27 | RFC-AAAA | 6369 | unused (was attachlite_ans) | 28 | RFC-AAAA | 6370 | app_attach_req | 29 | RFC-AAAA | 6371 | app_attach_ans | 30 | RFC-AAAA | 6372 | unused (was app_attachlite_req) | 31 | RFC-AAAA | 6373 | unused (was app_attachlite_ans) | 32 | RFC-AAAA | 6374 | config_update_req | 33 | RFC-AAAA | 6375 | config_update_ans | 34 | RFC-AAAA | 6376 | reserved | 0x8000..0xfffe | RFC-AAAA | 6377 | error | 0xffff | RFC-AAAA | 6378 +---------------------------------+----------------+----------+ 6380 13.9. Error Codes 6382 IANA SHALL create a "RELOAD Error Code" Registry. Entries in this 6383 registry are 16-bit integers denoting error codes. New entries SHALL 6384 be defined via RFC 5226 Standards Action. The initial contents of 6385 this registry are: 6387 +-------------------------------------+----------------+----------+ 6388 | Error Code Name | Code Value | RFC | 6389 +-------------------------------------+----------------+----------+ 6390 | invalid | 0 | RFC-AAAA | 6391 | Unused | 1 | RFC-AAAA | 6392 | Error_Forbidden | 2 | RFC-AAAA | 6393 | Error_Not_Found | 3 | RFC-AAAA | 6394 | Error_Request_Timeout | 4 | RFC-AAAA | 6395 | Error_Generation_Counter_Too_Low | 5 | RFC-AAAA | 6396 | Error_Incompatible_with_Overlay | 6 | RFC-AAAA | 6397 | Error_Unsupported_Forwarding_Option | 7 | RFC-AAAA | 6398 | Error_Data_Too_Large | 8 | RFC-AAAA | 6399 | Error_Data_Too_Old | 9 | RFC-AAAA | 6400 | Error_TTL_Exceeded | 10 | RFC-AAAA | 6401 | Error_Message_Too_Large | 11 | RFC-AAAA | 6402 | Error_Unknown_Kind | 12 | RFC-AAAA | 6403 | Error_Unknown_Extension | 13 | RFC-AAAA | 6404 | Error_Response_Too_Large | 14 | RFC-AAAA | 6405 | Error_Config_Too_Old | 15 | RFC-AAAA | 6406 | Error_Config_Too_New | 16 | RFC-AAAA | 6407 | Error_In_Progress | 17 | RFC-AAAA | 6408 | reserved | 0x8000..0xfffe | RFC-AAAA | 6409 +-------------------------------------+----------------+----------+ 6411 13.10. Overlay Link Types 6413 IANA shall create a "RELOAD Overlay Link." New entries SHALL be 6414 defined via RFC 5226 Standards Action. This registry SHALL be 6415 initially populated with the following values: 6417 +--------------------+------+---------------+ 6418 | Protocol | Code | Specification | 6419 +--------------------+------+---------------+ 6420 | reserved | 0 | RFC-AAAA | 6421 | DTLS-UDP-SR | 1 | RFC-AAAA | 6422 | DTLS-UDP-SR-NO-ICE | 3 | RFC-AAAA | 6423 | TLS-TCP-FH-NO-ICE | 4 | RFC-AAAA | 6424 | reserved | 255 | RFC-AAAA | 6425 +--------------------+------+---------------+ 6427 13.11. Overlay Link Protocols 6429 IANA shall create an "Overlay Link Protocol Registry". Entries in 6430 this registry SHALL be defined via RFC 5226 Standards Action. This 6431 registry SHALL be initially populated with the following value: 6432 "TLS". 6434 13.12. Forwarding Options 6436 IANA shall create a "Forwarding Option Registry". Entries in this 6437 registry between 1 and 127 SHALL be defined via RFC 5226 Standards 6438 Action. Entries in this registry between 128 and 254 SHALL be 6439 defined via RFC 5226 Specification Required. This registry SHALL be 6440 initially populated with the following values: 6442 +-------------------+------+---------------+ 6443 | Forwarding Option | Code | Specification | 6444 +-------------------+------+---------------+ 6445 | invalid | 0 | RFC-AAAA | 6446 | reserved | 255 | RFC-AAAA | 6447 +-------------------+------+---------------+ 6449 13.13. Probe Information Types 6451 IANA shall create a "RELOAD Probe Information Type Registry". 