idnits 2.17.1 draft-ietf-cdni-footprint-capabilities-semantics-04.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (October 27, 2014) is 3469 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- ** Obsolete normative reference: RFC 5226 (Obsoleted by RFC 8126) == Outdated reference: A later version (-27) exists of draft-ietf-cdni-logging-14 == Outdated reference: A later version (-21) exists of draft-ietf-cdni-metadata-07 Summary: 1 error (**), 0 flaws (~~), 3 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 CDNI J. Seedorf 3 Internet-Draft NEC 4 Intended status: Informational J. Peterson 5 Expires: April 30, 2015 Neustar 6 S. Previdi 7 Cisco 8 R. van Brandenburg 9 TNO 10 K. Ma 11 Ericsson 12 October 27, 2014 14 CDNI Request Routing: Footprint and Capabilities Semantics 15 draft-ietf-cdni-footprint-capabilities-semantics-04 17 Abstract 19 This document tries to capture the semantics of the "Footprint and 20 Capabilities Advertisement" part of the CDNI Request Routing 21 interface, i.e., the desired meaning and what "Footprint and 22 Capabilities Advertisement" is expected to offer within CDNI. The 23 discussion in this document has the goal to facilitate the choosing 24 of one or more suitable protocols for "Footprint and Capabilities 25 Advertisement" within CDNI Request Routing. 27 Requirements Language 29 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 30 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 31 document are to be interpreted as described in RFC 2119 [RFC2119]. 33 Status of This Memo 35 This Internet-Draft is submitted in full conformance with the 36 provisions of BCP 78 and BCP 79. 38 Internet-Drafts are working documents of the Internet Engineering 39 Task Force (IETF). Note that other groups may also distribute 40 working documents as Internet-Drafts. The list of current Internet- 41 Drafts is at http://datatracker.ietf.org/drafts/current/. 43 Internet-Drafts are draft documents valid for a maximum of six months 44 and may be updated, replaced, or obsoleted by other documents at any 45 time. It is inappropriate to use Internet-Drafts as reference 46 material or to cite them other than as "work in progress." 48 This Internet-Draft will expire on April 30, 2015. 50 Copyright Notice 52 Copyright (c) 2014 IETF Trust and the persons identified as the 53 document authors. All rights reserved. 55 This document is subject to BCP 78 and the IETF Trust's Legal 56 Provisions Relating to IETF Documents 57 (http://trustee.ietf.org/license-info) in effect on the date of 58 publication of this document. Please review these documents 59 carefully, as they describe your rights and restrictions with respect 60 to this document. Code Components extracted from this document must 61 include Simplified BSD License text as described in Section 4.e of 62 the Trust Legal Provisions and are provided without warranty as 63 described in the Simplified BSD License. 65 Table of Contents 67 1. Introduction and Scope . . . . . . . . . . . . . . . . . . . 3 68 2. Design Decisions for Footprint and Capabilities . . . . . . . 4 69 2.1. Advertising Limited Coverage . . . . . . . . . . . . . . 4 70 2.2. Capabilities and Dynamic Data . . . . . . . . . . . . . . 5 71 2.3. Advertisement versus Queries . . . . . . . . . . . . . . 6 72 2.4. Avoiding or Handling 'cheating' dCDNs . . . . . . . . . . 7 73 2.5. Focusing on Main Use Cases . . . . . . . . . . . . . . . 7 74 3. Main Use Case to Consider . . . . . . . . . . . . . . . . . . 8 75 4. Semantics for Footprint Advertisement . . . . . . . . . . . . 8 76 5. Semantics for Capabilities Advertisement . . . . . . . . . . 11 77 6. Negotiation of Support for Optional Types of 78 Footprint/Capabilities . . . . . . . . . . . . . . . . . . . 14 79 7. Capability Advertisement Object . . . . . . . . . . . . . . . 14 80 7.1. Base Advertisement Object . . . . . . . . . . . . . . . . 14 81 7.2. Redirection Mode Advertisement Object . . . . . . . . . . 15 82 7.3. Advertisement Object Serialization . . . . . . . . . . . 15 83 8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 16 84 8.1. Capabilities Registry . . . . . . . . . . . . . . . . . . 16 85 8.2. Footprint Sub-Registry . . . . . . . . . . . . . . . . . 16 86 8.3. Protocol Sub-Registry . . . . . . . . . . . . . . . . . . 17 87 8.4. Redirection Mode Sub-Registry . . . . . . . . . . . . . . 17 88 8.5. Logging Record Type Sub-Registry . . . . . . . . . . . . 17 89 8.6. Logging Field Name Sub-Registry . . . . . . . . . . . . . 17 90 8.7. Metadata Type Sub-Registry . . . . . . . . . . . . . . . 17 91 9. Security Considerations . . . . . . . . . . . . . . . . . . . 18 92 10. References . . . . . . . . . . . . . . . . . . . . . . . . . 18 93 10.1. Normative References . . . . . . . . . . . . . . . . . . 18 94 10.2. Informative References . . . . . . . . . . . . . . . . . 19 95 Appendix A. Acknowledgment . . . . . . . . . . . . . . . . . . . 19 96 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 20 98 1. Introduction and Scope 100 The CDNI working group is working on a set of protocols to enable the 101 interconnection of multiple CDNs to a CDN federation. This CDN- 102 federation should serve multiple purposes, as discussed in [RFC6770], 103 for instance, to extend the reach of a given CDN to areas in the 104 network which are not covered by this particular CDN. 106 The goal of this document is to achieve a clear understanding in the 107 CDNI WG about the semantics associated with the CDNI Request Routing 108 Footprint & Capabilities Advertisement Interface (from now on 109 referred to as FCI), in particular the type of information a 110 downstream CDN 'advertises' regarding its footprint and capabilities. 111 To narrow down undecided aspects of these semantics, this document 112 tries to establish a common understanding of what the FCI should 113 offer and accomplish in the context of CDN Interconnection. 