idnits 2.17.1 draft-ashw-mpls-te-feed-01.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** Looks like you're using RFC 2026 boilerplate. This must be updated to follow RFC 3978/3979, as updated by RFC 4748. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- == No 'Intended status' indicated for this document; assuming Proposed Standard Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) ** The document seems to lack separate sections for Informative/Normative References. All references will be assumed normative when checking for downward references. ** There are 2 instances of too long lines in the document, the longest one being 2 characters in excess of 72. ** The document seems to lack a both a reference to RFC 2119 and the recommended RFC 2119 boilerplate, even if it appears to use RFC 2119 keywords. RFC 2119 keyword, line 393: '... Implementations MUST NOT permit bandw...' RFC 2119 keyword, line 415: '...ected errors. An implementation SHOULD...' Miscellaneous warnings: ---------------------------------------------------------------------------- -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (January 2000) is 8839 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Missing Reference: 'OSPF' is mentioned on line 59, but not defined == Unused Reference: 'CR-LDP' is defined on line 457, but no explicit reference was found in the text == Unused Reference: 'LDP' is defined on line 460, but no explicit reference was found in the text == Outdated reference: A later version (-06) exists of draft-ietf-mpls-cr-ldp-04 == Outdated reference: A later version (-11) exists of draft-ietf-mpls-ldp-05 == Outdated reference: A later version (-05) exists of draft-ietf-isis-traffic-01 ** Downref: Normative reference to an Informational draft: draft-ietf-isis-traffic (ref. 'IS-IS') Summary: 6 errors (**), 0 flaws (~~), 7 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 MPLS Working Group Peter Ashwood-Smith 3 Internet Draft Bilel Jamoussi 4 Expiration Date: August 2000 Don Fedyk 5 Darek Skalecki 6 Nortel Networks 8 January 2000 10 IMPROVING TOPOLOGY DATA BASE ACCURACY WITH LSP FEEDBACK VIA CR-LDP 12 draft-ashw-mpls-te-feed-01.txt 14 Status of this Memo 16 This document is an Internet-Draft and is in full conformance with 17 all provisions of Section 10 of RFC2026. 19 Internet-Drafts are working documents of the Internet Engineering 20 Task Force (IETF), its areas, and its working groups. Note that 21 other groups may also distribute working documents as Internet- 22 Drafts. 24 Internet-Drafts are draft documents valid for a maximum of six 25 months and may be updated, replaced, or obsoleted by other documents 26 at any time. It is inappropriate to use Internet-Drafts as reference 27 material or to cite them other than as "work in progress." 29 The list of current Internet-Drafts can be accessed at 30 http://www.ietf.org/ietf/1id-abstracts.txt 32 The list of Internet-Draft Shadow Directories can be accessed at 33 http://www.ietf.org/shadow.html. 35 Abstract 37 One key component of traffic engineering is a concept known as 38 constraint based routing. In constraint based routing a topology 39 database is maintained on all participating nodes. This database 40 contains a complete list of all the links in the network that 41 participate in traffic engineering and for each of these links a set 42 of constraints which those links can meet. Bandwidth, for example, 43 is one essential constraint. Since the bandwidth available changes 44 as new LSPs are established and terminated the topology database 45 will develop inconsistencies with respect to the real network. It is 46 not possible to increase the flooding rates arbitrarily to keep the 47 database discrepancies from growing. We propose a new mechanism 48 whereby a source node can learn about the successes or failures of 49 its path selections by receiving feedback from the paths it is 50 attempting. This fed-back information can be incorporated into 51 subsequent route computations, which greatly improves the accuracy 52 of the overall routing solution by significantly reducing the 53 database discrepancies. 55 1. Introduction 57 Because the network is a distributed system, it is necessary to have 58 a mechanism to advertise information about links to all nodes in the 59 network [IS-IS], [OSPF]. A node can then build a topology map of 60 the network. This information is required to be as up-to-date as 61 possible for accurate traffic engineered paths. Information about 62 link or node failures must be rapidly propagated through the network 63 so that recovery can be initiated. Other information about links 64 that may be useful for reasons of quality of service include 65 parameters such as available bandwidth, and delay. The information 66 in this topology database is often out of date with respect to the 67 real network. Available bandwidth is the most critical of these 68 attributes and it can drift substantially with respect to reality 69 due to the low frequency of link state updates that can be sustained 70 in a very large topology. We refer to the deviation in the topology 71 database available bandwidth as being optimistic if the database 72 shows more available bandwidth than there really is, or pessimistic 73 if the topology database shows less bandwidth than there really is. 74 This distinction is important because we shall propose an efficient 75 algorithm to deal with optimistic databases without resorting to 76 shorter flooding intervals. 