6452 Entries in this registry SHALL be defined via RFC 5226 Standards 6453 Action. This registry SHALL be initially populated with the 6454 following values: 6456 +-----------------+------+---------------+ 6457 | Probe Option | Code | Specification | 6458 +-----------------+------+---------------+ 6459 | invalid | 0 | RFC-AAAA | 6460 | responsible_set | 1 | RFC-AAAA | 6461 | num_resources | 2 | RFC-AAAA | 6462 | uptime | 3 | RFC-AAAA | 6463 | reserved | 255 | RFC-AAAA | 6464 +-----------------+------+---------------+ 6466 13.14. Message Extensions 6468 IANA shall create a "RELOAD Extensions Registry". Entries in this 6469 registry SHALL be defined via RFC 5226 Specification Required. This 6470 registry SHALL be initially populated with the following values: 6472 +-----------------+--------+---------------+ 6473 | Extensions Name | Code | Specification | 6474 +-----------------+--------+---------------+ 6475 | invalid | 0 | RFC-AAAA | 6476 | reserved | 0xFFFF | RFC-AAAA | 6477 +-----------------+--------+---------------+ 6479 13.15. reload URI Scheme 6481 This section describes the scheme for a reload URI, which can be used 6482 to refer to either: 6484 o A peer. 6485 o A resource inside a peer. 6487 The reload URI is defined using a subset of the URI schema specified 6488 in Appendix A of RFC 3986 [RFC3986] and the associated URI Guidelines 6489 [RFC4395] per the following ABNF syntax: 6491 RELOAD-URI = "reload://" destination "@" overlay "/" 6492 [specifier] 6494 destination = 1 * HEXDIG 6495 overlay = reg-name 6496 specifier = 1*HEXDIG 6498 The definitions of these productions are as follows: 6500 destination: a hex-encoded Destination List object (i.e., multiple 6501 concatenated Destination objects with no length prefix prior to 6502 the object as a whole.) 6504 overlay: the name of the overlay. 6506 specifier : a hex-encoded StoredDataSpecifier indicating the data 6507 element. 6509 If no specifier is present then this URI addresses the peer which can 6510 be reached via the indicated destination list at the indicated 6511 overlay name. If a specifier is present, then the URI addresses the 6512 data value. 6514 13.15.1. URI Registration 6516 [[ Note to RFC Editor - please remove this paragraph before 6517 publication. ]] Review request was sent to uri-review@ietf.org on Oct 6518 7, 2010. 6520 The following summarizes the information necessary to register the 6521 reload URI. 6523 URI Scheme Name: reload 6524 Status: permanent 6525 URI Scheme Syntax: see Section 13.15 of RFC-AAAA 6526 URI Scheme Semantics: The reload URI is intended to be used as a 6527 reference to a RELOAD peer or resource. 6528 Encoding Considerations: The reload URI is not intended to be human- 6529 readable text, so it is encoded entirely in US-ASCII. 6530 Applications/protocols that use this URI scheme: The RELOAD protocol 6531 described in RFC-AAAA. 6532 Interoperability considerations: See RFC-AAAA. 6533 Security considerations: See RFC-AAAA 6534 Contact: Cullen Jennings 6535 Author/Change controller: IESG 6536 References: RFC-AAAA 6538 14. Acknowledgments 6540 This specification is a merge of the "REsource LOcation And Discovery 6541 (RELOAD)" draft by David A. Bryan, Marcia Zangrilli and Bruce B. 6542 Lowekamp, the "Address Settlement by Peer to Peer" draft by Cullen 6543 Jennings, Jonathan Rosenberg, and Eric Rescorla, the "Security 6544 Extensions for RELOAD" draft by Bruce B. Lowekamp and James Deverick, 6545 the "A Chord-based DHT for Resource Lookup in P2PSIP" by Marcia 6546 Zangrilli and David A. Bryan, and the Peer-to-Peer Protocol (P2PP) 6547 draft by Salman A. Baset, Henning Schulzrinne, and Marcin 6548 Matuszewski. Thanks to the authors of RFC 5389 for text included 6549 from that. Vidya Narayanan provided many comments and improvements. 