idnits 2.17.1 draft-bernardos-anima-fog-monitoring-00.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 doesn't use any RFC 2119 keywords, yet has text resembling RFC 2119 boilerplate text. -- The document date (March 10, 2019) is 1845 days in the past. Is this intentional? Checking references for intended status: Experimental ---------------------------------------------------------------------------- No issues found here. Summary: 0 errors (**), 0 flaws (~~), 2 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 ANIMA WG CJ. Bernardos 3 Internet-Draft UC3M 4 Intended status: Experimental A. Mourad 5 Expires: September 11, 2019 InterDigital 6 March 10, 2019 8 Autonomic setup of fog monitoring agents 9 draft-bernardos-anima-fog-monitoring-00 11 Abstract 13 The concept of fog computing has emerged driven by the Internet of 14 Things (IoT) due to the need of handling the data generated from the 15 end-user devices. The term fog is referred to any networked 16 computational resource in the continuum between things and cloud. In 17 fog computing, functions can be stiched together composing a service 18 function chain. These functions might be hosted on resources that 19 are inherently heterogeneous, volatile and mobile. This means that 20 resources might appear and disappear, and the connectivity 21 characteristics between these resources may also change dynamically. 22 This calls for new orchestration solutions able to cope with dynamic 23 changes to the resources in runtime or ahead of time (in anticipation 24 through prediction) as opposed to today's solutions which are 25 inherently reactive and static or semi-static. 27 A fog monitoring solution can be used to help predicting events so an 28 action can be taken before an event actually takes place. This 29 solution is composed of agents running on the fog nodes plus a 30 controller hosted at another device (running in the infrastructure or 31 in another fog node). Since fog environments are inherently volatile 32 and extremely dynamic, it is convenient to enable the use of 33 autonomic technologies to autonomously set-up the fog monitoring 34 platform. This document aims at presenting this use case as well as 35 specifying how to use GRASP as needed in this scenario. 37 Status of This Memo 39 This Internet-Draft is submitted in full conformance with the 40 provisions of BCP 78 and BCP 79. 42 Internet-Drafts are working documents of the Internet Engineering 43 Task Force (IETF). Note that other groups may also distribute 44 working documents as Internet-Drafts. The list of current Internet- 45 Drafts is at https://datatracker.ietf.org/drafts/current/. 47 Internet-Drafts are draft documents valid for a maximum of six months 48 and may be updated, replaced, or obsoleted by other documents at any 49 time. It is inappropriate to use Internet-Drafts as reference 50 material or to cite them other than as "work in progress." 52 This Internet-Draft will expire on September 11, 2019. 54 Copyright Notice 56 Copyright (c) 2019 IETF Trust and the persons identified as the 57 document authors. All rights reserved. 59 This document is subject to BCP 78 and the IETF Trust's Legal 60 Provisions Relating to IETF Documents 61 (https://trustee.ietf.org/license-info) in effect on the date of 62 publication of this document. Please review these documents 63 carefully, as they describe your rights and restrictions with respect 64 to this document. Code Components extracted from this document must 65 include Simplified BSD License text as described in Section 4.e of 66 the Trust Legal Provisions and are provided without warranty as 67 described in the Simplified BSD License. 69 Table of Contents 71 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 72 1.1. Problem statement . . . . . . . . . . . . . . . . . . . . 3 73 1.2. Fog monitoring framework . . . . . . . . . . . . . . . . 4 74 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 5 75 3. Autonomic setup of fog monitoring framework . . . . . . . . . 6 76 4. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 10 77 5. Security Considerations . . . . . . . . . . . . . . . . . . . 10 78 6. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 10 79 7. References . . . . . . . . . . . . . . . . . . . . . . . . . 10 80 7.1. Normative References . . . . . . . . . . . . . . . . . . 10 81 7.2. Informative References . . . . . . . . . . . . . . . . . 10 82 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 11 84 1. Introduction 86 The concept of fog computing has emerged driven by the Internet of 87 Things (IoT) due to the need of handling the data generated from the 88 end-user devices. The term fog is referred to any networked 89 computational resource in the continuum between things and cloud. A 90 fog node may therefore be an infrastructure network node such as an 91 eNodeB or gNodeB, an edge server, a customer premises equipment 92 (CPE), or even a user equipment (UE) terminal node such as a laptop, 93 a smartphone, or a computing unit on-board a vehicle, robot or drone. 