78 One of the major problems for a constraint based routing system is 79 dealing with changing constraints. Obviously, since bandwidth is one 80 of the essential constraints, dealing with the rapid changes in 81 reserved bandwidth poses some interesting challenges. In smaller 82 networks, one can resort to higher frequency flooding but this 83 obviously does not scale. 85 The basic proposal is to add to the signaling protocol the ability 86 to piggyback actual link bandwidth availability information at every 87 link that the signaling traverses. This is done as part of the 88 reverse messaging on success or failure (mapping, release, withdraw 89 or notification). What this means is that every time signaling 90 messages flow backwards toward a source to tell it of the success, 91 failure or termination of a request, that message contains a 92 detailed slice of bandwidth availability information for the exact 93 path that the message has followed. This slice of reservation 94 information, which is very up to date, is received by the source 95 node and inserted into the source node's topology database prior to 96 making any further source route computations. The result is that the 97 source node's topology database will tend to stay synchronized with 98 the slices of the network through which it is establishing paths. 99 This is nothing more than learning from successes and failures and 100 represents an intelligent alternative to either waiting for floods 101 or introducing non-determinism (guessing) into the source 102 algorithms. 104 Operating a constraint based routing system without such feedback is 105 inefficient at best since a source node will continue to give out 106 incorrect route over and over again until it gets an IGP update. 107 This could be minutes away and as a result the worst case blocking 108 time for a new route is the minimum repeatable flooding interval 109 (often several minutes in big networks). Alternatives to feedback 110 mechanisms involve adding some non-determinism (randomness) to the 111 routing algorithm in the hopes that it will stumble onto a path that 112 works. These sorts of approaches are seen in ATM dynamic routing 113 systems, which do not have these forms of feedback. 115 In order to get a good understanding of how the feedback works, 116 imagine a network with precisely one path (with sufficient 117 unreserved bandwidth) available from the source to the destination. 118 Further, imagine that the topology database at the source is 119 significantly out of date with respect to the real network in that 120 the source topology database sees sufficient bandwidth available on 121 many different routes to the destination. We call this being 122 optimistic with respect to the network since the source thinks that 123 more bandwidth is available than there really is. 125 When such an optimistic source selects its first path it will likely 126 contain links that do not in reality have sufficient unreserved 127 bandwidth. Therefore, the path is only established up to the link 128 that does not have sufficient bandwidth. A notification message is 129 formatted that contains the actual unreserved bandwidth for this 130 blocking link which flows back toward the source, collapsing the 131 partially created path as it goes. In addition, at every link that 132 this notification traverses, the current unreserved bandwidth 133 information for each corresponding link is appended to the vector of 134 unreserved bandwidth along the path. In this manner, an accurate 135 view of the slice through the network we are traversing is 136 constructed. Eventually this message arrives back at the source 137 node, where the vector is taken and inserted into the topology 138 database. This node has just learned from its mistake and is now 139 slightly less optimistic with respect to the real network 140 conditions. 142 Path selection can be attempted again but this time the node will 143 not make the same mistake it made the previous time. The link in 144 question, at which rejection occurred the first time, will not even 145 be eligible this time around, so a source route computation is 146 guaranteed to produce a different path (or none). The same procedure 147 may be repeated as many times as is necessary, each time learning 148 from its mistakes, until eventually no paths remain in the source 149 topology to the destination, or we actually find a path that works. 150 This tendency to converge either to a solution or determine that 151 there is no solution is an important property of a routing system 152 (it actually behaves a lot like a depth first search). This property 153 is not present with flooding mechanisms alone since the source node 154 must randomly hunt, or continually make the same mistakes, or abort 155 until the next flood arrives. 157 In addition to feeding back bandwidth on failure, we also recommend 158 feedback on success. This has important consequences on our ability 159 to spread load or to spill over to new links as existing links fill. 160 It is true that spilling over to new links does not require feedback 161 on success since we could simply wait for a feedback on failure, but 162 we can achieve better load spreading earlier. 