6551 The ideas and text for the Chord specific extension data to the Leave 6552 mechanisms was provided by J. Maenpaa, G. Camarillo, and J. 6553 Hautakorpi. 6555 Thanks to the many people who contributed including Ted Hardie, 6556 Michael Chen, Dan York, Das Saumitra, Lyndsay Campbell, Brian Rosen, 6557 David Bryan, Dave Craig, and Julian Cain. Extensive working last 6558 call comments were provided by: Jouni Maenpaa, Roni Even, Ari 6559 Keranen, John Buford, Michaelx Chen, Frederic-Philippe Met, and David 6560 Bryan. Special thanks to Marc Petit-Huguenin who provied an amazing 6561 amount to detailed review. 6563 15. References 6565 15.1. Normative References 6567 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 6568 Requirement Levels", BCP 14, RFC 2119, March 1997. 6570 [RFC2585] Housley, R. and P. Hoffman, "Internet X.509 Public Key 6571 Infrastructure Operational Protocols: FTP and HTTP", 6572 RFC 2585, May 1999. 6574 [RFC2818] Rescorla, E., "HTTP Over TLS", RFC 2818, May 2000. 6576 [RFC2988] Paxson, V. and M. Allman, "Computing TCP's Retransmission 6577 Timer", RFC 2988, November 2000. 6579 [RFC3174] Eastlake, D. and P. Jones, "US Secure Hash Algorithm 1 6580 (SHA1)", RFC 3174, September 2001. 6582 [RFC3447] Jonsson, J. and B. Kaliski, "Public-Key Cryptography 6583 Standards (PKCS) #1: RSA Cryptography Specifications 6584 Version 2.1", RFC 3447, February 2003. 6586 [RFC3986] Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform 6587 Resource Identifier (URI): Generic Syntax", STD 66, 6588 RFC 3986, January 2005. 6590 [RFC4279] Eronen, P. and H. Tschofenig, "Pre-Shared Key Ciphersuites 6591 for Transport Layer Security (TLS)", RFC 4279, 6592 December 2005. 6594 [RFC4347] Rescorla, E. and N. Modadugu, "Datagram Transport Layer 6595 Security", RFC 4347, April 2006. 6597 [RFC4395] Hansen, T., Hardie, T., and L. Masinter, "Guidelines and 6598 Registration Procedures for New URI Schemes", BCP 35, 6599 RFC 4395, February 2006. 6601 [RFC4634] Eastlake, D. and T. Hansen, "US Secure Hash Algorithms 6602 (SHA and HMAC-SHA)", RFC 4634, July 2006. 6604 [RFC4648] Josefsson, S., "The Base16, Base32, and Base64 Data 6605 Encodings", RFC 4648, October 2006. 6607 [RFC5245] Rosenberg, J., "Interactive Connectivity Establishment 6608 (ICE): A Protocol for Network Address Translator (NAT) 6609 Traversal for Offer/Answer Protocols", RFC 5245, 6610 April 2010. 6612 [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security 6613 (TLS) Protocol Version 1.2", RFC 5246, August 2008. 6615 [RFC5272] Schaad, J. and M. Myers, "Certificate Management over CMS 6616 (CMC)", RFC 5272, June 2008. 6618 [RFC5273] Schaad, J. and M. Myers, "Certificate Management over CMS 6619 (CMC): Transport Protocols", RFC 5273, June 2008. 6621 [RFC5348] Floyd, S., Handley, M., Padhye, J., and J. Widmer, "TCP 6622 Friendly Rate Control (TFRC): Protocol Specification", 6623 RFC 5348, September 2008. 6625 [RFC5389] Rosenberg, J., Mahy, R., Matthews, P., and D. Wing, 6626 "Session Traversal Utilities for NAT (STUN)", RFC 5389, 6627 October 2008. 6629 [RFC5766] Mahy, R., Matthews, P., and J. Rosenberg, "Traversal Using 6630 Relays around NAT (TURN): Relay Extensions to Session 6631 Traversal Utilities for NAT (STUN)", RFC 5766, April 2010. 6633 [RFC5952] Kawamura, S. and M. Kawashima, "A Recommendation for IPv6 6634 Address Text Representation", RFC 5952, August 2010. 6636 15.2. Informative References 6638 [Chord] Stoica, I., Morris, R., Liben-Nowell, D., Karger, D., 6639 Kaashoek, M., Dabek, F., and H. Balakrishnan, "Chord: A 6640 Scalable Peer-to-peer Lookup Protocol for Internet 6641 Applications", IEEE/ACM Transactions on Networking Volume 6642 11, Issue 1, 17-32, Feb 2003. 