95 In fog computing, functions might be organized in service function 96 chains (SFCs), hosted on resources that are inherently heterogeneous, 97 volatile and mobile. This means that resources might appear and 98 disappear, and the connectivity characteristics between these 99 resources may also change dynamically. This calls for new 100 orchestration solutions able to cope with dynamic changes to the 101 resources in runtime or ahead of time (in anticipation through 102 prediction) as opposed to today's solutions which are inherently 103 reactive and static or semi-static. 105 1.1. Problem statement 107 Figure 1 shows an exemplary scenario of a (robot) network service. A 108 robot device has its (navigation) control application running in the 109 fog away from the robot, as a network service in the form of an SFC 110 "F1-F2" (e.g., F1 might be in charge of identifying obstacles and F2 111 takes decisions on the robot navigation). Initially the function F1 112 is assumed to be hosted at a fog node A and F2 at fog node B. At a 113 given point of time, fog node A becomes unavailable (e.g., due to low 114 battery issues or the fog node A moving away from the coverage of the 115 robot). There is therefore a need to predict the need of migrating/ 116 moving the function F1 to another node (e.g., fog node C in the 117 figure), and this needs to be done prior to the fog/edge node 118 becoming no longer capable/available. Such dynamic migration cannot 119 be dealt with in today's orchestration solutions, which are rather 120 reactive and static or semi-static (e.g., resources may fail, but 121 this is an exceptional event, happening with low frequency, and only 122 scaling actions are supported to react to SLA-related events). 124 -------------- 125 | ==== | 126 ------+F1+---------- 127 / | | ==== | | \ 128 / | +------+ | \ 129 | | fog node C | \ 130 | -------------- \ 131 | \ 132 | -------------- ---\---------- 133 | | ==== | | \==== | 134 | -----------+F1+------------+F2| | 135 |/ | | ==== | | | | ==== | | 136 o | +------+ | | +------+ | 137 | | fog node A | | fog node B | 138 --------+- -------------- -------------- 139 | | 140 --0----0-- 142 Figure 1: Example scenario 144 Existing frameworks rely on monitoring platforms that react to 145 resource failure events and ensure that negotiated SLAs are met. 146 However these are not designed to predict events likely to happen in 147 a volatile fog environment, such as resources moving away, resources 148 becoming unavailable due to battery issues or just changes in 149 availability of the resources because of variations of the use of the 150 local resources on the nodes. Besides, it is not feasible in this 151 kind of volatile and extremely mobile environment to perform a 152 continuous monitoring and reporting of every possible parameter on 153 all the nodes hosting resources, as this would not scale and would 154 consume many resources and generate extra overhead. 156 In volatile and mobile environments, prediction (make-before-break) 157 is needed, as pure reaction (break-before-make) is not enough. This 158 prediction is not generic, and depends on the nature of the network 159 service/SFC: the functions of the SFC, the connectivity between them, 160 the service-specific requirements, etc. Monitoring has to be setup 161 differently on the nodes, depending on the specifics of the network 162 service. Besides, in order to act proactively and predict what might 163 need to be done, monitoring in such a volatile and mobile 164 environments does not only involve the nodes currently hosting the 165 resources running the network service/service function chain (i.e., 166 hosting a function), but also other nodes which are potential 167 candidates to join either in addition or in substitution to current 168 nodes for running the network service in accordance with the 169 orchestration decisions. 171 In the example of Figure 1, the fog node initially hosting function 172 F1 (fog node A) might be running out of battery and this should be 173 detected before the node A actually becomes unavailable, so the 174 function F1 can be effectively migrated in a time to a different fog 175 node C, capable of meeting the requirements of F1 (compute, 176 networking, location, expected availability, etc.). In order to be 177 able to predict the need for such a migration and have already 178 identified a target fog node where to move the function, it is needed 179 to have a monitoring solution in place that instructs each node 180 involved in the service (A and B), and also neighboring node 181 candidate (C) to host function (F1), to monitor and report on metrics 182 that are relevant for the specific network service "F1-F2" that is 183 currently running. 185 1.2. Fog monitoring framework 187 Fog environments differ from data-center ones on three key aspects: 188 heterogeneity, volatility and mobility. The fog monitoring framework 189 is used to predict events triggering and orchestration event (e.g., 190 migrating a function to a different resource). 