idnits 2.17.1 draft-ietf-i2rs-architecture-11.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 (December 16, 2015) is 3054 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Outdated reference: A later version (-11) exists of draft-ietf-i2rs-problem-statement-06 == Outdated reference: A later version (-18) exists of draft-ietf-netconf-restconf-09 == Outdated reference: A later version (-14) exists of draft-ietf-netmod-rfc6020bis-09 -- Obsolete informational reference (is this intentional?): RFC 6536 (Obsoleted by RFC 8341) -- Obsolete informational reference (is this intentional?): RFC 7223 (Obsoleted by RFC 8343) -- Obsolete informational reference (is this intentional?): RFC 7277 (Obsoleted by RFC 8344) Summary: 0 errors (**), 0 flaws (~~), 4 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group A. Atlas 3 Internet-Draft Juniper Networks 4 Intended status: Informational J. Halpern 5 Expires: June 18, 2016 Ericsson 6 S. Hares 7 Huawei 8 D. Ward 9 Cisco Systems 10 T. Nadeau 11 Brocade 12 December 16, 2015 14 An Architecture for the Interface to the Routing System 15 draft-ietf-i2rs-architecture-11 17 Abstract 19 This document describes the IETF architecture for a standard, 20 programmatic interface for state transfer in and out of the Internet 21 routing system. It describes the basic architecture, the components, 22 and their interfaces with particular focus on those to be 23 standardized as part of the Interface to Routing System (I2RS). 25 Status of This Memo 27 This Internet-Draft is submitted in full conformance with the 28 provisions of BCP 78 and BCP 79. 30 Internet-Drafts are working documents of the Internet Engineering 31 Task Force (IETF). Note that other groups may also distribute 32 working documents as Internet-Drafts. The list of current Internet- 33 Drafts is at http://datatracker.ietf.org/drafts/current/. 35 Internet-Drafts are draft documents valid for a maximum of six months 36 and may be updated, replaced, or obsoleted by other documents at any 37 time. It is inappropriate to use Internet-Drafts as reference 38 material or to cite them other than as "work in progress." 40 This Internet-Draft will expire on June 18, 2016. 42 Copyright Notice 44 Copyright (c) 2015 IETF Trust and the persons identified as the 45 document authors. All rights reserved. 47 This document is subject to BCP 78 and the IETF Trust's Legal 48 Provisions Relating to IETF Documents 49 (http://trustee.ietf.org/license-info) in effect on the date of 50 publication of this document. Please review these documents 51 carefully, as they describe your rights and restrictions with respect 52 to this document. Code Components extracted from this document must 53 include Simplified BSD License text as described in Section 4.e of 54 the Trust Legal Provisions and are provided without warranty as 55 described in the Simplified BSD License. 57 Table of Contents 59 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 60 1.1. Drivers for the I2RS Architecture . . . . . . . . . . . . 4 61 1.2. Architectural Overview . . . . . . . . . . . . . . . . . 5 62 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 9 63 3. Key Architectural Properties . . . . . . . . . . . . . . . . 11 64 3.1. Simplicity . . . . . . . . . . . . . . . . . . . . . . . 11 65 3.2. Extensibility . . . . . . . . . . . . . . . . . . . . . . 11 66 3.3. Model-Driven Programmatic Interfaces . . . . . . . . . . 12 67 4. Security Considerations . . . . . . . . . . . . . . . . . . . 13 68 4.1. Identity and Authentication . . . . . . . . . . . . . . . 14 69 4.2. Authorization . . . . . . . . . . . . . . . . . . . . . . 15 70 4.3. Client Redundancy . . . . . . . . . . . . . . . . . . . . 15 71 5. Network Applications and I2RS Client . . . . . . . . . . . . 16 72 5.1. Example Network Application: Topology Manager . . . . . . 16 73 6. I2RS Agent Role and Functionality . . . . . . . . . . . . . . 17 74 6.1. Relationship to its Routing Element . . . . . . . . . . . 17 75 6.2. I2RS State Storage . . . . . . . . . . . . . . . . . . . 17 76 6.2.1. I2RS Agent Failure . . . . . . . . . . . . . . . . . 18 77 6.2.2. Starting and Ending . . . . . . . . . . . . . . . . . 19 78 6.2.3. Reversion . . . . . . . . . . . . . . . . . . . . . . 19 79 6.3. Interactions with Local Configuration . . . . . . . . . . 19 80 6.4. Routing Components and Associated I2RS Services . . . . . 20 81 6.4.1. Routing and Label Information Bases . . . . . . . . . 21 82 6.4.2. IGPs, BGP and Multicast Protocols . . . . . . . . . . 22 83 6.4.3. MPLS . . . . . . . . . . . . . . . . . . . . . . . . 23 84 6.4.4. Policy and QoS Mechanisms . . . . . . . . . . . . . . 23 85 6.4.5. Information Modeling, Device Variation, and 86 Information Relationships . . . . . . . . . . . . . . 23 87 6.4.5.1. Managing Variation: Object Classes/Types and 88 Inheritance . . . . . . . . . . . . . . . . . . . 23 89 6.4.5.2. Managing Variation: Optionality . . . . . . . . . 24 90 6.4.5.3. Managing Variation: Templating . . . . . . . . . 24 91 6.4.5.4. Object Relationships . . . . . . . . . . . . . . 25 92 6.4.5.4.1. Initialization . . . . . . . . . . . . . . . 25 93 6.4.5.4.2. Correlation Identification . . . . . . . . . 25 94 6.4.5.4.3. Object References . . . . . . . . . . . . . . 26 95 6.4.5.4.4. Active Reference . . . . . . . . . . . . . . 26 96 7. I2RS Client Agent Interface . . . . . . . . . . . . . . . . . 26 97 7.1. One Control and Data Exchange Protocol . . . . . . . . . 26 98 7.2. Communication Channels . . . . . . . . . . . . . . . . . 26 99 7.3. Capability Negotiation . . . . . . . . . . . . . . . . . 27 100 7.4. Scope Policy Specifications . . . . . . . . . . . . . . . 27 101 7.5. Connectivity . . . . . . . . . . . . . . . . . . . . . . 28 102 7.6. Notifications . . . . . . . . . . . . . . . . . . . . . . 28 103 7.7. Information collection . . . . . . . . . . . . . . . . . 29 104 7.8. Multi-Headed Control . . . . . . . . . . . . . . . . . . 29 105 7.9. Transactions . . . . . . . . . . . . . . . . . . . . . . 30 106 8. Operational and Manageability Considerations . . . . . . . . 30 107 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 31 108 10. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 31 109 11. Informative References . . . . . . . . . . . . . . . . . . . 32 110 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 33 112 1. Introduction 114 Routers that form the internet routing infrastructure maintain state 115 at various layers of detail and function. For example, a typical 116 router maintains a Routing Information Base (RIB), and implements 117 routing protocols such as OSPF, IS-IS, and BGP to exchange 118 reachability information, topology information, protocol state, and 119 other information about the state of the network with other routers. 121 Routers convert all of this information into forwarding entries which 122 are then used to forward packets and flows between network elements. 123 The forwarding plane and the specified forwarding entries then 124 contain active state information that describes the expected and 125 observed operational behavior of the router and which is also needed 126 by the network applications. Network-oriented applications require 127 easy access to this information to learn the network topology, to 128 verify that programmed state is installed in the forwarding plane, to 129 measure the behavior of various flows, routes or forwarding entries, 130 as well as to understand the configured and active states of the 131 router. 133 This document sets out an architecture for a common, standards-based 134 interface to this information. This Interface to the Routing System 135 (I2RS) facilitates control and observation of the routing-related 136 state (for example, a Routing Element RIB manager's state), as well 137 as enabling network-oriented applications to be built on top of 138 today's routed networks. The I2RS is a programmatic asynchronous 139 interface for transferring state into and out of the internet routing 140 system. This I2RS architecture recognizes that the routing system 141 and a router's OS provide useful mechanisms that applications could 142 harness to accomplish application-level goals. These network- 143 oriented applications can leverage the I2RS programmatic interface to 144 create new ways of combining retrieval of internet routing data, 145 analyzing this data, setting state within routers. 147 Fundamental to the I2RS are clear data models that define the 148 semantics of the information that can be written and read. The I2RS 149 provides a framework for registering and for requesting the 150 appropriate information for each particular application. The I2RS 151 provides a way for applications to customize network behavior while 152 leveraging the existing routing system as desired. 154 Although the I2RS architecture is general enough to support 155 information and data models for a variety of data, and aspects of the 156 I2RS solution may be useful in domains other than routing, I2RS and 157 this document are specifically focused on an interface for routing 158 data. 160 1.1. Drivers for the I2RS Architecture 162 There are four key drivers that shape the I2RS architecture. First 163 is the need for an interface that is programmatic, asynchronous, and 164 offers fast, interactive access for atomic operations. Second is the 165 access to structured information and state that is frequently not 166 directly configurable or modeled in existing implementations or 167 configuration protocols. Third is the ability to subscribe to 168 structured, filterable event notifications from the router. Fourth, 169 the operation of I2RS is to be data-model driven to facilitate 170 extensibility and provide standard data-models to be used by network 171 applications. 173 I2RS is described as an asynchronous programmatic interface, the key 174 properties of which are described in Section 5 of 175 [I-D.ietf-i2rs-problem-statement]. 