6644 [Eclipse] Singh, A., Ngan, T., Druschel, T., and D. Wallach, 6645 "Eclipse Attacks on Overlay Networks: Threats and 6646 Defenses", INFOCOM 2006, April 2006. 6648 [I-D.baset-tsvwg-tcp-over-udp] 6649 Baset, S. and H. Schulzrinne, "TCP-over-UDP", 6650 draft-baset-tsvwg-tcp-over-udp-01 (work in progress), 6651 June 2009. 6653 [I-D.ietf-hip-bone] 6654 Camarillo, G., Nikander, P., Hautakorpi, J., Keranen, A., 6655 and A. Johnston, "HIP BONE: Host Identity Protocol (HIP) 6656 Based Overlay Networking Environment", 6657 draft-ietf-hip-bone-07 (work in progress), June 2010. 6659 [I-D.ietf-hip-reload-instance] 6660 Keranen, A., Camarillo, G., and J. Maenpaa, "Host Identity 6661 Protocol-Based Overlay Networking Environment (HIP BONE) 6662 Instance Specification for REsource LOcation And Discovery 6663 (RELOAD)", draft-ietf-hip-reload-instance-03 (work in 6664 progress), January 2011. 6666 [I-D.ietf-mmusic-ice-tcp] 6667 Rosenberg, J., Keranen, A., Lowekamp, B., and A. Roach, 6668 "TCP Candidates with Interactive Connectivity 6669 Establishment (ICE)", draft-ietf-mmusic-ice-tcp-12 (work 6670 in progress), February 2011. 6672 [I-D.ietf-p2psip-concepts] 6673 Bryan, D., Matthews, P., Shim, E., Willis, D., and S. 6674 Dawkins, "Concepts and Terminology for Peer to Peer SIP", 6675 draft-ietf-p2psip-concepts-03 (work in progress), 6676 October 2010. 6678 [I-D.ietf-p2psip-sip] 6679 Jennings, C., Lowekamp, B., Rescorla, E., Baset, S., and 6680 H. Schulzrinne, "A SIP Usage for RELOAD", 6681 draft-ietf-p2psip-sip-05 (work in progress), July 2010. 6683 [I-D.jiang-p2psip-relay] 6684 Jiang, X., Zong, N., Even, R., and Y. Zhang, "An extension 6685 to RELOAD to support Direct Response and Relay Peer 6686 routing", draft-jiang-p2psip-relay-04 (work in progress), 6687 April 2010. 6689 [I-D.maenpaa-p2psip-self-tuning] 6690 Maenpaa, J., Camarillo, G., and J. Hautakorpi, "A Self- 6691 tuning Distributed Hash Table (DHT) for REsource LOcation 6692 And Discovery (RELOAD)", 6693 draft-maenpaa-p2psip-self-tuning-01 (work in progress), 6694 October 2009. 6696 [I-D.maenpaa-p2psip-service-discovery] 6697 Maenpaa, J. and G. Camarillo, "Service Discovery Usage for 6698 REsource LOcation And Discovery (RELOAD)", 6699 draft-maenpaa-p2psip-service-discovery-00 (work in 6700 progress), October 2009. 6702 [I-D.pascual-p2psip-clients] 6703 Pascual, V., Matuszewski, M., Shim, E., Zhang, H., and S. 6704 Yongchao, "P2PSIP Clients", 6705 draft-pascual-p2psip-clients-01 (work in progress), 6706 February 2008. 6708 [RFC1122] Braden, R., "Requirements for Internet Hosts - 6709 Communication Layers", STD 3, RFC 1122, October 1989. 6711 [RFC2311] Dusse, S., Hoffman, P., Ramsdell, B., Lundblade, L., and 6712 L. Repka, "S/MIME Version 2 Message Specification", 6713 RFC 2311, March 1998. 6715 [RFC4086] Eastlake, D., Schiller, J., and S. Crocker, "Randomness 6716 Requirements for Security", BCP 106, RFC 4086, June 2005. 6718 [RFC4145] Yon, D. and G. Camarillo, "TCP-Based Media Transport in 6719 the Session Description Protocol (SDP)", RFC 4145, 6720 September 2005. 6722 [RFC4787] Audet, F. and C. Jennings, "Network Address Translation 6723 (NAT) Behavioral Requirements for Unicast UDP", BCP 127, 6724 RFC 4787, January 2007. 6726 [RFC4828] Floyd, S. and E. Kohler, "TCP Friendly Rate Control 6727 (TFRC): The Small-Packet (SP) Variant", RFC 4828, 6728 April 2007. 6730 [RFC5054] Taylor, D., Wu, T., Mavrogiannopoulos, N., and T. Perrin, 6731 "Using the Secure Remote Password (SRP) Protocol for TLS 6732 Authentication", RFC 5054, November 2007. 