164 Finally, when a path is torn down the release/withdraw messages also 165 contain bandwidth information that can be fed back into the source 166 topology database. This is very important during failure scenarios 167 where the links we need to use to reroute the path share common sub- 168 segments with the failed path. Without the feedback, the common sub- 169 segments may not indicate sufficient available bandwidth until we 170 get a flood that may mean many seconds without a connection. With 171 feedback at least we will be up to date with respect to available 172 bandwidth up to the point of failure in the path. Also since failure 173 involves many paths tearing down and re-establishing this is the 174 time that it is most critical to have an accurate view. 176 When preemption is being employed it is also extremely important 177 that the topology database inconsistencies be small. If not, high 178 setup priority LSPs may unnecessarily preempt lower holding priority 179 LSPs to obtain bandwidth that, had they had a more up to date view 180 of unreserved bandwidth, they would have been able to find 181 elsewhere. Since preempted LSPs may in turn preempt other LSPs in a 182 domino like effect, the results of such database inconsistencies can 183 have wide reaching ripple like impacts. These feedback mechanisms 184 help reduce these occurrences significantly. 186 There are a number of network conditions where feedback shows its 187 value. One can think of a constraint-based network as being in one 188 of three conditions. The first is called ramp-up, this is when the 189 rate of arriving reservations exceeds the rate of departing 190 reservations. The second is called steady-state, this is when the 191 rate of arriving reservations is about the same as the rate of 192 departing reservations. Finally, the ramp-down condition is that 193 which has a greater rate of departing reservations than arriving 194 reservations. 196 These three network conditions show distinctly different types of 197 error in the topology databases. In particular an optimistic view of 198 available bandwidth by a source node is characteristic of the ramp- 199 up condition of a network. A pessimistic view of available bandwidth 200 by a source node is characteristic of the ramp-down condition of a 201 network. If one plots the average error in the topology databases 202 with respect to the real network for the three different network 203 conditions, one will see the error slowly go positive during ramp 204 up, slowly go negative during ramp down, and drift slowly around 0 205 for the steady condition. The effect of flooding on this plot is to 206 periodically snap the error back to 0 at flooding intervals. The 207 effect of the feedback algorithm is to bring an optimistic error 208 back to zero without having to wait for the flood interval. On 209 average then, the feedback algorithm tends to halve the absolute 210 error, keeping it mostly negative or pessimistic. This makes sense 211 since a routing system will never give paths to links that it thinks 212 do not have resources and as a result its pessimistic view of the 213 world stays that way until it gets a flood. This relieves the IGP 214 updates of the most urgent requirement of flooding when bandwidth is 215 consumed. Availability of new bandwidth occurs when paths are 216 released or new links become available. New links are accompanied 217 by floods. Significant releases of bandwidth can be broadcast at 218 relatively low frequencies in the order of several minutes with 219 little operational impact. 221 Extensive operational experience with this feedback protocol in 222 proprietary Nortel Networks (pre-standard CR-LDP) products has shown 223 it to work very well for networks up to 1000 nodes with significant 224 flooding intervals damped to several minutes. Without this protocol, 225 these networks would block setups for up to several minutes. With 226 this protocol, the blocking in most cases is reduced to a small 227 number of retry attempts which is usually sub-second depending 228 mostly on the propagation delays in the network. 230 These feedback algorithms have been particularly beneficial in cases 231 of failure recovery during which the network is in a sudden 232 condition of ramp-up. Since a large number of reservations must be 233 remade, it is highly likely that we will exceed the limits of 234 certain key links in the network. Without feedback, the rerouting 235 must block until a flood arrives telling us of the situation at 236 those key links at which time rerouting can continue. With feedback, 237 the rerouting simply continues until a feedback indicates that a 238 link is full. In addition since reservation-balancing algorithms are 239 also often used, feedback allows the balancing algorithms to make 240 better distribution decisions based on immediate feedback. 242 We have also explored through simulation and implementation a 243 variety of mechanisms to deal with the pessimistic error in the 244 database. One simple proposal is to use selective forgetting. In 245 this algorithm, a reserved bandwidth value slowly drops back to zero 246 over a relatively short time interval. The theory being that you 247 shift the network back to an optimistic state (by forgetting your 248 pessimism) where the feedback algorithm will again correctly 249 operate. These algorithms have not shown any great advantage and are 250 actually non-optimal when the error is purely optimistic. 252 Other algorithmic permutations we have explored include such 253 variations as: 255 Feeding-back to all intermediate nodes, information learned from 256 control messages upstream of that intermediate node. 258 Feeding back in both directions so that both the source and 259 destination node's databases stay synchronized. 261 Allowing a request to continue to its destination despite there 262 being insufficient bandwidth at some intermediate hop. Then, 263 rejecting the request with a full bandwidth vector slice all the way 264 to the destination instead of just to the point of rejection. 