177 The I2RS architecture facilitates obtaining information from the 178 router. The I2RS architecture provides the ability to not only read 179 specific information, but also to subscribe to targeted information 180 streams, filtered events, and thresholded events. 182 Such an interface also facilitates the injection of ephemeral state 183 into the routing system. A non-routing protocol or application could 184 inject state into a routing element via the state-insertion 185 functionality of the I2RS and that state could then be distributed in 186 a routing or signaling protocol and/or be used locally (e.g. to 187 program the co-located forwarding plane). I2RS will only permit 188 modification of state that would be safe, conceptually, to modify via 189 local configuration; no direct manipulation of protocol-internal 190 dynamically determined data is envisioned. 192 1.2. Architectural Overview 194 Figure 1 shows the basic architecture for I2RS between applications 195 using I2RS, their associated I2RS Clients, and I2RS Agents. 196 Applications access I2RS services through I2RS clients. A single 197 client can provide access to one or more applications. In the 198 figure, Clients A and B each provide access to a single application 199 (application A and B respectively), while Client P provides access to 200 multiple applications. 202 Applications can access I2RS services through local or remote 203 clients. In the figure, Applications A and B access I2RS services 204 through local clients, while Applications C, D and E access I2RS 205 services through a remote client. The details of how applications 206 communicate with a remote client is out of scope for I2RS. 208 An I2RS Client can access one or more I2RS agents. In the figure 1, 209 Clients B and P access I2RS Agents 1 and 2. Likewise, an I2RS Agent 210 can provide service to one or more clients. In the figure, I2RS 211 Agent 1 provides services to Clients A, B and P while Agent 2 212 provides services to only Clients B and P. 214 I2RS agents and clients communicate with one another using an 215 asynchronous protocol. Therefore, a single client can post multiple 216 simultaneous requests, either to a single agent or to multiple 217 agents. Furthermore, an agent can process multiple requests, either 218 from a single client or from multiple clients, simultaneously. 220 The I2RS agent provides read and write access to selected data on the 221 routing element that are organized into I2RS Services. Section 4 222 describes how access is mediated by authentication and access control 223 mechanisms. Figure 1 shows I2RS agents being able to write ephemeral 224 static state (e.g. RIB entries), and to read from dynamic static 225 (e.g. MPLS LSP-ID or number of active BGP peers). 227 In addition to read and write access, the I2RS agent allows clients 228 to subscribe to different types of notifications about events 229 affecting different object instances. One example of a notification 230 of such an event (which is unrelated to an object creation, 231 modification or deletion) is when a next-hop in the RIB is resolved 232 enough to be used by a RIB manager for installation in the forwarding 233 plane as part of a particular route. Please see Section 7.6 and 234 Section 7.7 for details. 236 The scope of I2RS is to define the interactions between the I2RS 237 agent and the I2RS client and the associated proper behavior of the 238 I2RS agent and I2RS client. 240 ****************** ***************** ***************** 241 * Application C * * Application D * * Application E * 242 ****************** ***************** ***************** 243 ^ ^ ^ 244 | | | 245 |--------------| | |--------------| 246 | | | 247 v v v 248 *************** 249 * Client P * 250 *************** 251 ^ ^ 252 | |-------------------------| 253 *********************** | *********************** | 254 * Application A * | * Application B * | 255 * * | * * | 256 * +----------------+ * | * +----------------+ * | 257 * | Client A | * | * | Client B | * | 258 * +----------------+ * | * +----------------+ * | 259 ******* ^ ************* | ***** ^ ****** ^ ****** | 260 | | | | | 261 | |-------------| | | |-----| 262 | | -----------------------| | | 263 | | | | | 264 ************ v * v * v ********* ***************** v * v ******** 265 * +---------------------+ * * +---------------------+ * 266 * | Agent 1 | * * | Agent 2 | * 267 * +---------------------+ * * +---------------------+ * 268 * ^ ^ ^ ^ * * ^ ^ ^ ^ * 269 * | | | | * * | | | | * 270 * v | | v * * v | | v * 271 * +---------+ | | +--------+ * * +---------+ | | +--------+ * 272 * | Routing | | | | Local | * * | Routing | | | | Local | * 273 * | and | | | | Config | * * | and | | | | Config | * 274 * |Signaling| | | +--------+ * * |Signaling| | | +--------+ * 275 * +---------+ | | ^ * * +---------+ | | ^ * 276 * ^ | | | * * ^ | | | * 277 * | |----| | | * * | |----| | | * 278 * v | v v * * v | v v * 279 * +----------+ +------------+ * * +----------+ +------------+ * 280 * | Dynamic | | Static | * * | Dynamic | | Static | * 281 * | System | | System | * * | System | | System | * 282 * | State | | State | * * | State | | State | * 283 * +----------+ +------------+ * * +----------+ +------------+ * 284 * * * * 285 * Routing Element 1 * * Routing Element 2 * 286 ******************************** ******************************** 287 Figure 1: Architecture of I2RS clients and agents 289 Routing Element: A Routing Element implements some subset of the 290 routing system. It does not need to have a forwarding plane 291 associated with it. Examples of Routing Elements can include: 293 * A router with a forwarding plane and RIB Manager that runs IS- 294 IS, OSPF, BGP, PIM, etc., 296 * A BGP speaker acting as a Route Reflector, 298 * An LSR that implements RSVP-TE, OSPF-TE, and PCEP and has a 299 forwarding plane and associated RIB Manager, 301 * A server that runs IS-IS, OSPF, BGP and uses ForCES to control 302 a remote forwarding plane, 304 A Routing Element may be locally managed, whether via CLI, SNMP, 305 or NETCONF. 307 Routing and Signaling: This block represents that portion of the 308 Routing Element that implements part of the internet routing 309 system. It includes not merely standardized protocols (i.e. IS- 310 IS, OSPF, BGP, PIM, RSVP-TE, LDP, etc.), but also the RIB Manager 311 layer. 313 Local Configuration: A Routing Element will provide the ability to 314 configure and manage it. The Local Configuration is defined in 315 this architecture as the configuration parameters associated with 316 a routing system. The Local Configuration may be read, written, 317 or changed via a combination of CLI, NETCONF, SNMP and other 318 protocols. The black box behavior for interactions between the 319 ephemeral state that I2RS installs into the routing element and 320 the Local Configuration must be defined in I2RS data models. 322 Dynamic System State: An I2RS agent needs access to state on a 323 routing element beyond what is contained in the routing subsystem. 324 Such state may include various counters, statistics, flow data, 325 and local events. This is the subset of operational state that is 326 needed by network applications based on I2RS that is not contained 327 in the routing and signaling information. How this information is 328 provided to the I2RS agent is out of scope, but the standardized 329 information and data models for what is exposed are part of I2RS. 331 Static System State: An I2RS agent needs access to static state on 332 a routing element beyond what is contained in the routing 333 subsystem. An example of such state is specifying queueing 334 behavior for an interface or traffic. How the I2RS agent modifies 335 or obtains this information is out of scope, but the standardized 336 information and data models for what is exposed are part of I2RS. 338 I2RS Agent: See the definition in Section 2. 340 Application: A network application that needs to observe the 341 network or manipulate the network to achieve its service 342 requirements. 344 I2RS Client: See the definition in Section 2. 346 As can be seen in Figure 1, an I2RS client can communicate with 347 multiple I2RS agents. An I2RS client may connect to one or more I2RS 348 agents based upon its needs. Similarly, an I2RS agent may 349 communicate with multiple I2RS clients - whether to respond to their 350 requests, to send notifications, etc. Timely notifications are 351 critical so that several simultaneously operating applications have 352 up-to-date information on the state of the network. 354 As can also be seen in Figure 1, an I2RS Agent may communicate with 355 multiple clients. Each client may send the agent a variety of write 356 operations. In order to keep the protocol simple, two clients should 357 not attempt to write (modify) the same piece of information on an 358 I2RS Agent. This is considered an error. However, such collisions 359 may happen and section 7.8 (multi-headed control) describes how the 360 I2RS agent resolves collision by first utilizing priority to resolve 361 collisions, and second by servicing the requests in a first in, first 362 served basis. The i2rs architecture includes this definition of 363 behavior for this case simply for predictability not because this is 364 an intended result. This predictability will simplify the error 365 handling and suppress oscillations. If additional error cases beyond 366 this simple treatment are required, these error cases should be 367 resolved by the network applications and management systems. 369 In contrast, although multiple I2RS clients may need to supply data 370 into the same list (e.g. a prefix or filter list), this is not 371 considered an error and must be correctly handled. The nuances so 372 that writers do not normally collide should be handled in the 373 information models. 