115 It is explicitly outside the scope of this document to decide on 116 specific protocols to use for the FCI. 118 General assumptions in this document: 120 o The CDNs participating in the CDN federation have already 121 performed a boot strap process, i.e., they have connected to each 122 other, either directly or indirectly, and can exchange information 123 amongst each other. 125 o The uCDN has received footprint and/or capability advertisements 126 from a set of dCDNs. Footprint advertisement and capability 127 advertisement need not use the same underlying protocol. 129 o The upstream CDN (uCDN) receives the initial request-routing 130 request from the endpoint requesting the resource. 132 The CDNI Problem Statement [RFC6707] describes footprint and 133 capabilities advertisement as: "[enabling] a Request Routing function 134 in an Upstream CDN to query a Request Routing function in a 135 Downstream CDN to determine if the Downstream CDN is able (and 136 willing) to accept the delegated Content Request". In addition, the 137 RFC says "the CDNI Request Routing interface is also expected to 138 enable a downstream CDN to provide to the upstream CDN (static or 139 dynamic) information (e.g., resources, footprint, load) to facilitate 140 selection of the downstream CDN by the upstream CDN request routing 141 system when processing subsequent content requests from User Agents". 142 It thus considers "resources" and "load" as capabilities to be 143 advertised by the downstream CDN. 145 The range of different footprint definitions and possible 146 capabilities is very broad. Attempting to define a comprehensive 147 advertisement solution quickly becomes intractable. The CDNI 148 requirements draft [I-D.ietf-cdni-requirements] lists the specific 149 requirements for the CDNI Footprint & Capabilities Advertisement 150 Interface in order to disambiguate footprints and capabilities with 151 respect to CDNI. This document attempts to distill the apparent 152 common understanding of what the terms 'footprint' and 'capabilities' 153 mean in the context of CDNI, and detail the semantics of the 154 footprint advertisement mechanism and the capability advertisement 155 mechanism. 157 2. Design Decisions for Footprint and Capabilities 159 A large part of the difficulty in discussing the FCI lies in 160 understanding what exactly is meant when trying to define footprint 161 in terms of "coverage" or "reachability." While the operators of 162 CDNs pick strategic locations to situate caches, a cache with a 163 public IPv4 address is reachable by any endpoint on the Internet 164 unless some policy enforcement precludes the use of the cache. 166 Some CDNs aspire to cover the entire world, which we will henceforth 167 call global CDNs. The footprint advertised by such a CDN in the CDNI 168 environment would, from a coverage or reachability perspective, 169 presumably cover all prefixes. Potentially more interesting for CDNI 170 use cases, however, are CDNs that claim a more limited coverage, but 171 seek to federate with other CDNs in order to create a single CDN 172 fabric which shares resources. 174 Futhermore, not all capabilities need be footprint restricted. 175 Depending upon the use case, the optimal semantics of "footprints 176 with capability attributes" vs. "capabilities with footprint 177 restrictions" are not clear. 179 The key to understanding the semantics of footprint and capability 180 advertisement lies in understand why a dCDN would advertise a limited 181 coverage area, and how a uCDN would use such advertisements to decide 182 among one of several dCDNs. The following section will discuss some 183 of the trade-offs and design decisions that need to be decided upon 184 for the CDNI FCI. 186 2.1. Advertising Limited Coverage 188 The basic use case that would motivate a dCDN to advertise a limited 189 coverage is that the CDN was built to cover only a particular portion 190 of the Internet. For example, an ISP could purpose-build a CDN to 191 serve only their own customers by situating caches in close 192 topological proximity to high concentrations of their subscribers. 194 The ISP knows the prefixes it has allocated to end users and thus can 195 easily construct a list of prefixes that its caches were positioned 196 to serve. 198 When such a purpose-built CDN joins a federation, however, and 199 advertises its footprint to a uCDN, the original intended coverage of 200 the CDN might not represent its actual value to the federation of 201 CDNs. Consider an ISP-A and ISP-B that both field their own CDNs, 202 which they federate through CDNI. A given user E, who is customer of 203 ISP-B, might happen to be topologically closest to a cache fielded by 204 ISP-A, if E happens to live in a region where ISP-B has few customers 205 and ISP-A has many. In this case, should ISP-A's CDN "cover" E? If 206 ISP-B's CDN has a failure condition, should the uCDN understand that 207 ISP-A's caches are potentially available back-ups - and if so, how 208 does ISP-A advertise itself as a "standby" for E? What about the 209 case where CDNs advertising to the same uCDN express overlapping 210 coverage (for example, a federation mixing global and limited CDNs)? 