6734 [RFC5201] Moskowitz, R., Nikander, P., Jokela, P., and T. Henderson, 6735 "Host Identity Protocol", RFC 5201, April 2008. 6737 [RFC5280] Cooper, D., Santesson, S., Farrell, S., Boeyen, S., 6738 Housley, R., and W. Polk, "Internet X.509 Public Key 6739 Infrastructure Certificate and Certificate Revocation List 6740 (CRL) Profile", RFC 5280, May 2008. 6742 [RFC5765] Schulzrinne, H., Marocco, E., and E. Ivov, "Security 6743 Issues and Solutions in Peer-to-Peer Systems for Realtime 6744 Communications", RFC 5765, February 2010. 6746 [RFC5785] Nottingham, M. and E. Hammer-Lahav, "Defining Well-Known 6747 Uniform Resource Identifiers (URIs)", RFC 5785, 6748 April 2010. 6750 [Sybil] Douceur, J., "The Sybil Attack", IPTPS 02, March 2002. 6752 [UnixTime] 6753 Wikipedia, "Unix Time", . 6756 [bryan-design-hotp2p08] 6757 Bryan, D., Lowekamp, B., and M. Zangrilli, "The Design of 6758 a Versatile, Secure P2PSIP Communications Architecture for 6759 the Public Internet", Hot-P2P'08. 6761 [handling-churn-usenix04] 6762 Rhea, S., Geels, D., Roscoe, T., and J. Kubiatowicz, 6763 "Handling Churn in a DHT", In Proc. of the USENIX Annual 6764 Technical Conference June 2004 USENIX 2004. 6766 [lookups-churn-p2p06] 6767 Wu, D., Tian, Y., and K. Ng, "Analytical Study on 6768 Improving DHT Lookup Performance under Churn", IEEE 6769 P2P'06. 6771 [minimizing-churn-sigcomm06] 6772 Godfrey, P., Shenker, S., and I. Stoica, "Minimizing Churn 6773 in Distributed Systems", SIGCOMM 2006. 6775 [non-transitive-dhts-worlds05] 6776 Freedman, M., Lakshminarayanan, K., Rhea, S., and I. 6777 Stoica, "Non-Transitive Connectivity and DHTs", 6778 WORLDS'05. 6780 [opendht-sigcomm05] 6781 Rhea, S., Godfrey, B., Karp, B., Kubiatowicz, J., 6782 Ratnasamy, S., Shenker, S., Stoica, I., and H. Yu, 6783 "OpenDHT: A Public DHT and its Uses", SIGCOMM'05. 6785 [vulnerabilities-acsac04] 6786 Srivatsa, M. and L. Liu, "Vulnerabilities and Security 6787 Threats in Structured Peer-to-Peer Systems: A Quantitative 6788 Analysis", ACSAC 2004. 6790 Appendix A. Change Log 6792 A.1. Changes since draft-ietf-p2psip-reload-12 6794 o Clarified lifetime management. 6795 o Made it clear how direct return response could be put in an 6796 extension . 6797 o Clarified distinction between enrollment and configuration 6798 servers. 6799 o Clarified that the KindId is 32 bits long. The -12 draft had some 6800 text that said 32 bits and one typo that said 16. Earlier drafts 6801 were said 32 bits. 6802 o Miscellaneous editorial. 6803 o Specified the Chord hash was SHA-1 truncated 6804 o Clarified Attache procedures. 6805 o Changed name in XML configuration from chord-reload-reactive to 6806 chord-reactive. Made other chord-reload- names consistent. 6807 o Fixed the relax NG and made a more complex configuration example. 6809 o Added Error_In_Progress to Error Code table 6811 Appendix B. Routing Alternatives 6813 Significant discussion has been focused on the selection of a routing 6814 algorithm for P2PSIP. This section discusses the motivations for 6815 selecting symmetric recursive routing for RELOAD and describes the 6816 extensions that would be required to support additional routing 6817 algorithms. 6819 B.1. Iterative vs Recursive 6821 Iterative routing has a number of advantages. It is easier to debug, 6822 consumes fewer resources on intermediate peers, and allows the 6823 querying peer to identify and route around misbehaving peers 6824 [non-transitive-dhts-worlds05]. However, in the presence of NATs, 6825 iterative routing is intolerably expensive because a new connection 6826 must be established for each hop (using ICE) [bryan-design-hotp2p08]. 