375 The architectural goal for the I2RS is that such errors should 376 produce predictable behaviors, and be reportable to interested 377 clients. The details of the associated policy is discussed in 378 Section 7.8. The same policy mechanism (simple priority per I2RS 379 client) applies to interactions between the I2RS agent and the 380 CLI/SNMP/NETCONF as described in Section 6.3. 382 In addition it must be noted that there may be indirect interactions 383 between write operations. A basic example of this is when two 384 different but overlapping prefixes are written with different 385 forwarding behavior. Detection and avoidance of such interactions is 386 outside the scope of the I2RS work and is left to agent design and 387 implementation. 389 2. Terminology 391 The following terminology is used in this document. 393 agent or I2RS Agent: An I2RS agent provides the supported I2RS 394 services from the local system's routing sub-systems by 395 interacting with the routing element to provide specified 396 behavior. The I2RS agent understands the I2RS protocol and can be 397 contacted by I2RS clients. 399 client or I2RS Client: A client implements the I2RS protocol, uses 400 it to communicate with I2RS Agents, and uses the I2RS services to 401 accomplish a task. It interacts with other elements of the 402 policy, provisioning, and configuration system by means outside of 403 the scope of the I2RS effort. It interacts with the I2RS agents 404 to collect information from the routing and forwarding system. 405 Based on the information and the policy oriented interactions, the 406 I2RS client may also interact with I2RS agents to modify the state 407 of their associated routing systems to achieve operational goals. 408 An I2RS client can be seen as the part of an application that uses 409 and supports I2RS and could be a software library. 411 service or I2RS Service: For the purposes of I2RS, a service refers 412 to a set of related state access functions together with the 413 policies that control their usage. The expectation is that a 414 service will be represented by a data-model. For instance, 'RIB 415 service' could be an example of a service that gives access to 416 state held in a device's RIB. 418 read scope: The read scope of an I2RS client within an I2RS agent 419 is the set of information which the I2RS client is authorized to 420 read within the I2RS agent. The read scope specifies the access 421 restrictions to both see the existence of data and read the value 422 of that data. 424 notification scope: The set of events and associated information 425 that the I2RS Client can request be pushed by the I2RS Agent. 426 I2RS Clients have the ability to register for specific events and 427 information streams, but must be constrained by the access 428 restrictions associated with their notification scope. 430 write scope: The set of field values which the I2RS client is 431 authorized to write (i.e. add, modify or delete). This access can 432 restrict what data can be modified or created, and what specific 433 value sets and ranges can be installed. 435 scope: When unspecified as either read scope, write scope, or 436 notification scope, the term scope applies to the read scope, 437 write scope, and notification scope. 439 resources: A resource is an I2RS-specific use of memory, storage, 440 or execution that a client may consume due to its I2RS operations. 441 The amount of each such resource that a client may consume in the 442 context of a particular agent may be constrained based upon the 443 client's security role. An example of such a resource could 444 include the number of notifications registered for. These are not 445 protocol-specific resources or network-specific resources. 447 role or security role: A security role specifies the scope, 448 resources, priorities, etc. that a client or agent has. Multiple 449 identities may use the same security role. A single identity may 450 have may have multiple roles. 452 identity: A client is associated with exactly one specific 453 identity. State can be attributed to a particular identity. It 454 is possible for multiple communication channels to use the same 455 identity; in that case, the assumption is that the associated 456 client is coordinating such communication. 458 Identity and scope: A single identity can be associated with 459 multiple roles. Each role has its own scope and an identity 460 associated with multiple roles can use the combined scope of all 461 its roles. More formally, each identity has: 463 a read-scope that is the logical OR of the read-scopes 464 associated with its roles, 466 a write-scope that is the logical OR of the write-scopes 467 associated with its roles, and 469 a notification-scope that is the logical OR of the 470 notification-scopes associated with its roles. 472 secondary identity: An I2RS Client may supply a secondary opaque 473 identity that is not interpreted by the I2RS Agent. An example 474 use is when the I2RS Client is a go-between for multiple 475 applications and it is necessary to track which application has 476 requested a particular operation. 478 Groups: NETCONF Network Access [RFC6536] refers uses the term group 479 in terms of an Administrative group which supports support the 480 well-established distinction between a root account and other 481 types of less-privileged conceptual user accounts. Group still 482 refers to a single identity (e.g. root) which is shared by a group 483 of users. 485 3. Key Architectural Properties 487 Several key architectural properties for the I2RS protocol are 488 elucidated below (simplicity, extensibility, and model-driven 489 programmatic interfaces). However, some architecture principles such 490 as performance and scaling are not described below because they are 491 discussed in [I-D.ietf-i2rs-problem-statement] and because the 492 performance and scaling requires varies based on the particular use- 493 cases. 495 3.1. Simplicity 497 There have been many efforts over the years to improve the access to 498 the information available to the routing and forwarding system. 499 Making such information visible and usable to network management and 500 applications has many well-understood benefits. There are two 501 related challenges in doing so. First, the quantity and diversity of 502 information potentially available is very large. Second, the 503 variation both in the structure of the data and in the kinds of 504 operations required tends to introduce protocol complexity. 506 While the types of operations contemplated here are complex in their 507 nature, it is critical that I2RS be easily deployable and robust. 508 Adding complexity beyond what is needed to satisfy well known and 509 understood requirements would hinder the ease of implementation, the 510 robustness of the protocol, and the deployability of the protocol. 511 Overly complex data models tend to ossify information sets by 512 attempting to describe and close off every possible option, 513 complicating extensibility. 515 Thus, one of the key aims for I2RS is the keep the protocol and 516 modeling architecture simple. So for each architectural component or 517 aspect, we ask ourselves "do we need this complexity, or is the 518 behavior merely nice to have?" Protocol parsimony is clearly a goal. 520 3.2. Extensibility 522 Naturally, extensibility of the protocol and data model is very 523 important. In particular, given the necessary scope limitations of 524 the initial work, it is critical that the initial design include 525 strong support for extensibility. 527 The scope of the I2RS work is being restricted in the interests of 528 achieving a deliverable and deployable result. The I2RS Working 529 Group is modeling only a subset of the data of interest. It is 530 clearly desirable for the data models defined in the I2RS to be 531 useful in more general settings. It should be easy to integrate data 532 models from the I2RS with other data. Other work should be able to 533 easily extend it to represent additional aspects of the network 534 elements or network systems. This reinforces the criticality of 535 designing the data models to be highly extensible, preferably in a 536 regular and simple fashion. 538 The I2RS Working Group is defining operations for the I2RS protocol. 539 It would be optimistic to assume that more and different ones may not 540 be needed when the scope of I2RS increases. Thus, it is important to 541 consider extensibility not only of the underlying services' data 542 models, but also of the primitives and protocol operations. 544 3.3. Model-Driven Programmatic Interfaces 546 A critical component of I2RS is the standard information and data 547 models with their associated semantics. While many components of the 548 routing system are standardized, associated data models for them are 549 not yet available. Instead, each router uses different information, 550 different mechanisms, and different CLI which makes a standard 551 interface for use by applications extremely cumbersome to develop and 552 maintain. Well-known data modeling languages exist and may be used 553 for defining the data models for I2RS. 555 There are several key benefits for I2RS in using model-driven 556 architecture and protocol(s). First, it allows for transferring 557 data-models whose content is not explicitly implemented or 558 understood. Second, tools can automate checking and manipulating 559 data; this is particularly valuable for both extensibility and for 560 the ability to easily manipulate and check proprietary data-models. 562 The different services provided by I2RS can correspond to separate 563 data-models. An I2RS agent may indicate which data-models are 564 supported. 566 The purpose of the data model is to provide an definition of the 567 information regarding the routing system that can be used in 568 operational networks. In some cases, the I2RS Working group may 569 define an information model for a set of routing information in order 570 to define in general terms the information for a data model. If 571 routing information is being modeled for the first time, a logical 572 information model may be standardized prior to creating the data 573 model. 