212 The answers to these questions greatly depend on how much information 213 we want the uCDN to use to make a selection of a dCDN. If a uCDN has 214 three dCDNs to choose from that "cover" the IP address of user E, 215 obviously the uCDN might be interested to know how optimal the 216 coverage is from each of the dCDNs - coverage need not be binary, 217 either provided or not provided. dCDNs could advertise a coverage 218 "score," for example, and provided that they all reported scores 219 fairly on the same scale, uCDNs could use that to make their 220 topological optimality decision. Alternatively, dCDNs could for 221 their footprint advertise the IP addresses of their caches rather 222 than prefix "coverage," and let the uCDN decide for itself (based on 223 its own topological intelligence) which dCDN has better resources to 224 serve a given user. 226 In summary, the semantics of advertising footprint depend on whether 227 such qualitative metrics for expressing footprint (such as the 228 coverage 'score' mentioned above) should be part of the CDNI FCI, or 229 if it should focus just on 'binary' footprint. 231 2.2. Capabilities and Dynamic Data 233 In cases where the apparent footprint of dCDNs overlaps, uCDNs might 234 also want to rely on a host of other factors to evaluate the 235 respective merits of dCDNs. These include facts related to the 236 caches themselves, to the network where the cache is deployed, to the 237 nature of the resource sought and to the administrative policies of 238 the respective networks. 240 In the absence of network-layer impediments to reaching caches, the 241 choice to limit coverage is necessarily an administrative policy. 243 Much policy must be agreed upon before CDNs can merge into 244 federations, including questions of membership, compensation, volumes 245 and so on. A uCDN certainly will factor these sorts of 246 considerations into its decision to select a dCDN, but there is 247 probably little need for dCDNs to actually advertise them through an 248 interface - they will be settled out of band as a precondition for 249 federating. 251 Other facts about the dCDN would be expressed through the interface 252 to the uCDN. Some capabilities of a dCDN are static, and some are 253 highly dynamic. Expressing the total storage built into its caches, 254 for example, changes relatively rarely, whereas the amount of storage 255 in use at any given moment is highly volatile. Network bandwidth 256 similarly could be expressed as either total bandwidth available to a 257 cache, or based on the current state of the network. A cache may at 258 one moment lack a particular resource in storage, but have it the 259 next. 261 The semantics of the capabilities interface will depend on how much 262 of the dCDN state needs to be pushed to the uCDN and qualitatively 263 how often that information should be updated. 265 2.3. Advertisement versus Queries 267 In a federated CDN environment, each dCDN shares some of its state 268 with the uCDN, which the uCDN uses to build a unified picture of all 269 of the dCDNs available to it. In architectures that share detailed 270 capability information, the uCDN could basically perform the entire 271 request-routing intelligence down to selecting a particular cache 272 before sending the request to the dCDN (note that within the current 273 CDNI WG scope, such direct selection of specific caches by the uCDN 274 is out of scope). However, when the uCDN must deal with many 275 potential dCDNs, this approach does not scale. Especially as CDNs 276 scale up from dozens or hundreds of caches to thousands or tens of 277 thousands, the volume of updates to footprint and capability may 278 become onerous. 280 Were the volume of updates to exceed the volumes of requests to the 281 uCDN, it might make more sense for the uCDN to query dCDNs upon 282 receiving requests (as is the case in the recursive redirection mode 283 described in [I-D.ietf-cdni-framework]), instead of receiving 284 advertisements and tracking the state of dCDNs itself. The advantage 285 of querying dCDNs would be that much of the dynamic data that dCDNs 286 cannot share with the uCDN would now be factored into the uCDN's 287 decision. dCDNs need not replicate any state to the uCDN - uCDNs 288 could effectively operate in a stateless mode. 290 The semantics of both footprint and capability advertisement depend 291 on the service model here: are there cases where a synchronous query/ 292 response model would work better for the uCDN decision than a state 293 replication model? 295 2.4. Avoiding or Handling 'cheating' dCDNs 297 In a situation where more than one dCDN is willing to serve a given 298 end user request, it might be attractive for a dCDN to 'cheat' in the 299 sense that the dCDN provides inaccurate information to the uCDN in 300 order to convince the uCDN to select it opposed to 'competing' dCDNs. 301 It could therefore be desirable to take away the incentive for dCDNs 302 to cheat (in information advertised) as much as possible. One option 303 here is to make the information the dCDN advertises somehow 304 verifiable for the uCDN. One the other hand, a cheating dCDN might 305 be avoided or handled by the fact that there will be strong 306 contractual agreements between a uCDN and a dCDN, so that a dCDN 307 would risk severe penalties or legal consequences when caught 308 cheating. 310 Overall, it seems that information a dCDN advertises should (in the 311 long run) be somehow qualitatively verifiable by the uCDN, though 312 possibly through non-real-time out-of-band audits. It is probably an 313 overly strict requirement to mandate that such verification be 314 possible "immediately", i.e., during the request routing process 315 itself. If the uCDN can detect a cheating dCDN at a later stage, it 316 should suffice for the uCDN to "de-incentivize" cheating because it 317 would negatively affect the long-term business relationship with a 318 particular dCDN. 320 2.5. Focusing on Main Use Cases 322 To narrow down semantics for "footprint" and "capabilities" in the 323 CDNI context, it can be useful to initially focus on key use cases to 324 be addressed by the CDNI WG that are to be envisioned the main 325 deployments in the foreseeable future. In this regard, a main 326 realistic use case is the existence of ISP-owned CDNs, which 327 essentially cover a certain