6828 Iterative routing is supported through the RouteQuery mechanism and 6829 is primarily intended for debugging. It also allows the querying 6830 peer to evaluate the routing decisions made by the peers at each hop, 6831 consider alternatives, and perhaps detect at what point the 6832 forwarding path fails. 6834 B.2. Symmetric vs Forward response 6836 An alternative to the symmetric recursive routing method used by 6837 RELOAD is Forward-Only routing, where the response is routed to the 6838 requester as if it were a new message initiated by the responder (in 6839 the previous example, Z sends the response to A as if it were sending 6840 a request). Forward-only routing requires no state in either the 6841 message or intermediate peers. 6843 The drawback of forward-only routing is that it does not work when 6844 the overlay is unstable. For example, if A is in the process of 6845 joining the overlay and is sending a Join request to Z, it is not yet 6846 reachable via forward routing. Even if it is established in the 6847 overlay, if network failures produce temporary instability, A may not 6848 be reachable (and may be trying to stabilize its network connectivity 6849 via Attach messages). 6851 Furthermore, forward-only responses are less likely to reach the 6852 querying peer than symmetric recursive ones are, because the forward 6853 path is more likely to have a failed peer than is the request path 6854 (which was just tested to route the request) 6855 [non-transitive-dhts-worlds05]. 6857 An extension to RELOAD that supports forward-only routing but relies 6858 on symmetric responses as a fallback would be possible, but due to 6859 the complexities of determining when to use forward-only and when to 6860 fallback to symmetric, we have chosen not to include it as an option 6861 at this point. 6863 B.3. Direct Response 6865 Another routing option is Direct Response routing, in which the 6866 response is returned directly to the querying node. In the previous 6867 example, if A encodes its IP address in the request, then Z can 6868 simply deliver the response directly to A. In the absence of NATs or 6869 other connectivity issues, this is the optimal routing technique. 6871 The challenge of implementing direct response is the presence of 6872 NATs. There are a number of complexities that must be addressed. In 6873 this discussion, we will continue our assumption that A issued the 6874 request and Z is generating the response. 6876 o The IP address listed by A may be unreachable, either due to NAT 6877 or firewall rules. Therefore, a direct response technique must 6878 fallback to symmetric response [non-transitive-dhts-worlds05]. 6879 The hop-by-hop ACKs used by RELOAD allow Z to determine when A has 6880 received the message (and the TLS negotiation will provide earlier 6881 confirmation that A is reachable), but this fallback requires a 6882 timeout that will increase the response latency whenever A is not 6883 reachable from Z. 6884 o Whenever A is behind a NAT it will have multiple candidate IP 6885 addresses, each of which must be advertised to ensure 6886 connectivity; therefore Z will need to attempt multiple 6887 connections to deliver the response. 6888 o One (or all) of A's candidate addresses may route from Z to a 6889 different device on the Internet. In the worst case these nodes 6890 may actually be running RELOAD on the same port. Therefore, it is 6891 absolutely necessary to establish a secure connection to 6892 authenticate A before delivering the response. This step 6893 diminishes the efficiency of direct response because multiple 6894 roundtrips are required before the message can be delivered. 