575 4. Security Considerations 577 This I2RS architecture describes interfaces that clearly require 578 serious consideration of security. As an architecture, I2RS has been 579 designed to re-utilize existing protocols that carry network 580 management information. Two of existing protocol which the I2RS WG 581 has selected to attempt to re-use are NETCONF [RFC6241] and RESTCONF 582 [I-D.ietf-netconf-restconf]. The I2RS protocol design process is to 583 specify additional requirements which will include security for an 584 existing protocol in order to support the I2RS architecture. After 585 an existing protocol, e.g. NETCONF or RESTCONF, has been alter to 586 fit the I2RS requirements, then this protocol will be reviewed to 587 determine if it meets the I2RS security requirements. 589 Due to the re-use strategy of the I2RS architecture, this security 590 section describes the assumed security environment for I2RS with 591 additional detail on: a) identity and authentication, b) 592 authorization, and c) client redundancy. Each protocol proposed for 593 inclusions as an I2RS protocol will need to be evaluated for the 594 security constraints of the protocol. The detailed requirements for 595 the I2RS protocol and the I2RS security environment will be defined 596 within these gloal security environments. 598 First, here is a brief description of the assumed security 599 environment for I2RS. The I2RS Agent associated with a Routing 600 Element is a trusted part of that Routing Element. For example, it 601 may be part of a vendor-distributed signed software image for the 602 entire Routing Element or it may be trusted signed application that 603 an operator has installed. The I2RS Agent is assumed to have a 604 separate authentication and authorization channel by which it can 605 validate both the identity and permissions associated with an I2RS 606 Client. To support numerous and speedy interactions between the I2RS 607 Agent and I2RS Client, it is assumed that the I2RS Agent can also 608 cache that particular I2RS Clients are trusted and their associated 609 authorized scope. This implies that the permission information may 610 be old either in a pull model until the I2RS Agent re-requests it, or 611 in a push model until the authentication and authorization channel 612 can notify the I2RS Agent of changes. 614 Mutual authentication between the I2RS Client and I2RS Agent is 615 required. An I2RS Client must be able to trust that the I2RS Agent 616 is attached to the relevant Routing Element so that write/modify 617 operations are correctly applied and so that information received 618 from the I2RS Agent can be trusted by the I2RS Client. 620 An I2RS Client is not automatically trustworthy. It has identity 621 information and applications using that I2RS Client should be aware 622 of the scope limitations of that I2RS Client. If the I2RS Client is 623 acting as a broker for multiple applications, managing the security, 624 authentication and authorization for that communication is out of 625 scope; nothing prevents I2RS and a separate authentication and 626 authorization channel from being used. Regardless of mechanism, an 627 I2RS Client that is acting as a broker is responsible for determining 628 that applications using it are trusted and permitted to make the 629 particular requests. 631 Different levels of integrity, confidentiality, and replay protection 632 are relevant for different aspects of I2RS. The primary 633 communication channel that is used for client authentication and then 634 used by the client to write data requires integrity, confidentiality 635 and replay protection. Appropriate selection of a default required 636 transport protocol is the preferred way of meeting these 637 requirements. 639 Other communications via I2RS may not require integrity, 640 confidentiality, and replay protection. For instance, if an I2RS 641 Client subscribes to an information stream of prefix announcements 642 from OSPF, those may require integrity but probably not 643 confidentiality or replay protection. Similarly, an information 644 stream of interface statistics may not even require guaranteed 645 delivery. In Section 7.2, more reasoning for multiple communication 646 channels is provided. From the security perspective, it is critical 647 to realize that an I2RS Agent may open a new communication channel 648 based upon information provided by an I2RS Client (as described in 649 Section 7.2). For example, a I2RS client may request notifications 650 of certain events and the agent will open a communication channel to 651 report such events. Therefore, to avoid an indirect attack, such a 652 request must be done in the context of an authenticated and 653 authorized client whose communications cannot have been altered. 655 4.1. Identity and Authentication 657 As discussed above, all control exchanges between the I2RS client and 658 agent should be authenticated and integrity protected (such that the 659 contents cannot be changed without detection). Further, manipulation 660 of the system must be accurately attributable. In an ideal 661 architecture, even information collection and notification should be 662 protected; this may be subject to engineering tradeoffs during the 663 design. 665 I2RS clients may be operating on behalf of other applications. While 666 those applications' identities are not needed for authentication or 667 authorization, each application should have a unique opaque 668 identifier that can be provided by the I2RS client to the I2RS agent 669 for purposes of tracking attribution of operations to support 670 functionality such as troubleshooting and logging of network changes. 672 4.2. Authorization 674 All operations using I2RS, both observation and manipulation, should 675 be subject to appropriate authorization controls. Such authorization 676 is based on the identity and assigned role of the I2RS client 677 performing the operations and the I2RS agent in the network element. 678 Multiple Identities may use the same role. An identity may utilize 679 several roles. For example, an I2RS client identity may use an MPLS- 680 Monitor role to have a read-scope that includes created MPLS LSPs and 681 also use a Troubleshooting role to have write access to trigger 682 active OAM such as LSP-ping. 684 I2RS Agents, in performing information collection and manipulation, 685 will be acting on behalf of the I2RS clients. As such, each 686 operation authorization will be based on the lower of the two 687 permissions of the agent itself and of the authenticated client. The 688 mechanism by which this authorization is applied within the device is 689 outside of the scope of I2RS. 691 The appropriate or necessary level of granularity for scope can 692 depend upon the particular I2RS Service and the implementation's 693 granularity. An approach to a similar access control problem is 694 defined in the NetConf Access Control Model (NACM) [RFC6536]; it 695 allows arbitrary access to be specified for a data node instance 696 identifier while defining meaningful manipulable defaults. The 697 identity within NACM [RFC6536] can be specify as either a user name 698 or a group user name (e.g. Root), and this name is linked a scope 699 policy that contained in a set of access control rules. Similarly, 700 it is expected the I2RS identity links to one role which has a scope 701 policy specified by a set of access control rules. This scope policy 702 is can be provided via Local Configuration, exposed as an I2RS 703 Service for manipulation by authorized clients, or via some other 704 method (e.g. AAA service) 706 When an I2RS client is authenticated, its identity is provided to the 707 I2RS Agent, and this identity links to a role which links to the 708 scope policy. Multiple identities may belong to the same role; for 709 example, such a role might be an Internal-Routes-Monitor that allows 710 reading of the portion of the I2RS RIB associated with IP prefixes 711 used for internal device addresses in the AS." 713 4.3. Client Redundancy 715 I2RS must support client redundancy. At the simplest, this can be 716 handled by having a primary and a backup network application that 717 both use the same client identity and can successfully authenticate 718 as such. Since I2RS does not require a continuous transport 719 connection and supports multiple transport sessions, this can provide 720 some basic redundancy. However, it does not address the need for 721 troubleshooting and logging of network changes to be informed about 722 which network application is actually active. At a minimum, basic 723 transport information about each connection and time can be logged 724 with the identity. 726 5. Network Applications and I2RS Client 728 I2RS is expected to be used by network-oriented applications in 729 different architectures. While the interface between a network- 730 oriented application and the I2RS client is outside the scope of 731 I2RS, considering the different architectures is important to 732 sufficiently specify I2RS. 734 In the simplest architecture of direct access, a network-oriented 735 application has an I2RS client as a library or driver for 736 communication with routing elements. 738 In the broker architecture, multiple network-oriented applications 739 communicate in an unspecified fashion to a broker application that 740 contains an I2RS Client. That broker application requires additional 741 functionality for authentication and authorization of the network- 742 oriented applications; such functionality is out of scope for I2RS 743 but similar considerations to those described in Section 4.2 do 744 apply. As discussed in Section 4.1, the broker I2RS Client should 745 determine distinct opaque identifiers for each network-oriented 746 application that is using it. The broker I2RS Client can pass along 747 the appropriate value as a secondary identifier which can be used for 748 tracking attribution of operations. 