operator's network. At the same time, 328 however, the possibility of overlapping footprints should not be 329 excluded, i.e., the scenario where more than one dCDN claims it can 330 serve a given end user request. The ISPs may also choose to federate 331 with a fallback global CDN. 333 It seems reasonable to assume that in most use cases it is the uCDN 334 that makes the decision on selecting a certain dCDN for request 335 routing based on information the uCDN has received from this 336 particular dCDN. It may be assumed that 'cheating' CDNs will be 337 dealt with via means outside the scope of CDNI and that the 338 information advertised between CDNs is accurate. In addition, 339 excluding the use of qualitative information (e.g., cache proximity, 340 delivery latency, cache load) to predict the quality of delivery 341 would further simplify the use case allowing it to better focus on 342 the basic functionality of the FCI. 344 3. Main Use Case to Consider 346 Focusing on a main use case that contains a simple (yet somewhat 347 challenging), realistic, and generally imaginable scenario can help 348 in narrowing down the requirements for the CDNI FCI. To this end, 349 the following (simplified) use case can help in clarifying the 350 semantics of footprint and capabilities for CDNI. In particular, the 351 intention of the use case is to clarify what information needs to be 352 exchanged on the CDNI FCI, what types of information need to be 353 supported in a mandatory fashion (and which should be considered 354 optional), and what types of information need to be updated with 355 respect to a priori established CDNI contracts. 357 In short, one can imagine the following use case: A given uCDN has 358 several dCDNs. It selects one dCDN for delivery protocol A and 359 footprint 1 and another dCDN for delivery protocol B and footprint 1. 360 The dCDN that serves delivery protocol B has a further, transitive 361 (level-2) dCDN, that serves delivery protocol B in a subset of 362 footprint 1 where the first-level dCDN cannot serve delivery protocol 363 B itself. What happens if capabilities change in the transitive 364 level-2 dCDN that might affect how the uCDN selects a level-1 dCDN 365 (e.g., in case the level-2 dCDN cannot serve delivery protocol B 366 anymore)? How will these changes be conveyed to the uCDN? In 367 particular, what information does the uCDN need to be able to select 368 a new first-level dCDN, either for all of footprint 1 or only for the 369 subset of footprint 1 that the transitive level-2 dCDN served on 370 behalf of the first-level dCDN? 372 4. Semantics for Footprint Advertisement 374 Roughly speaking, "footprint" can be defined as "ability and 375 willingness to serve" by a downstream CDN. However, in addition to 376 simple "ability and willingness to serve", the uCDN may wish to have 377 additional information to make a dCDN selection decision, e.g., "how 378 well" a given dCDN can actually serve a given end user request. The 379 "ability and willingness" to serve should be distinguished from the 380 subjective qualitative measurement of "how well" it was served. One 381 can imagine that such additional information is implicitly associated 382 with a given footprint, e.g., due to contractual agreements (e.g., 383 SLAs), business relationships, or perceived dCDN quality in the past. 384 As an alternative, such additional information could also be 385 explicitly tagged along with the footprint. 387 It is reasonable to assume that a significant part of the actual 388 footprint advertisement will happen in contractual agreements between 389 participating CDNs, i.e., prior to the advertisement phase using the 390 CDNI FCI. The reason for this assumption is that any contractual 391 agreement is likely to contain specifics about the dCDN coverage 392 (i.e., the dCDN footprint) to which the contractual agreement 393 applies. In particular, additional information to judge the delivery 394 quality associated with a given dCDN footprint might be defined in 395 contractual agreements (i.e. outside of the CDNI FCI). Further, one 396 can assume that dCDN contractual agreements about the delivery 397 quality associated with a given footprint will probably be based on 398 high-level aggregated statistics (i.e., not too detailed). 400 Given that a large part of footprint advertisement will actually 401 happen in contractual agreements, the semantics of CDNI footprint 402 advertisement refer to answering the following question: what exactly 403 still needs to be advertised by the CDNI FCI? For instance, updates 404 about temporal failures of part of a footprint can be useful 405 information to convey via the CDNI request routing interface. Such 406 information would provide updates on information previously agreed in 407 contracts between the participating CDNs. In other words, the CDNI 408 FCI is a means for a dCDN to provide changes/updates regarding a 409 footprint it has prior agreed to serve in a contract with a uCDN. 411 Generally speaking, one can imagine two categories of footprint to be 412 advertised by a dCDN: 414 o Footprint could be defined based on "coverage/reachability", where 415 coverage/reachability refers to a set of prefixes, a geographic 416 region, or similar boundary. The dCDN claims that it can cover/ 417 reach 'end user requests coming from this footprint'. 419 o Footprint could be defined based on "resources", where resources 420 refers to surrogates/caches a dCDN claims to have (e.g., the 421 location of surrogates/resources). The dCDN claims that 'from 422 this footprint' it can serve incoming end user requests. 424 For each of these footprint types, there are capabilities associated 425 with a given footprint, i.e., the capabilities (e.g., delivery 426 protocol, redirection mode, metadata) supported in the coverage area 427 for a "coverage/reachability" defined footprint, or the capabilities 428 of resources (e.g., delivery protocol, redirection mode, metadata 429 support) for a "resources" defined footprint. 