6895 o If A is behind a NAT and does not have a connection already 6896 established with Z, there are only two ways the direct response 6897 will work. The first is that A and Z both be behind the same NAT, 6898 in which case the NAT is not involved. In the more common case, 6899 when Z is outside A's NAT, the response will only be received if 6900 A's NAT implements endpoint-independent filtering. As the choice 6901 of filtering mode conflates application transparency with security 6902 [RFC4787], and no clear recommendation is available, the 6903 prevalence of this feature in future devices remains unclear. 6905 An extension to RELOAD that supports direct response routing but 6906 relies on symmetric responses as a fallback would be possible, but 6907 due to the complexities of determining when to use direct response 6908 and when to fallback to symmetric, and the reduced performance for 6909 responses to peers behind restrictive NATs, we have chosen not to 6910 include it as an option at this point. 6912 B.4. Relay Peers 6914 [I-D.jiang-p2psip-relay] has proposed implementing a form of direct 6915 response by having A identify a peer, Q, that will be directly 6916 reachable by any other peer. A uses Attach to establish a connection 6917 with Q and advertises Q's IP address in the request sent to Z. Z 6918 sends the response to Q, which relays it to A. This then reduces the 6919 latency to two hops, plus Z negotiating a secure connection to Q. 6921 This technique relies on the relative population of nodes such as A 6922 that require relay peers and peers such as Q that are capable of 6923 serving as a relay peer. It also requires nodes to be able to 6924 identify which category they are in. This identification problem has 6925 turned out to be hard to solve and is still an open area of 6926 exploration. 6928 An extension to RELOAD that supports relay peers is possible, but due 6929 to the complexities of implementing such an alternative, we have not 6930 added such a feature to RELOAD at this point. 6932 A concept similar to relay peers, essentially choosing a relay peer 6933 at random, has previously been suggested to solve problems of 6934 pairwise non-transitivity [non-transitive-dhts-worlds05], but 6935 deterministic filtering provided by NATs makes random relay peers no 6936 more likely to work than the responding peer. 6938 B.5. Symmetric Route Stability 6940 A common concern about symmetric recursive routing has been that one 6941 or more peers along the request path may fail before the response is 6942 received. The significance of this problem essentially depends on 6943 the response latency of the overlay. An overlay that produces slow 6944 responses will be vulnerable to churn, whereas responses that are 6945 delivered very quickly are vulnerable only to failures that occur 6946 over that small interval. 6948 The other aspect of this issue is whether the request itself can be 6949 successfully delivered. Assuming typical connection maintenance 6950 intervals, the time period between the last maintenance and the 6951 request being sent will be orders of magnitude greater than the delay 6952 between the request being forwarded and the response being received. 6954 Therefore, if the path was stable enough to be available to route the 6955 request, it is almost certainly going to remain available to route 6956 the response. 