750 In a third architecture, a routing element or network-oriented 751 application that uses an I2RS Client to access services on a 752 different routing element may also contain an I2RS agent to provide 753 services to other network-oriented applications. However, where the 754 needed information and data models for those services differs from 755 that of a conventional routing element, those models are, at least 756 initially, out of scope for I2RS. Below is an example of such a 757 network application 759 5.1. Example Network Application: Topology Manager 761 A Topology Manager includes an I2RS client that uses the I2RS data 762 models and protocol to collect information about the state of the 763 network by communicating directly with one or more I2RS agents. From 764 these I2RS agents, the Topology Manager collects routing 765 configuration and operational data, such as interface and label- 766 switched path (LSP) information. In addition, the Topology Manager 767 may collect link-state data in several ways - either via I2RS models, 768 by peering with BGP-LS[I-D.ietf-idr-ls-distribution] or listening 769 into the IGP. 771 The set of functionality and collected information that is the 772 Topology Manager may be embedded as a component of a larger 773 application, such as a path computation application. As a stand- 774 alone application, the Topology Manager could be useful to other 775 network applications by providing a coherent picture of the network 776 state accessible via another interface. That interface might use the 777 same I2RS protocol and could provide a topology service using 778 extensions to the I2RS data models. 780 6. I2RS Agent Role and Functionality 782 The I2RS Agent is part of a routing element. As such, it has 783 relationships with that routing element as a whole, and with various 784 components of that routing element. 786 6.1. Relationship to its Routing Element 788 A Routing Element may be implemented with a wide variety of different 789 architectures: an integrated router, a split architecture, 790 distributed architecture, etc. The architecture does not need to 791 affect the general I2RS agent behavior. 793 For scalability and generality, the I2RS agent may be responsible for 794 collecting and delivering large amounts of data from various parts of 795 the routing element. Those parts may or may not actually be part of 796 a single physical device. Thus, for scalability and robustness, it 797 is important that the architecture allow for a distributed set of 798 reporting components providing collected data from the I2RS agent 799 back to the relevant I2RS clients. There may be multiple I2RS Agents 800 within the same router. In such a case, they must have non- 801 overlapping sets of information which they manipulate. 803 To facilitate operations, deployment and troubleshooting, it is 804 important that traceability of the requests received by I2RS Agent's 805 and actions taken be supported via a common data model. 807 6.2. I2RS State Storage 809 State modification requests are sent to the I2RS agent in a routing 810 element by I2RS clients. The I2RS agent is responsible for applying 811 these changes to the system, subject to the authorization discussed 812 above. The I2RS agent will retain knowledge of the changes it has 813 applied, and the client on whose behalf it applied the changes. The 814 I2RS agent will also store active subscriptions. These sets of data 815 form the I2RS data store. This data is retained by the agent until 816 the state is removed by the client, overridden by some other 817 operation such as CLI, or the device reboots. Meaningful logging of 818 the application and removal of changes is recommended. I2RS applied 819 changes to the routing element state will not be retained across 820 routing element reboot. The I2RS data store is not preserved across 821 routing element reboots; thus the I2RS agent will not attempt to 822 reapply such changes after a reboot. 824 6.2.1. I2RS Agent Failure 826 It is expected that an I2RS Agent may fail independently of the 827 associated routing element. This could happen because I2RS is 828 disabled on the routing element or because the I2RS Agent, a separate 829 process or even running on a separate processor, experiences an 830 unexpected failure. Just as routing state learned from a failed 831 source is removed, the ephemeral I2RS state will usually be removed 832 shortly after the failure is detected or as part of a graceful 833 shutdown process. To handle I2RS Agent failure, the I2RS Agent must 834 use two different notifications. 836 NOTIFICATION_I2RS_AGENT_STARTING: This notification identifies that 837 the associated I2RS Agent has started. It includes an agent-boot- 838 count that indicates how many times the I2RS Agent has restarted 839 since the associated routing element restarted. The agent-boot- 840 count allows an I2RS Client to determine if the I2RS Agent has 841 restarted. 843 NOTIFICATION_I2RS_AGENT_TERMINATING: This notification reports that 844 the associated I2RS Agent is shutting down gracefully. Ephemeral 845 state will be removed. It can optionally include a timestamp 846 indicating when the I2RS Agent will shutdown. Use of this 847 timestamp assumes that time synchronization has been done and the 848 timestamp should not have granularity finer than one second 849 because better accuracy of shutdown time is not guaranteed. 851 There are two different failure types that are possible and each has 852 different behavior. 854 Unexpected failure: In this case, the I2RS Agent has unexpectedly 855 crashed and thus cannot notify its clients of anything. Since 856 I2RS does not require a persistent connection between the I2RS 857 Client and I2RS Agent, it is necessary to have a mechanism for the 858 I2RS Agent to notify I2RS Clients that had subscriptions or 859 written ephemeral state; such I2RS Clients should be cached by the 860 I2RS Agent's system in persistent storage. When the I2RS Agent 861 starts, it should send a NOTIFICATION_I2RS_AGENT_STARTING to each 862 cached I2RS Client. 864 Graceful failure: In this case, the I2RS Agent can do specific 865 limited work as part of the process of being disabled. The I2RS 866 Agent must send a NOTIFICATION_I2RS_AGENT_TERMINATING to all its 867 cached I2RS Clients. 869 6.2.2. Starting and Ending 871 When an I2RS client applies changes via the I2RS protocol, those 872 changes are applied and left until removed or the routing element 873 reboots. The network application may make decisions about what to 874 request via I2RS based upon a variety of conditions that imply 875 different start times and stop times. That complexity is managed by 876 the network application and is not handled by I2RS. 878 6.2.3. Reversion 880 An I2RS Agent may decide that some state should no longer be applied. 881 An I2RS Client may instruct an Agent to remove state it has applied. 882 In all such cases, the state will revert to what it would have been 883 without the I2RS client-agent interaction; that state is generally 884 whatever was specified via the CLI, NETCONF, SNMP, etc. I2RS Agents 885 will not store multiple alternative states, nor try to determine 886 which one among such a plurality it should fall back to. Thus, the 887 model followed is not like the RIB, where multiple routes are stored 888 at different preferences. (For I2RS state in the presence of two 889 I2RS clients, please see section 1.2 and section 7.8) 891 An I2RS Client may register for notifications, subject to its 892 notification scope, regarding state modification or removal by a 893 particular I2RS Client. 895 6.3. Interactions with Local Configuration 897 Changes may originate from either Local Configuration or from I2RS. 898 The modifications and data stored by I2RS are separate from the local 899 device configuration, but conflicts between the two must be resolved 900 in a deterministic manner that respects operator-applied policy. 901 That policy can determine whether Local Configuration overrides a 902 particular I2RS client's request or vice versa. To achieve this end, 903 either by default Local Configuration always wins or, optionally, a 904 routing element may permit a priority to be configured on the device 905 for the Local Configuration mechanism. The policy mechanism in the 906 later case is comparing the I2RS client's priority with that priority 907 assigned to the Local Configuration. 909 When the Local Configuration always wins, some communication between 910 that subsystem and the I2RS Agent is still necessary. That 911 communication contains the details of each specific device 912 configuration change that the I2RS Agent is permitted to modify. In 913 addition, when the system determines, that a client's I2RS state is 914 preempted, the I2RS agent must notify the affected I2RS clients; how 915 the system determines this is implementation-dependent. 917 It is critical that policy based upon the source is used because the 918 resolution cannot be time-based. Simply allowing the most recent 919 state to prevail could cause race conditions where the final state is 920 not repeatably deterministic. 922 6.4. Routing Components and Associated I2RS Services 924 For simplicity, each logical protocol or set of functionality that 925 can be compactly described in a separable information and data model 926 is considered as a separate I2RS Service. A routing element need not 927 implement all routing components described nor provide the associated 928 I2RS services. I2RS Services should include a capability model so 929 that peers can determine which parts of the service are supported. 