431 It seems clear that "coverage/reachability" types of footprint must 432 be supported within CDNI. The following such types of footprint are 433 mandatory and must be supported by the CDNI FCI: 435 o List of ISO Country Codes 437 o List of AS numbers 439 o Set of IP-prefixes 441 A 'set of IP-prefixes' must be able to contain full IP addresses, 442 i.e., a /32 for IPv4 and a /128 for IPv6, and also IP prefixes with 443 an arbitrary prefix length. There must also be support for multiple 444 IP address versions, i.e., IPv4 and IPv6, in such a footprint. 446 "Resource" types of footprints are more specific than "coverage/ 447 reachability" types of footprints, where the actual coverage/ 448 reachability are extrapolated from the resource location (e.g., 449 netmask applied to resource IP address to derive IP-prefix). The 450 specific methods for extrapolating coverage/reachability from 451 resource location are beyond the scope of this document. In the 452 degenerate case, the resource address could be specified as a 453 coverage/reachability type of footprint, in which case no 454 extrapolation is necessary. Resource types of footprints may expose 455 the internal structure of a CDN network which may be undesirable. As 456 such, the resource types of footprints are not considered mandatory 457 to support for CDNI. 459 For all of these mandatory-to-implement footprint types, footprints 460 can be viewed as constraints for delegating requests to a dCDN: A 461 dCDN footprint advertisement tells the uCDN the limitations for 462 delegating a request to the dCDN. For IP prefixes or ASN(s), the 463 footprint signals to the uCDN that it should consider the dCDN a 464 candidate only if the IP address of the request routing source falls 465 within the prefix set (or ASN, respectively). The CDNI 466 specifications do not define how a given uCDN determines what address 467 ranges are in a particular ASN. Similarly, for country codes a uCDN 468 should only consider the dCDN a candidate if it covers the country of 469 the request routing source. The CDNI specifications do not define 470 how a given uCDN determines the country of the request routing 471 source. Multiple footprint constraints are additive, i.e., the 472 advertisement of different types of footprint narrows the dCDN 473 candidacy cumulatively. 475 In addition to these mandatory "coverage/reachability" types of 476 footprint, other optional "coverage/reachability" types of footprint 477 or "resource" types of footprint may defined by future 478 specifications. To facilitate this, a clear process for specifying 479 optional footprint types in a IANA registry must be specified. This 480 includes the specification of the level of oversight necessary (e.g., 481 WG decision or expert review) for adding new optional footprints to a 482 IANA registry as well as the specification of a template regarding 483 design choices that must be captured by new optional types of 484 footprints. 486 Independent of the exact type of a footprint, a footprint might also 487 include the connectivity of a given dCDN to other CDNs that may be 488 able to serve content to users on behalf of that dCDN, to cover cases 489 where there is a transitive CDN interconnection. Further, the 490 downstream CDN must be able to express its footprint to an interested 491 upstream CDN (uCDN) in a comprehensive form, e.g., as a data set 492 containing the complete footprint. Making incremental updates, 493 however, to express dynamic changes in state is also desirable. 495 5. Semantics for Capabilities Advertisement 497 In general, the dCDN must be able to express its general capabilities 498 to the uCDN. These general capabilities could express if the dCDN 499 supports a given service, for instance, HTTP delivery, RTP/RTSP 500 delivery or RTMP. Furthermore, the dCDN must be able to express 501 particular capabilities for the delivery in a particular footprint 502 area. For example, the dCDN might in general offer RTMP but not in 503 some specific areas, either for maintenance reasons or because the 504 caches covering this particular area cannot deliver this type of 505 service. Hence, in certain cases footprint and capabilities are tied 506 together and cannot be interpreted independently from each other. In 507 such cases, i.e., where capabilities must be expressed on a per 508 footprint basis, it may be beneficial to combine footprint and 509 capabilities advertisement. 511 A high-level and very rough semantic for capabilities is thus the 512 following: Capabilities are types of information that allow a uCDN to 513 determine if a downstream CDN is able (and willing) to accept (and 514 properly handle) a delegated content request. In addition, 515 Capabilities are characterized by the fact that this information may 516 possibly change over time based on the state of the network or 517 caches. 519 At a first glance, several broad categories of capabilities seem 520 useful to convey via an advertisement interface, however, advertising 521 capabilities that change highly dynamically (e.g., real-time delivery 522 performance metrics, CDN resource load, or other highly dynamically 523 changing QoS information) should probably not be in scope for the 524 CDNI FCI. First, out of the multitude of possible metrics and 525 capabilities, it is hard to agree on a subset and the precise metrics 526 to be used. Second, and perhaps more importantly, it seems not 527 feasible to specify such highly dynamically changing capabilities and 528 the corresponding metrics within the CDNI charter time-frame. 