192 The monitoring framework we propose for fog environments is composed 193 of 2 logical components: 195 o Fog agents running on each fog node. An agent is responsible for 196 sending information to a fog monitoring controller and to other 197 fog agents. What to monitor and what information to send 198 (including frequency) is configured per agent considering the 199 specifics of the network service/SFC. A fog agent might also take 200 some autonomous actions (such as request migration of a function 201 to a neighbor node) in certain situations where connectivity with 202 the fog monitoring controller is temporarily unavailable. 204 o A fog monitoring controller (e.g., running at the edge or at a fog 205 node). This node obtains input from the orchestration logic (MANO 206 stack) and autonomously decides what information to monitor, where 207 and how, based on the requirements provided by the orchestration 208 logic managing the network services instantiated in the fog. This 209 configuration is network service/function specific. 211 * It interacts with the orchestration logic to coordinate and 212 trigger orchestration events, such as function migration, 213 connectivity updates, etc. In some deployments, this entity 214 might be co-located with the orchestration logic (e.g., the 215 NFVO). 217 * It interacts with the fog agents to instruct what information 218 and parameters need to be monitored, as well as to obtain such 219 information. This interaction is not limited to fog agents at 220 nodes currently involved in a given network service/SFC, but 221 also includes other nodes that are suitable for hosting a 222 function that needs to be migrated. This allows to provide the 223 orchestration logic with candidate nodes in a pro-active way. 225 * It is capable of autonomously discover and set up fog agents. 227 2. Terminology 229 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 230 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", 231 and"OPTIONAL" in this document are to be interpreted as described in 232 BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all 233 capitals, as shown here. 235 The following terms are using in ths document: 237 fog: Fog goes to the Extreme Edge, that is the closest 238 possible to the user including on the user device 239 itself. 241 fog node: Any device that is capable of participating in the Fog. 242 A Fog node might be volatile, mobile and constrained 243 (in terms of computing resources). Fog nodes may be 244 heterogeneous and may belong to different owners. 246 orchestrator: In this document we use orchestrator and NFVO terms 247 interchangeably. 249 3. Autonomic setup of fog monitoring framework 251 Fog nodes autonomously start fog agents at the bootstrapping, then 252 start looking for other agents and the fog monitoring controller. 253 This autonomic setup can be performed using GRASP. The procedure is 254 represented in Figure 2. The different steps are described next: 256 +--------+ +--------+ +--------+ 257 | fog | | fog | | fog | 258 | node C | | node A | | node B | +------+ 259 | | | | | | | fog | 260 | | | | | | | | | | | | +------+ | mon. | 261 | +----+ | | +----+ | | +----+ | | NFVO | | ctrl | 262 +--------+ +--------+ +--------+ +------+ +------+ 263 | | | | 264 (fog nodes A & B bootstrap) | | 265 | | | | 266 | | periodic mcast advertisement| 267 | | (ID, fog_scope) | 268 | | <----------------------------+ 269 | Mcast discovery (fog_node_ID, scope) | 270 +-------------------------------------------->| 271 +------------>| | | 272 | Mcast discovery (fog_node_ID, scope) | 273 | +------------------------------>| 274 |<------------+ | | 275 | | | | 276 | Unicast advertisement (ID, fog_scope) | 277 | |<------------------------------+ 278 |<--------------------------------------------+ 279 | | | | 280 | Unicast registration (ID, fog_node_ID | 281 | | fog_scope, capab.) | 282 | +------------------------------>| 283 +-------------------------------------------->| 284 | | | | 285 (fog nodes A & B registered) | | 286 | | | | 287 (fog node C bootstraps) | | | 288 | | | | | 289 | Mcast discovery (fog_node_ID, scope) | | 290 +---------------------------------------------------------->| 291 +-------------------------->| | | 292 +------------>| Unicast advertisement (ID, fog_scope) | 293 |<----------------------------------------------------------+ 294 |<--------------------------+ | | 295 |<------------+ Unicast registration (ID, fog_node_ID | 296 | | | fog_scope, capab.) | 297 +---------------------------------------------------------->| 298 (fog node C registered) | | | 299 | | | | | 301 Figure 2: Autonomic setup of fog agents 303 o The fog monitoring controller is regularly sending periodic 304 multicast advertisement messages, which include its ID as well as 305 the scope for the advertisement messages (i.e., the scope of where 306 the messages have to be flooded). 308 M_DISCOVERY messages are used, with new objectives and objective 309 options. GRASP specifies that "an objective option is used to 310 identify objectives for the purposes of discovery, negotiation or 311 synchronization". New objective options are defined for the 312 purposes of discovering potential fog agents with certain 313 characteristics. Non-limiting examples of these options are 314 listed below (note that the names are just examples, and the ones 315 used have to be registered by the IANA): 317 * FOGNODERADIO: used to specify a given type of radio technology, 318 e.g.,: WiFi (version), D2D, LTE, 5G, Bluetooth (version), etc. 320 * FOGNODECONNECTIVITY: used to specify a given type of 321 connectivity, e.g., layer-2, IPv4, IPv6. 