930 Each I2RS Service requires an information model that describes at 931 least the following: data that can be read, data that can be written, 932 notifications that can be subscribed to, and the capability model 933 mentioned above. 935 The initial services included in the I2RS architecture are as 936 follows. 938 *************************** ************** ***************** 939 * I2RS Protocol * * * * Dynamic * 940 * * * Interfaces * * Data & * 941 * +--------+ +-------+ * * * * Statistics * 942 * | Client | | Agent | * ************** ***************** 943 * +--------+ +-------+ * 944 * * ************** ************* 945 *************************** * * * * 946 * Policy * * Base QoS * 947 ******************** ******** * Templates * * Templates * 948 * +--------+ * * * * * ************* 949 * BGP | BGP-LS | * * PIM * ************** 950 * +--------+ * * * 951 ******************** ******** **************************** 952 * MPLS +---------+ +-----+ * 953 ********************************** * | RSVP-TE | | LDP | * 954 * IGPs +------+ +------+ * * +---------+ +-----+ * 955 * +--------+ | OSPF | |IS-IS | * * +--------+ * 956 * | Common | +------+ +------+ * * | Common | * 957 * +--------+ * * +--------+ * 958 ********************************** **************************** 960 ************************************************************** 961 * RIB Manager * 962 * +-------------------+ +---------------+ +------------+ * 963 * | Unicast/multicast | | Policy-Based | | RIB Policy | * 964 * | RIBs & LIBs | | Routing | | Controls | * 965 * | route instances | | (ACLs, etc) | +------------+ * 966 * +-------------------+ +---------------+ * 967 ************************************************************** 969 Figure 2: Anticipated I2RS Services 971 There are relationships between different I2RS Services - whether 972 those be the need for the RIB to refer to specific interfaces, the 973 desire to refer to common complex types (e.g. links, nodes, IP 974 addresses), or the ability to refer to implementation-specific 975 functionality (e.g. pre-defined templates to be applied to interfaces 976 or for QoS behaviors that traffic is direct into). Section 6.4.5 977 discusses information modeling constructs and the range of 978 relationship types that are applicable. 980 6.4.1. Routing and Label Information Bases 982 Routing elements may maintain one or more Information Bases. 983 Examples include Routing Information Bases such as IPv4/IPv6 Unicast 984 or IPv4/IPv6 Multicast. Another such example includes the MPLS Label 985 Information Bases, per-platform or per-interface or per-context. 987 This functionality, exposed via an I2RS Service, must interact 988 smoothly with the same mechanisms that the routing element already 989 uses to handle RIB input from multiple sources, so as to safely 990 change the system state. Conceptually, this can be handled by having 991 the I2RS Agent communicate with a RIB Manager as a separate routing 992 source. 994 The point-to-multipoint state added to the RIB does not need to match 995 to well-known multicast protocol installed state. The I2RS Agent can 996 create arbitrary replication state in the RIB, subject to the 997 advertised capabilities of the routing element. 999 6.4.2. IGPs, BGP and Multicast Protocols 1001 A separate I2RS Service can expose each routing protocol on the 1002 device. Such I2RS services may include a number of different kinds 1003 of operations: 1005 o reading the various internal RIB(s) of the routing protocol is 1006 often helpful for understanding the state of the network. 1007 Directly writing to these protocol-specific RIBs or databases is 1008 out of scope for I2RS. 1010 o reading the various pieces of policy information the particular 1011 protocol instance is using to drive its operations. 1013 o writing policy information such as interface attributes that are 1014 specific to the routing protocol or BGP policy that may indirectly 1015 manipulate attributes of routes carried in BGP. 1017 o writing routes or prefixes to be advertised via the protocol. 1019 o joining/removing interfaces from the multicast trees 1021 o subscribing to an information stream of route changes 1023 o receiving notifications about peers coming up or going down 1025 For example, the interaction with OSPF might include modifying the 1026 local routing element's link metrics, announcing a locally-attached 1027 prefix, or reading some of the OSPF link-state database. However, 1028 direct modification of the link-state database must not be allowed in 1029 order to preserve network state consistency. 1031 6.4.3. MPLS 1033 I2RS Services will be needed to expose the protocols that create 1034 transport LSPs (e.g. LDP and RSVP-TE) as well as protocols (e.g. 1035 BGP, LDP) that provide MPLS-based services (e.g. pseudowires, L3VPNs, 1036 L2VPNs, etc). This should include all local information about LSPs 1037 originating in, transiting, or terminating in this Routing Element. 1039 6.4.4. Policy and QoS Mechanisms 1041 Many network elements have separate policy and QoS mechanisms, 1042 including knobs which affect local path computation and queue control 1043 capabilities. These capabilities vary widely across implementations, 1044 and I2RS cannot model the full range of information collection or 1045 manipulation of these attributes. A core set does need to be 1046 included in the I2RS information models and supported in the expected 1047 interfaces between the I2RS Agent and the network element, in order 1048 to provide basic capabilities and the hooks for future extensibility. 1050 By taking advantage of extensibility and sub-classing, information 1051 models can specify use of a basic model that can be replaced by a 1052 more detailed model. 1054 6.4.5. Information Modeling, Device Variation, and Information 1055 Relationships 1057 I2RS depends heavily on information models of the relevant aspects of 1058 the Routing Elements to be manipulated. These models drive the data 1059 models and protocol operations for I2RS. It is important that these 1060 informational models deal well with a wide variety of actual 1061 implementations of Routing Elements, as seen between different 1062 products and different vendors. There are three ways that I2RS 1063 information models can address these variations: class or type 1064 inheritance, optional features, and templating. 1066 6.4.5.1. Managing Variation: Object Classes/Types and Inheritance 1068 Information modelled by I2RS from a Routing Element can be described 1069 in terms of classes or types or object. Different valid inheritance 1070 definitions can apply. What is appropriate for I2RS to use is not 1071 determined in this architecture; for simplicity, class and subclass 1072 will be used as the example terminology. This I2RS architecture does 1073 require the ability to address variation in Routing Elements by 1074 allowing information models to define parent or base classes and 1075 subclasses. 1077 The base or parent class defines the common aspects that all Routing 1078 Elements are expected to support. Individual subclasses can 1079 represent variations and additional capabilities. When applicable, 1080 there may be several levels of refinement. The I2RS protocol can 1081 then provide mechanisms to allow an I2RS client to determine which 1082 classes a given I2RS Agent has available. Clients which only want 1083 basic capabilities can operate purely in terms of base or parent 1084 classes, while a client needing more details or features can work 1085 with the supported sub-class(es). 1087 As part of I2RS information modeling, clear rules should be specified 1088 for how the parent class and subclass can relate; for example, what 1089 changes can a subclass make to its parent? The description of such 1090 rules should be done so that it can apply across data modeling tools 1091 until the I2RS data modeling language is selected. 1093 6.4.5.2. Managing Variation: Optionality 1095 I2RS Information Models must be clear about what aspects are 1096 optional. For instance, must an instance of a class always contain a 1097 particular data field X? If so, must the client provide a value for 1098 X when creating the object or is there a well-defined default value? 1099 From the Routing Element perspective, in the above example, each 1100 Information model should provide information that: 1102 o Is X required for the data field to be accepted and applied? 1104 o If X is optional, then how does "X" as an optional portion of data 1105 field interact with the required aspects of the data field? 1107 o Does the data field have defaults for the mandatory portion of the 1108 field and the optional portions of the field 1110 o Is X required to be within a particular set of values (e.g. range, 1111 length of strings)? 1113 The information model needs to be clear about what read or write 1114 values are set by client and what responses or actions are required 1115 by the agent. It is important to indicate what is required or 1116 optional in client values and agent responses/actions. 1118 6.4.5.3. Managing Variation: Templating 1120 A template is a collection of information to address a problem; it 1121 cuts across the notions of class and object instances. A template 1122 provides a set of defined values for a set of information fields and 1123 can specify a set of values that must be provided to complete the 1124 template. Further, a flexible template scheme may that some of the 1125 defined values can be over-written. 1127 For instance, assigning traffic to a particular service class might 1128 be done by specifying a template Queueing with a parameter to 1129 indicate Gold, Silver, or Best Effort. The details of how that is 1130 carried out are not modeled. This does assume that the necessary 1131 templates are made available on the Routing Element via some 1132 mechanism other than I2RS. The idea is that by providing suitable 1133 templates for tasks that need to be accomplished, with templates 1134 implemented differently for different kinds of Routing Elements, the 1135 client can easily interact with the Routing Element without concern 1136 for the variations which are handled by values included in the 1137 template. 