530 Useful capabilities refer to information that does not change highly 531 dynamically and which in many cases is absolutely necessary to decide 532 on a particular dCDN for a given end user request. For instance, if 533 an end user request concerns the delivery of a video file with a 534 certain protocol (e.g., RTMP), the uCDN needs to know if a given dCDN 535 has the capabilitity of supporting this delivery protocol. 537 Similar to footprint advertisement, it is reasonable to assume that a 538 significant part of the actual (resource) capabilities advertisement 539 will happen in contractual agreements between participating CDNs, 540 i.e. prior to the advertisement phase using the CDNI FCI. The role 541 of capability advertisement is hence rather to enable the dCDN to 542 update a uCDN on changes since a contract has been set up (e.g., in 543 case a new delivery protocol is suddenly being added to the list of 544 supported delivery protocols of a given dCDN, or in case a certain 545 delivery protocol is suddenly not being supported anymore due to 546 failures). Capabilities advertisement thus refers to conveying 547 information to a uCDN about changes/updates of certain capabilities 548 with respect to a given contract. 550 Given these semantics, it needs to be decided what exact capabilities 551 are useful and how these can be expressed. Since the details of CDNI 552 contracts are not known at the time of this writing (and the CDNI 553 interface should probably be agnostic to these contracts anyway), it 554 remains to be seen what capabilities will be used to define 555 agreements between CDNs in practice. One implication for 556 standardization may be to initially only specify a very limited set 557 of mandatory capabilities for advertisement and have on top of that a 558 flexible data model that allows exchanging additional capabilities 559 when needed. Still, agreement needs to be found on which 560 capabilities (if any) should be mandatory among CDNs. As discussed 561 in Section 2.5, finding the concrete answers to these questions can 562 benefit from focusing on a small number of key use cases that are 563 highly relevant and contain enough complexity to help in 564 understanding what concrete capabilities are needed to facilitate CDN 565 Interconnection. 567 Under the above considerations, the following capabilities seem 568 useful as 'base' capabilities, i.e., ones that are needed in any case 569 and therefore constitute mandatory capabilities to be supported by 570 the CDNI FCI: 572 o Delivery Protocol (e.g., HTTP vs. RTMP) 574 o Acquisition Protocol (for aquiring content from a uCDN) 576 o Redirection Mode (e.g., DNS Redirection vs. HTTP Redirection as 577 discussed in [I-D.ietf-cdni-framework]) 579 o CDNI Logging (i.e., supported logging fields) 581 o CDNI Metadata (i.e., supported GenericMetadata types) 583 It is not feasable to enumerate all the possible options for the 584 mandatory capabilities listed above (e.g., all the potential delivery 585 protocols or metadata options) or anticipate all the future needs for 586 additional capabilities. It would be unreasonable to burden the CDNI 587 FCI specification with defining each supported capability. Instead, 588 the CDNI FCI specification should define a generic protocol for 589 conveying any capability information. In this respect, it seems 590 reasonable to define a registry which initially contains the 591 mandatory capabilities listed above, but may be extended as needs 592 dictate. This document defines the registry (and the rules for 593 adding new entries to the registry) for the different capability 594 types (see Section 8). Each capability type MAY have a list of valid 595 values. The individual CDNI interface specifications which define a 596 given capability SHOULD define any necessary registries (and the 597 rules for adding new entries to the registry) for the values 598 advertised for a given capability type. 600 The "CDNI Logging Fields Names" registry defines all supported 601 logging fields, including mandatory-to-implement logging fields. 602 Advertisement of support for mandatory-to-implement logging fields 603 SHOULD be supported but would be redundant. CDNs SHOULD NOT 604 advertise support for mandatory-to-implement logging fields. The 605 following logging fields are defined as optional in the CDNI Logging 606 Interface document [I-D.ietf-cdni-logging]: 608 o c-ip-anonimizing 610 o s-ccid 612 o s-sid 614 The "CDNI GenericMetadata Types" registry defines all supported 615 GenericMetadats types, including mandatory-to-implement 616 GenericMetadata types. Advertisement of support for mandatory-to- 617 implement GenericMetadata types SHOULD be supported but would be 618 redundant. CDNs SHOULD NOT advertise support for mandatory-to- 619 implement GenericMetadata types. The CDNI Metadata Interface 620 document [I-D.ietf-cdni-metadata] does not define any optional 621 GenericMetadata types. 623 6. Negotiation of Support for Optional Types of Footprint/Capabilities 625 The notion of optional types of footprint and capabilities implies 626 that certain implementations may not support all kinds of footprint 627 and capabilities. Therefore, any FCI solution protocol must define 628 how the support for optional types of footprint/capabilities will be 629 negotiated between a uCDN and a dCDN that use the particular FCI 630 protocol. In particular, any FCI solution protocol needs to specify 631 how to handle failure cases or non-supported types of footprint/ 632 capabilities. 