323 * FOGNODEVIRTUALIZATION: used to specify a given type of 324 virtualization supported by the node where the agent runs. 325 Examples are: hypervisor (type), container, micro-kernel, bare- 326 metal, etc. 328 * FOGNODEDOMAIN: used to specify the domain/owner of the node. 329 This is useful to support operation of multiple domains/ 330 operators simultaneously on the same fog network. 332 An example of discovery message using GRASP would be the following 333 (in this example, the fog monitoring controller is identified by 334 its IPv6 address: 2001:DB8:1111:2222:3333:4444:5555:6666): 336 [M_DISCOVERY, 13948745, h'20010db8111122223333444455556666', 337 ["FOGDOMAIN", F_SYNCH_bits, 2, "operator1"]] 339 GRASP is used to allow the fog agents and the controller discovery 340 in an autonomic way. The extensions defined above, together with 341 the use of properly scoped multicast addresses (as explained 342 below), allow to precisely define which nodes participate in the 343 monitoring and to gather their principal characteristics. 345 o When a fog node bootstraps, such as nodes A and B in the figure, 346 they start sending multicast discovery messages within a given 347 scope, that is, the intended area that composes the fog. The 348 definition of the scope depends on the scenario, and examples of 349 possible scopes are: 351 * All-resources of a given manufacturer. 353 * All-resources of a given type. 355 * All-resources of a given administrative domain. 357 * All-resources of a given user. 359 * All-resources within a topological network distance (e.g., 360 number of hops). 362 * All-resources within a geographical location. 364 * Etc. 366 Combination of previous scopes are also possible. 368 The discovery messages are multicast within the scope, reaching 369 all the nodes that compose the specified fog resources. This can 370 be done for example using well defined IPv6 multicast addresses, 371 specified for each of the different scopes. This signaling is 372 based on GRASP. Different IPv6 multicast addresses need to be 373 defined to reach each different scope, using scopes equal or 374 larger than Admin-Local according to [RFC7346]. 376 o In response to multicast fog discovery messages, the fog 377 monitoring controller replies with unicast information messages. 379 o Fog agents can then register with a controller. The registration 380 message is unicast, and includes information on the capabilities 381 of the fog node, such as: 383 * Type of node. 385 * Vendor. 387 * Energy source: battery-powered or not. 389 * Connectivity (number of network interfaces and information 390 associated to them, such as radio technology type, layer-2 and 391 layer-3 addresses, etc.). 393 * Etc. 395 Note that registration to multiple fog monitoring controller 396 instances could also be possible if a fog node wants to belong to 397 several fog domains at the same time (but note that how the 398 orchestration of the same resource is done by multiple 399 orchestrators is not covered by this invention). The defined 400 mechanisms support this via the use of fog IDs and FOGNODEDOMAIN 401 options. 403 o A fog node C bootstraps after nodes A and B are already 404 registered. The same discovery process is followed by fog node C, 405 but in addition to the regular advertisement, registration 406 procedures described before, existing neighboring fog agents (such 407 as A and B in this example), might also respond to discovery 408 messages sent by bootstrapping nodes to provide required 409 information. This makes the procedure faster, more efficient and 410 reliable. In addition to helping the fog monitoring controller in 411 the fog agent discovery process, fog agents learn themselves about 412 the existence and associated capabilities of other fog agents. 413 This can be used to allow autonomous monitoring by the fog agents 414 without the involvement of the central controller. 416 4. IANA Considerations 418 TBD. 420 5. Security Considerations 422 TBD. 424 6. Acknowledgments 426 The work in this draft will be further developed and explored under 427 the framework of the H2020 5G-CORAL project (Grant 761586). 429 7. References 431 7.1. Normative References 433 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 434 Requirement Levels", BCP 14, RFC 2119, 435 DOI 10.17487/RFC2119, March 1997, 436 . 438 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 439 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 440 May 2017, . 442 7.2. Informative References 444 [RFC7346] Droms, R., "IPv6 Multicast Address Scopes", RFC 7346, 445 DOI 10.17487/RFC7346, August 2014, 446 . 448 Authors' Addresses 450 Carlos J. Bernardos 451 Universidad Carlos III de Madrid 452 Av. Universidad, 30 453 Leganes, Madrid 28911 454 Spain 456 Phone: +34 91624 6236 457 Email: cjbc@it.uc3m.es 458 URI: http://www.it.uc3m.es/cjbc/ 460 Alain Mourad 461 InterDigital Europe 463 Email: Alain.Mourad@InterDigital.com 464 URI: http://www.InterDigital.com/