1139 If implementation variation can be exposed in other ways, templates 1140 may not be needed. However, templates themselves could be objects 1141 referenced in the protocol messages, with Routing Elements being 1142 configured with the proper templates to complete the operation. This 1143 is a topic for further discussion. 1145 6.4.5.4. Object Relationships 1147 Objects (in a Routing Element or otherwise) do not exist in 1148 isolation. They are related to each other. One of the important 1149 things a class definition does is represent the relationships between 1150 instances of different classes. These relationships can be very 1151 simple, or quite complicated. The following lists the information 1152 relationships that the information models need to support. 1154 6.4.5.4.1. Initialization 1156 The simplest relationship is that one object instance is initialized 1157 by copying another. For example, one may have an object instance 1158 that represents the default setup for a tunnel, and all new tunnels 1159 have fields copied from there if they are not set as part of 1160 establishment. This is closely related to the templates discussed 1161 above, but not identical. Since the relationship is only momentary 1162 it is often not formally represented in modeling, but only captured 1163 in the semantic description of the default object. 1165 6.4.5.4.2. Correlation Identification 1167 Often, it suffices to indicate in one object that it is related to a 1168 second object, without having a strong binding between the two. So 1169 an Identifier is used to represent the relationship. This can be 1170 used to allow for late binding, or a weak binding that does not even 1171 need to exist. A policy name in an object might indicate that if a 1172 policy by that name exists, it is to be applied under some 1173 circumstance. In modeling this is often represented by the type of 1174 the value. 1176 6.4.5.4.3. Object References 1178 Sometimes the relationship between objects is stronger. A valid ARP 1179 entry has to point to the active interface over which it was derived. 1180 This is the classic meaning of an object reference in programming. 1181 It can be used for relationships like containment or dependence. 1182 This is usually represented by an explicit modeling link. 1184 6.4.5.4.4. Active Reference 1186 There is an even stronger form of coupling between objects if changes 1187 in one of the two objects are always to be reflected in the state of 1188 the other. For example, if a Tunnel has an MTU (maximum transmit 1189 unit), and link MTU changes need to immediately propagate to the 1190 Tunnel MTU, then the tunnel is actively coupled to the link 1191 interface. This kind of active state coupling implies some sort of 1192 internal bookkeeping to ensure consistency, often conceptualized as a 1193 subscription model across objects. 1195 7. I2RS Client Agent Interface 1197 7.1. One Control and Data Exchange Protocol 1199 This I2RS architecture assumes a data-model driven protocol where the 1200 data-models are defined in Yang 1.1 ([RFC6020]), Yang 1.1 1201 ([I-D.ietf-netmod-rfc6020bis]), and associated Yang based model 1202 drafts ([RFC6991], [RFC7223], [RFC7224], [RFC7277], [RFC7317]). Two 1203 the protocols to be expanded to support the I2RS protocol are NETCONF 1204 [RFC6241] and RESTCONF [I-D.ietf-netconf-restconf]. This helps meet 1205 the goal of simplicity and thereby enhances deployability. The I2RS 1206 protocol may need to use several underlying transports (TCP, SCTP 1207 (stream control transport protocol), DCCP (Datagram Congestion 1208 Control Protocol)), with suitable authentication and integrity 1209 protection mechanisms. These different transports can support 1210 different types of communication (e.g. control, reading, 1211 notifications, and information collection) and different sets of 1212 data. Whatever transport is used for the data exchange, it must also 1213 support suitable congestion control mechanisms. The transports 1214 chosen should be operator and implementor friendly to ease adoption. 1216 7.2. Communication Channels 1218 Multiple communication channels and multiple types of communication 1219 channels are required. There may be a range of requirements (e.g. 1220 confidentiality, reliability), and to support the scaling there may 1221 need to be channels originating from multiple sub-components of a 1222 routing element and/or to multiple parts of an I2RS client. All such 1223 communication channels will use the same higher level protocol. Use 1224 of additional channels for communication will be coordinated between 1225 the I2RS client and the I2RS agent. 1227 I2RS protocol communication can be delivered in-band via the routing 1228 system's data plane. I2RS protocol communication might be delivered 1229 out-of-band via a management interface. Depending on what operations 1230 are requested, it is possible for the I2RS protocol communication to 1231 cause the in-band communication channels to stop working; this could 1232 cause the I2RS agent to become unreachable across that communication 1233 channel. 1235 7.3. Capability Negotiation 1237 The support for different protocol capabilities and I2RS Services 1238 will vary across I2RS Clients and Routing Elements supporting I2RS 1239 Agents. Since each I2RS Service is required to include a capability 1240 model (see Section 6.4), negotiation at the protocol level can be 1241 restricted to protocol specifics and which I2RS Services are 1242 supported. 1244 Capability negotiation (such as which transports are supported beyond 1245 the minimum required to implement) will clearly be necessary. It is 1246 important that such negotiations be kept simple and robust, as such 1247 mechanisms are often a source of difficulty in implementation and 1248 deployment. 1250 The protocol capability negotiation can be segmented into the basic 1251 version negotiation (required to ensure basic communication), and the 1252 more complex capability exchange which can take place within the base 1253 protocol mechanisms. In particular, the more complex protocol and 1254 mechanism negotiation can be addressed by defining information models 1255 for both the I2RS Agent and the I2RS Client. These information 1256 models can describe the various capability options. This can then 1257 represent and be used to communicate important information about the 1258 agent, and the capabilities thereof. 1260 7.4. Scope Policy Specifications 1262 As section 4.1 and 4.2 describe, each I2RS Client will have a unique 1263 identity and it may have a secondary identity (see section 2) to aid 1264 in troubleshooting. As section 4 indicates, all authentication and 1265 authorization mechanisms are based on the primary Identity which 1266 links to a role with scope policy for reading data, for writing data, 1267 and limitations on the resources that can be consumed. 1268 Specifications for scope policy need to specify the data and value 1269 ranges for portion of scope policy. 1271 7.5. Connectivity 1273 A client may or may not maintain an active communication channel with 1274 an agent. Therefore, an agent may need to open a communication 1275 channel to the client to communicate previously requested 1276 information. The lack of an active communication channel does not 1277 imply that the associated client is non-functional. When 1278 communication is required, the agent or client can open a new 1279 communication channel. 1281 State held by an agent that is owned by a client should not be 1282 removed or cleaned up when a client is no longer communicating - even 1283 if the agent cannot successfully open a new communication channel to 1284 the client. 1286 For many applications, it may be desirable to clean up state if a 1287 network application dies before removing the state it has created. 1288 Typically, this is dealt with in terms of network application 1289 redundancy. If stronger mechanisms are desired, mechanisms outside 1290 of I2RS may allow a supervisory network application to monitor I2RS 1291 clients, and based on policy known to the supervisor clean up state 1292 if applications die. More complex mechanism instantiated in the I2RS 1293 agent would add complications to the I2RS protocol and are thus left 1294 for future work. 1296 Some examples of such a mechanism include the following. In one 1297 option, the client could request state clean-up if a particular 1298 transport session is terminated. The second is to allow state 1299 expiration, expressed as a policy associated with the I2RS client's 1300 role. The state expiration could occur after there has been no 1301 successful communication channel to or from the I2RS client for the 1302 policy-specified duration. 1304 7.6. Notifications 1306 As with any policy system interacting with the network, the I2RS 1307 Client needs to be able to receive notifications of changes in 1308 network state. Notifications here refers to changes which are 1309 unanticipated, represent events outside the control of the systems 1310 (such as interface failures on controlled devices), or are 1311 sufficiently sparse as to be anomalous in some fashion. A 1312 notification may also be due to a regular event. 1314 Such events may be of interest to multiple I2RS Clients controlling 1315 data handled by an I2RS Agent, and to multiple other I2RS clients 1316 which are collecting information without exerting control. The 1317 architecture therefore requires that it be practical for I2RS Clients 1318 to register for a range of notifications, and for the I2RS Agents to 1319 send notifications to a number of Clients. The I2RS Client should be 1320 able to filter the specific notifications that will be received; the 1321 specific types of events and filtering operations can vary by 1322 information model and need to be specified as part of the information 1323 model. 