266 Our simulations have not show significant benefits relative to the 267 simpler algorithm proposed here. However, it is an interesting 268 research topic to explore and quantify the different feedback 269 algorithms and their impacts on blocking times so we do not want to 270 discourage the interested reader from exploring these concepts more 271 fully. 273 2. Adding feedback TLVs to CR-LDP 275 Two new TLVs are optionally added to the CR-LDP mapping, 276 notification, and withdraw messages. There may be an arbitrary 277 number of these TLV in any order or position in the message. It is 278 recommended that they be placed such that they can be read and 279 applied to a topology database by scanning the message forwards and 280 walking the topology database from the point where the last link 281 feedback TLV left off. 283 Each TLV consists of the 8 unreserved bandwidth values for each 284 holding priority 0 through 7 as IEEE floating point numbers (the 285 units are unidirectional bytes per second). Following this are the 286 IP addresses of the two ends of the interface. Two TLVs are 287 possible, one for IPV4 and one for IPV6 addressing of the link. 289 2.1 Bandwidth directionality considerations 291 The order of the two addresses in the feedback TLV implies the 292 direction in which the bandwidth is available. For example if the 293 first address is A and the second address is B the bandwidth is 294 unreserved in the A to B direction. 296 It is possible for an implementation to provide both the A to B 297 direction and the B to A direction as part of the same feedback 298 message. This is done by simply including a TLV with A,B as the 299 addresses of the link and a different TLV with B,A as the addresses 300 of the link. Should CR-LDP evolve to be able to support bi- 301 directional traffic flow and reservations it is expected that bi- 302 directional feedback would also be implemented via this mechanism. 304 3. IPV4 specified link feedback TLV 306 0 1 2 3 307 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 308 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 309 |U|F| 0x830 | Length | 310 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 311 | BANDWIDTH UNRESERVED AT HOLDING PRIORITY 0 (IEEE float) | 312 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 313 . . . . . . . . 314 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 315 | BANDWIDTH UNRESERVED AT HOLDING PRIORITY 7 (IEEE float) | 316 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 317 | IPV4 address of interface (near end) | 318 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 319 | IPV4 address of interface (far end) | 320 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 322 4. IPV6 specified link feedback TLV 324 0 1 2 3 325 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 326 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 327 |U|F| 0x831 | Length | 328 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 329 | BANDWIDTH UNRESERVED AT HOLDING PRIORITY 0 (IEEE float) | 330 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 331 . . . . . . . . 332 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 333 | BANDWIDTH UNRESERVED AT HOLDING PRIORITY 7 (IEEE float) | 334 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 335 | IPV6 address of interface (near end) | 336 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 337 | IPV6 address of interface (far end) | 338 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 340 5. Detailed Procedures 342 On receipt of a withdraw, notification, or mapping message 343 pertaining to a request made by CR-LDP (as opposed to LDP), a 344 feedback TLV of the appropriate format for the interface over which 345 the message was received is inserted into the message before 346 forwarding it back to the source of the request. The 8 bandwidth 347 values are filled in with the outgoing bandwidth available on this 348 interface for each of the 8 holding priorities in bytes per second. 349 Finally the interface's address and far end address are placed in 350 the TLV. 352 On receipt of a CR-LDP request message which cannot be satisfied. A 353 notification message is formatted normally. The 8-bandwidth values 354 are filled in with the outgoing bandwidth available on this 355 interface for each of the 8 holding priorities in bytes per second. 356 Finally, the interface's address and far end address are placed in 357 the TLV. 359 On receipt of a CR-LDP request message which has been satisfied and 360 which results in a mapping being generated. No feedback TLV is added 361 since the previous node will insert the proper TLV when it receives 362 the reverse flowing mapping. 364 When an LDP session goes down either because of a link failure, 365 TCP/IP timeout, keepalive timeout, adjacency timeout etc. Other LDP 366 sessions in the module must generate either notification, withdraw 367 or release messages for LSPs that traversed the LDP in question. In 368 the case that the LSP was created by CR-LDP and that a withdraw or 369 notification is about to be generated, LDP will insert a feedback 370 TLV for the interface which just went down that contains 0's for all 371 the bandwidth values and attach to it the proper interface 372 addresses. 374 When the LDP session that originated a CR-LDP label request receives 375 a mapping that contains feedback TLV's it is recommended that these 376 bandwidth values overwrite the corresponding values in the node's 377 topology database. Doing so permits this node to immediately 378 synchronize its topology with respect to the real bandwidth 379 reservations along the path that was just established. 