634 In general, a uCDN may ignore capabilities or types of footprint it 635 does not understand; in this case it only selects a suitable 636 downstream CDN based on the types of capabilities and footprint it 637 understands. Similarly, if a dCDN does not use an optional 638 capability or footprint which is, however, supported by a uCDN, this 639 causes no problem for the FCI functionality because the uCDN decides 640 on the remaining capabilities/footprint information that is being 641 conveyed by the dCDN. 643 7. Capability Advertisement Object 645 To support extensibility, the FCI defines a generic base object 646 (similar to the CDNI Metadata interface GenericMetadata object) 647 [I-D.ietf-cdni-metadata] to facilitate a uniform set of mandatory 648 parsing requirements for all future FCI objects. 650 [Ed. Note: MI/LI object definitions will build off the base object 651 defined here, but will be specified in a separate draft.] 653 7.1. Base Advertisement Object 655 The FCIBase object is an abstraction for managing individual CDNI 656 capabilities (see Section 8.1) in an opaque manner. 658 Property: capability-type 660 Description: CDNI Capability object type. 662 Type: MIME Type String (from Section 8) 664 Mandatory-to-Specify: Yes. 666 Property: capabilitiy-value 668 Description: CDNI Capability object. 670 Type: Format/Type is defined by the value of capability-type 671 property above. 673 Mandatory-to-Specify: Yes. 675 7.2. Redirection Mode Advertisement Object 677 The Redirection Mode advertisement object is used to indicate support 678 for one or more of the modes listed in the CDNI Capabilities 679 Redirection Modes registry (see Section 8.4). 681 Property: redirection-modes 683 Description: List of supported CDNI Redirection Modes. 685 Type: List of Redirection Modes (from Section 8.4) 687 Mandatory-to-Specify: Yes. 689 [Ed. Note: Redirection Mode is not in MI/LI; it is FCI specific. We 690 define it here as an example for the MI/LI FCI drafts.] 692 7.3. Advertisement Object Serialization 694 { 695 "capabilities": [ 696 { 697 "capability-type": "application/cdni.FCI.RedirectionMode.v1" 698 "capability-value": { 699 "redirection-modes": [ 700 "DNS-I", 701 "HTTP-I" 702 ] 703 } 704 }, 705 { 706 "capability-type": "application/cdni.FCI.LI.s-ccid.v1" 707 "capability-value": { 708 "s-ccid-support": true 709 } 710 } 711 ] 712 } 714 [Ed. Note: Need to add JSON serialization discussion.] 716 8. IANA Considerations 718 This document requests the registration of the following MIME Media 719 Type under the IANA MIME Media Type registry 720 (http://www.iana.org/assignments/media-types/index.html). 722 application/cdni.FCI.RedirectionMode.v1 724 8.1. Capabilities Registry 726 IANA registries are to be used for mandatory and optional types of 727 footprint and capabilities. Therefore, the mandatory types of 728 capabilities listed in this document (see Section 5) are to be 729 registered with IANA. In order to prevent namespace collisions for 730 capabilities a new IANA registry is requested for the "CDNI 731 Capabilities" namespace. The namespace shall be split into two 732 partitions: standard and optional. 734 The "standard" namespace partition is intended to contain mandatory 735 to implement capabilities and conforms to the "IETF Review" policy as 736 defined in [RFC5226]. The registry contains the name of the standard 737 capability, the RFC number of the specification defining the 738 capability (including the capability advertisement object format and 739 serialization). 741 The following table defines the initial capabilities for the standard 742 partition: 744 +-----------------------------------------+---------+ 745 | Capability | RFC | 746 +-----------------------------------------+---------+ 747 | application/cdni.FCI.RedirectionMode.v1 | RFCthis | 748 +-----------------------------------------+---------+ 750 The "optional" namespace partition conforms to the "Expert Review" 751 policy as defined in [RFC5226]. The expert review is intended to 752 prevent namespace hoarding and to prevent the definition of redundant 753 capabilities. Vendors defining new capabilities which conflict with 754 existing capabilities follow the guidelines for the "Specification 755 Required" policy as defined in [RFC5226]. 757 8.2. Footprint Sub-Registry 759 The "CDNI Metadata Footprint Types" namespace defined in the CDNI 760 Metadata Interface document [I-D.ietf-cdni-metadata] contains the 761 supported footprint formats for use in footprint advertisement. No 762 further IANA action is required here. 764 8.3. Protocol Sub-Registry 766 The "CDNI Metadata Protocols" namespace defined in the CDNI Metadata 767 Interface document [I-D.ietf-cdni-metadata] contains the supported 768 protocol values for the Delivery Protocol and Acquisition Protocol 769 capabilities. No further IANA action is required here. 771 8.4. Redirection Mode Sub-Registry 773 The "CDNI Capabilities Redirection Modes" namespace defines the valid 774 redirection modes that may be advertised as supported by a CDN. 775 Additions to the Redirection Mode namespace conform to the "IETF 776 Review" policy as defined in [RFC5226]. 778 The following table defines the initial Redirection Modes: 780 +------------------+----------------------------------+---------+ 781 | Redirection Mode | Description | RFC | 782 +------------------+----------------------------------+---------+ 783 | DNS-I | Iterative DNS-based Redirection | RFCthis | 784 | | | | 785 | DNS-R | Recursive DNS-based Redirection | RFCthis | 786 | | | | 787 | HTTP-I | Iterative HTTP-based Redirection | RFCthis | 788 | | | | 789 | HTTP-R | Recursive HTTP-based Redirection | RFCthis | 790 +------------------+----------------------------------+---------+ 792 8.5. Logging Record Type Sub-Registry 794 The "CDNI Logging Record-Types" namespace defined in the CDNI Logging 795 Interface document [I-D.ietf-cdni-logging] contains the types of all 796 supported logging record-types. No further IANA action is required 797 here. 799 8.6. Logging Field Name Sub-Registry 801 The "CDNI Logging Fields Names" namespace defined in the CDNI Logging 802 Interface document [I-D.ietf-cdni-logging] contains the names of all 803 supported logging fields. No further IANA action is required here. 805 8.7. Metadata Type Sub-Registry 807 The "CDNI GenericMetadata Types" namespace defined in the CDNI 808 Metadata Interface document [I-D.ietf-cdni-metadata] contains the 809 names of the supported GenericMetadata objects. No further IANA 810 action is required here. 812 9. Security Considerations 814 This specification describes the semantics for capabilities and 815 footprint advertisement objects in content distribution networks. It 816 does not, however, specify a concrete protocol for transporting those 817 objects, or even a specific object syntax. Specific security 818 mechanisms can only be selected for concrete protocols that 819 instantiate these semantics. This document does, however, place some 820 high-level security constraints on such protocols. 