6958 An overlay that is unstable enough to suffer this type of failure 6959 frequently is unlikely to be able to support reliable functionality 6960 regardless of the routing mechanism. However, regardless of the 6961 stability of the return path, studies show that in the event of high 6962 churn, iterative routing is a better solution to ensure request 6963 completion [lookups-churn-p2p06] [non-transitive-dhts-worlds05] 6965 Finally, because RELOAD retries the end-to-end request, that retry 6966 will address the issues of churn that remain. 6968 Appendix C. Why Clients? 6970 There are a wide variety of reasons a node may act as a client rather 6971 than as a peer [I-D.pascual-p2psip-clients]. This section outlines 6972 some of those scenarios and how the client's behavior changes based 6973 on its capabilities. 6975 C.1. Why Not Only Peers? 6977 For a number of reasons, a particular node may be forced to act as a 6978 client even though it is willing to act as a peer. These include: 6980 o The node does not have appropriate network connectivity, typically 6981 because it has a low-bandwidth network connection. 6982 o The node may not have sufficient resources, such as computing 6983 power, storage space, or battery power. 6984 o The overlay algorithm may dictate specific requirements for peer 6985 selection. These may include participating in the overlay to 6986 determine trustworthiness; controlling the number of peers in the 6987 overlay to reduce overly-long routing paths; or ensuring minimum 6988 application uptime before a node can join as a peer. 6990 The ultimate criteria for a node to become a peer are determined by 6991 the overlay algorithm and specific deployment. A node acting as a 6992 client that has a full implementation of RELOAD and the appropriate 6993 overlay algorithm is capable of locating its responsible peer in the 6994 overlay and using Attach to establish a direct connection to that 6995 peer. In that way, it may elect to be reachable under either of the 6996 routing approaches listed above. Particularly for overlay algorithms 6997 that elect nodes to serve as peers based on trustworthiness or 6998 population, the overlay algorithm may require such a client to locate 6999 itself at a particular place in the overlay. 7001 C.2. Clients as Application-Level Agents 7003 SIP defines an extensive protocol for registration and security 7004 between a client and its registrar/proxy server(s). Any SIP device 7005 can act as a client of a RELOAD-based P2PSIP overlay if it contacts a 7006 peer that implements the server-side functionality required by the 7007 SIP protocol. In this case, the peer would be acting as if it were 7008 the user's peer, and would need the appropriate credentials for that 7009 user. 7011 Application-level support for clients is defined by a usage. A usage 7012 offering support for application-level clients should specify how the 7013 security of the system is maintained when the data is moved between 7014 the application and RELOAD layers. 7016 Authors' Addresses 7018 Cullen Jennings 7019 Cisco 7020 170 West Tasman Drive 7021 MS: SJC-21/2 7022 San Jose, CA 95134 7023 USA 7025 Phone: +1 408 421-9990 7026 Email: fluffy@cisco.com 7028 Bruce B. Lowekamp (editor) 7029 Skype 7030 Palo Alto, CA 7031 USA 7033 Email: bbl@lowekamp.net 7035 Eric Rescorla 7036 RTFM, Inc. 7037 2064 Edgewood Drive 7038 Palo Alto, CA 94303 7039 USA 7041 Phone: +1 650 678 2350 7042 Email: ekr@rtfm.com 7043 Salman A. Baset 7044 Columbia University 7045 1214 Amsterdam Avenue 7046 New York, NY 7047 USA 7049 Email: salman@cs.columbia.edu 7051 Henning Schulzrinne 7052 Columbia University 7053 1214 Amsterdam Avenue 7054 New York, NY 7055 USA 7057 Email: hgs@cs.columbia.edu