1325 The I2RS information model needs to include representation of these 1326 events. As discussed earlier, the capability information in the 1327 model will allow I2RS clients to understand which events a given I2RS 1328 Agent is capable of generating. 1330 For performance and scaling by the I2RS client and general 1331 information confidentiality, an I2RS Client needs to be able to 1332 register for just the events it is interested in. It is also 1333 possible that I2RS might provide a stream of notifications via a 1334 publish/subscribe mechanism that is not amenable to having the I2RS 1335 agent do the filtering. 1337 7.7. Information collection 1339 One of the other important aspects of the I2RS is that it is intended 1340 to simplify collecting information about the state of network 1341 elements. This includes both getting a snapshot of a large amount of 1342 data about the current state of the network element, and subscribing 1343 to a feed of the ongoing changes to the set of data or a subset 1344 thereof. This is considered architecturally separate from 1345 notifications due to the differences in information rate and total 1346 volume. 1348 7.8. Multi-Headed Control 1350 As was described earlier, an I2RS Agent interacts with multiple I2RS 1351 Clients who are actively controlling the network element. From an 1352 architecture and design perspective, the assumption is that by means 1353 outside of this system the data to be manipulated within the network 1354 element is appropriately partitioned so that any given piece of 1355 information is only being manipulated by a single I2RS Client. 1357 Nonetheless, unexpected interactions happen and two (or more) I2RS 1358 clients may attempt to manipulate the same piece of data. This is 1359 considered an error case. This architecture does not attempt to 1360 determine what the right state of data should be when such a 1361 collision happens. Rather, the architecture mandates that there be 1362 decidable means by which I2RS Agents handle the collisions. The 1363 mechanism for ensuring predictability is to have a simple priority 1364 associated with each I2RS clients, and the highest priority change 1365 remains in effect. In the case of priority ties, the first client 1366 whose attribution is associated with the data will keep control. 1368 In order for this approach to multi-headed control to be useful for 1369 I2RS Clients, it is important that it is possible for an I2RS Client 1370 to register for changes to any changes made by I2RS to data that it 1371 may care about. This is included in the I2RS event mechanisms. This 1372 also needs to apply to changes made by CLI/NETCONF/SNMP within the 1373 write-scope of the I2RS Agent, as the same priority mechanism (even 1374 if it is "CLI always wins") applies there. The I2RS client may then 1375 respond to the situation as it sees fit. 1377 7.9. Transactions 1379 In the interest of simplicity, the I2RS architecture does not include 1380 multi-message atomicity and rollback mechanisms. Rather, it includes 1381 a small range of error handling for a set of operations included in a 1382 single message. An I2RS Client may indicate one of the following 1383 three error handling for a given message with multiple operations 1384 which it sends to an I2RS Agent: 1386 Perform all or none: This traditional SNMP semantic indicates that 1387 other I2RS agent will keep enough state when handling a single 1388 message to roll back the operations within that message. Either 1389 all the operations will succeed, or none of them will be applied 1390 and an error message will report the single failure which caused 1391 them not to be applied. This is useful when there are, for 1392 example, mutual dependencies across operations in the message. 1394 Perform until error: In this case, the operations in the message 1395 are applied in the specified order. When an error occurs, no 1396 further operations are applied, and an error is returned 1397 indicating the failure. This is useful if there are dependencies 1398 among the operations and they can be topologically sorted. 1400 Perform all storing errors: In this case, the I2RS Agent will 1401 attempt to perform all the operations in the message, and will 1402 return error indications for each one that fails. This is useful 1403 when there is no dependency across the operation, or where the 1404 client would prefer to sort out the effect of errors on its own. 1406 In the interest of robustness and clarity of protocol state, the 1407 protocol will include an explicit reply to modification or write 1408 operations even when they fully succeed. 1410 8. Operational and Manageability Considerations 1412 In order to facilitate troubleshooting of routing elements 1413 implementing I2RS agents, those routing elements should provide for a 1414 mechanism to show actively provisioned I2RS state and other I2RS 1415 Agent internal information. Note that this information may contain 1416 highly sensitive material subject to the Security Considerations of 1417 any data models implemented by that Agent and thus must be protected 1418 according to those considerations. Preferably, this mechanism should 1419 use a different privileged means other than simply connecting as an 1420 I2RS client to learn the data. Using a different mechanism should 1421 improve traceability and failure management. 1423 Manageability plays a key aspect in I2RS. Some initial examples 1424 include: 1426 Resource Limitations: Using I2RS, applications can consume 1427 resources, whether those be operations in a time-frame, entries in 1428 the RIB, stored operations to be triggered, etc. The ability to 1429 set resource limits based upon authorization is important. 1431 Configuration Interactions: The interaction of state installed via 1432 the I2RS and via a router's configuration needs to be clearly 1433 defined. As described in this architecture, a simple priority 1434 that is configured is used to provide sufficient policy 1435 flexibility. 1437 Traceability of Interactions: The ability to trace the interactions 1438 of the requests received by the I2RS Agent's and actions taken by 1439 the I2RS agents is needed so that operations can monitor I2RS 1440 Agents during deployment, and troubleshoot software or network 1441 problems. 1443 Notification Subscription Service: The ability for an I2RS Client to 1444 subscribe to a notification stream pushed from the I2RS Agent 1445 (rather than having I2RS client poll the I2RS agent) provides a 1446 more scalable notification handling for the I2RS Agent-Client 1447 interactions. 1449 9. IANA Considerations 1451 This document includes no request to IANA. 1453 10. Acknowledgements 1455 Significant portions of this draft came from draft-ward-i2rs- 1456 framework-00 and draft-atlas-i2rs-policy-framework-00. 1458 The authors would like to thank Nitin Bahadur, Shane Amante, Ed 1459 Crabbe, Ken Gray, Carlos Pignataro, Wes George, Ron Bonica, Joe 1460 Clarke, Juergen Schoenwalder, Jeff Haas, Jamal Hadi Salim, Scott 1461 Brim, Thomas Narten, Dean Bogdanovi, Tom Petch, Robert Raszuk, 1462 Sriganesh Kini, John Mattsson, Nancy Cam-Winget, DaCheng Zhang, Qin 1463 Wu, Ahmed Abro, Salman Asadullah, and Eric Yu. for their suggestions 1464 and review. 1466 11. Informative References 1468 [I-D.ietf-i2rs-problem-statement] 1469 Atlas, A., Nadeau, T., and D. Ward, "Interface to the 1470 Routing System Problem Statement", draft-ietf-i2rs- 1471 problem-statement-06 (work in progress), January 2015. 1473 [I-D.ietf-idr-ls-distribution] 1474 Gredler, H., Medved, J., Previdi, S., Farrel, A., and S. 1475 Ray, "North-Bound Distribution of Link-State and TE 1476 Information using BGP", draft-ietf-idr-ls-distribution-13 1477 (work in progress), October 2015. 1479 [I-D.ietf-netconf-restconf] 1480 Bierman, A., Bjorklund, M., and K. Watsen, "RESTCONF 1481 Protocol", draft-ietf-netconf-restconf-09 (work in 1482 progress), December 2015. 1484 [I-D.ietf-netmod-rfc6020bis] 1485 Bjorklund, M., "The YANG 1.1 Data Modeling Language", 1486 draft-ietf-netmod-rfc6020bis-09 (work in progress), 1487 December 2015. 1489 [RFC6020] Bjorklund, M., Ed., "YANG - A Data Modeling Language for 1490 the Network Configuration Protocol (NETCONF)", RFC 6020, 1491 DOI 10.17487/RFC6020, October 2010, 1492 . 1494 [RFC6241] Enns, R., Ed., Bjorklund, M., Ed., Schoenwaelder, J., Ed., 1495 and A. Bierman, Ed., "Network Configuration Protocol 1496 (NETCONF)", RFC 6241, DOI 10.17487/RFC6241, June 2011, 1497 . 1499 [RFC6536] Bierman, A. and M. Bjorklund, "Network Configuration 1500 Protocol (NETCONF) Access Control Model", RFC 6536, 1501 DOI 10.17487/RFC6536, March 2012, 1502 . 1504 [RFC6991] Schoenwaelder, J., Ed., "Common YANG Data Types", 1505 RFC 6991, DOI 10.17487/RFC6991, July 2013, 1506 . 1508 [RFC7223] Bjorklund, M., "A YANG Data Model for Interface 1509 Management", RFC 7223, DOI 10.17487/RFC7223, May 2014, 1510 . 1512 [RFC7224] Bjorklund, M., "IANA Interface Type YANG Module", 1513 RFC 7224, DOI 10.17487/RFC7224, May 2014, 1514 . 1516 [RFC7277] Bjorklund, M., "A YANG Data Model for IP Management", 1517 RFC 7277, DOI 10.17487/RFC7277, June 2014, 1518 . 1520 [RFC7317] Bierman, A. and M. Bjorklund, "A YANG Data Model for 1521 System Management", RFC 7317, DOI 10.17487/RFC7317, August 1522 2014, . 1524 Authors' Addresses 1526 Alia Atlas 1527 Juniper Networks 1528 10 Technology Park Drive 1529 Westford, MA 01886 1530 USA 1532 Email: akatlas@juniper.net 1534 Joel Halpern 1535 Ericsson 1537 Email: Joel.Halpern@ericsson.com 1539 Susan Hares 1540 Huawei 1542 Email: shares@ndzh.com 1544 Dave Ward 1545 Cisco Systems 1546 Tasman Drive 1547 San Jose, CA 95134 1548 USA 1550 Email: wardd@cisco.com 1552 Thomas D. Nadeau 1553 Brocade 1555 Email: tnadeau@lucidvision.com