381 When the LDP session that originated a CR-LDP label request receives 382 a notification that contains feedback TLV's it is recommended that 383 these bandwidth values overwrite the corresponding values in the 384 nodes topology database. Doing so permits this node to immediately 385 synchronize its topology with respect to the real bandwidth 386 reservations along the path that just failed to establish. The 387 source node may then re-compute a path knowing that the computation 388 will take into account the failure if it was caused by the topology 389 database being in error with respect to the real network state. 391 6. IGP considerations 393 Implementations MUST NOT permit bandwidth information learned by 394 this feedback mechanism to be re-flooded via IS-IS, OSPF or any 395 other IGP. The bandwidth information learned via these feedback 396 mechanisms is to be used ONLY for source route computations on the 397 nodes that are directly on the path that fed back the bandwidth. 398 Normally only the source node of the LSP, or perhaps intermediate 399 gateway nodes will use this information. It is however permitted for 400 intermediate nodes that are forwarding this feedback information to 401 store it for their own local source route computations. 403 There is a possibility of a race condition between the bandwidth 404 information that is received via feedback and that which is received 405 via a normal IGP flood. While there may be a discrepancy between the 406 two, both are within a few 100 milliseconds of being correct. 407 Solutions to allow us to determine which information is most up to 408 date (say by adding a sequence number) do not add any significant 409 benefit. Constraint based, source routed systems will always have 410 errors in the local topology database with respect to the real 411 network. We can reduce these errors through reduced flooding 412 intervals, path following feedback and selective flooding but we 413 cannot realistically reduce the errors below the second or so range. 414 As a result propagation delay order race conditions are noise with 415 respect to the average expected errors. An implementation SHOULD 416 therefore consider the most recently received update (IGP or 417 feedback) as being the most up to date. 419 7. Future considerations 421 Constraint based routing systems such as CR-LDP will in the future 422 offer other forms of constraint than simply reserved bandwidth. 423 Actual utilization levels, current congestion levels, number of 424 discrete channels/wavelengths available etc. are all possible 425 constraints that change rapidly and which must be taken into 426 consideration when computing a route. It is expected that this 427 mechanism will be used to feedback these and other new forms of link 428 constraining data. 430 8. RSVP consideration 432 Nothing precludes the use of such feedback mechanisms with a similar 433 TLV structure in the RSVP Resv and other reverse flowing message 434 although repeatedly applying a fed-back update into a local topology 435 database is wasteful and probably should be damped. 437 9. Intellectual Property Consideration 439 The IETF has been notified of intellectual property rights claimed 440 in regard to some or all of the specification contained in this 441 document. For more information consult the online list of claimed 442 rights. 444 10. Security Considerations 446 This document raises no new security considerations for CR-LDP, RSVP 447 or MPLS in general. 449 11. Acknowledgments 451 The authors would like to thank Keith Dysart for his guidance, and 452 Jerzy Miernik for helping implement these concepts and bringing them 453 to life. 455 12. References 457 [CR-LDP] Constraint-Based LSP Setup using LDP, draft-ietf-mpls-cr- 458 ldp-04.txt 460 [LDP] LDP Specification, draft-ietf-mpls-ldp-05.txt 462 [IS-IS] Extensions to IS-IS for traffic engineering, draft-ietf- 463 isis-traffic-01.txt 465 13. Author's Addresses 467 Peter Ashwood-Smith Bilel Jamoussi 468 Nortel Networks Corp. Nortel Networks Corp. 469 P.O. Box 3511 Station C, 600 Technology Park Drive 470 Ottawa, ON K1Y 4H7 Billerica, MA 01821 471 Canada USA 472 Phone: +1 613-763-4534 phone: +1 978-288-4506 473 petera@nortelnetworks.com jamoussi@nortelnetworks.com 475 Darek Skalecki Don Fedyk 476 Nortel Networks Corp. Nortel Networks Corp. 477 P.O. Box 3511 Station C, 600 Technology Park Drive 478 Ottawa, On K1Y 4H7 Billerica, MA 01821 479 Canada USA 480 Phone: +1 613-765-2252 Phone: +1 978-228-3041 481 dareks@nortelnetworks.com dwfedyk@nortelnetworks.com 483 Full Copyright Statement 485 Copyright (C) The Internet Society (date). All Rights Reserved. This 486 document and translations of it may be copied and furnished to 487 others, and derivative works that comment on or otherwise explain it 488 or assist in its implementation may be prepared, copied, published 489 and distributed, in whole or in part, without restriction of any 490 kind, provided that the above copyright notice and this paragraph 491 are included on all such copies and derivative works. However, this 492 document itself may not be modified in any way, such as by removing 493 the copyright notice or references to the Internet Society or other 494 Internet organizations, except as needed for the purpose of 495 developing Internet standards in which case the procedures for 496 copyrights defined in the Internet Standards process must be 497 followed, or as required to translate it into languages other than 498 English. 500 The limited permissions granted above are perpetual and will not be 501 revoked by the Internet Society or its successors or assigns.