822 All protocols that implement these semantics are REQUIRED to provide 823 integrity and authentication services. Without authentication and 824 integrity, an attacker could trivially deny service by forging a 825 footprint advertisement from a dCDN which claims the network has no 826 footprint or capability. This would prevent the uCDN from delegating 827 any requests to the dCDN. Since a pre-existing relationship between 828 all dCDNs and uCDNs is assumed by CDNi, the exchange of any necessary 829 credentials could be conducted before the FCI interface is brought 830 online. The authorization decision to accept advertisements would 831 also follow this pre-existing relationship and any contractual 832 obligations that it stipulates. 834 It is not believed that there are any serious privacy risks in 835 sharing footprint or capability information: it will represent highly 836 aggregated data about networks and at best policy-related information 837 about media, rather than any personally identifying information. 838 However, particular dCDNs may wish to share information about their 839 footprint with a uCDN but not with other, competing dCDNs. For 840 example, if a dCDN incurs an outage that reduces footprint coverage 841 temporarily, that may be information the dCDN would want to share 842 confidentially with the uCDN. Protocols implementing these semantics 843 SHOULD provide confidentiality services. 845 As specified in this document, the security requirements of the FCI 846 could be met by hop-by-hop transport-layer security mechanisms 847 coupled with domain certificates as credentials. There is no 848 apparent need for further object-level security in this framework, as 849 the trust relationships it defines are bilateral relationships 850 between uCDNs and dCDNs rather than transitive relationships. 852 10. References 854 10.1. Normative References 856 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 857 Requirement Levels", BCP 14, RFC 2119, March 1997. 859 [RFC5226] Narten, T. and H. Alvestrand, "Guidelines for Writing an 860 IANA Considerations Section in RFCs", BCP 26, RFC 5226, 861 May 2008. 863 [RFC6707] Niven-Jenkins, B., Le Faucheur, F., and N. Bitar, "Content 864 Distribution Network Interconnection (CDNI) Problem 865 Statement", RFC 6707, September 2012. 867 [RFC6770] Bertrand, G., Stephan, E., Burbridge, T., Eardley, P., Ma, 868 K., and G. Watson, "Use Cases for Content Delivery Network 869 Interconnection", RFC 6770, November 2012. 871 10.2. Informative References 873 [I-D.ietf-cdni-framework] 874 Peterson, L., Davie, B., and R. Brandenburg, "Framework 875 for CDN Interconnection", draft-ietf-cdni-framework-14 876 (work in progress), June 2014. 878 [I-D.ietf-cdni-logging] 879 Faucheur, F., Bertrand, G., Oprescu, I., and R. 880 Peterkofsky, "CDNI Logging Interface", draft-ietf-cdni- 881 logging-14 (work in progress), October 2014. 883 [I-D.ietf-cdni-metadata] 884 Niven-Jenkins, B., Murray, R., Caulfield, M., Leung, K., 885 and K. Ma, "CDN Interconnection Metadata", draft-ietf- 886 cdni-metadata-07 (work in progress), July 2014. 888 [I-D.ietf-cdni-requirements] 889 Leung, K. and Y. Lee, "Content Distribution Network 890 Interconnection (CDNI) Requirements", draft-ietf-cdni- 891 requirements-17 (work in progress), January 2014. 893 Appendix A. Acknowledgment 895 Jan Seedorf is partially supported by the GreenICN project (GreenICN: 896 Architecture and Applications of Green Information Centric 897 Networking), a research project supported jointly by the European 898 Commission under its 7th Framework Program (contract no. 608518) and 899 the National Institute of Information and Communications Technology 900 (NICT) in Japan (contract no. 167). The views and conclusions 901 contained herein are those of the authors and should not be 902 interpreted as necessarily representing the official policies or 903 endorsements, either expressed or implied, of the GreenICN project, 904 the European Commission, or NICT. 906 Martin Stiemerling provided initial input to this document and 907 valuable comments to the ongoing discussions among the authors of 908 this document. Thanks to Francois Le Faucheur and Scott Wainner for 909 providing valuable comments and suggestions to the text. 911 Authors' Addresses 913 Jan Seedorf 914 NEC 915 Kurfuerstenanlage 36 916 Heidelberg 69115 917 Germany 919 Phone: +49 6221 4342 221 920 Fax: +49 6221 4342 155 921 Email: seedorf@neclab.eu 923 Jon Peterson 924 NeuStar 925 1800 Sutter St Suite 570 926 Concord CA 94520 927 USA 929 Email: jon.peterson@neustar.biz 931 Stefano Previdi 932 Cisco Systems 933 Via Del Serafico 200 934 Rome 0144 935 Italy 937 Email: sprevidi@cisco.com 939 Ray van Brandenburg 940 TNO 941 Brassersplein 2 942 Delft 2612CT 943 The Netherlands 945 Phone: +31-88-866-7000 946 Email: ray.vanbrandenburg@tno.nl 947 Kevin J. Ma 948 Ericsson 949 43 Nagog Park 950 Acton, MA 01720 951 USA 953 Phone: +1 978-844-5100 954 Email: kevin.j.ma@ericsson.com