idnits 2.17.1 draft-ietf-ospf-te-metric-extensions-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 : ---------------------------------------------------------------------------- ** The abstract seems to contain references ([RFC3630]), which it shouldn't. Please replace those with straight textual mentions of the documents in question. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (November 1, 2011) is 4531 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: 'RFC2119' is mentioned on line 158, but not defined == Missing Reference: 'RFC4203' is mentioned on line 168, but not defined == Missing Reference: 'RFC4206' is mentioned on line 399, but not defined == Missing Reference: 'RFC5329' is mentioned on line 498, but not defined == Unused Reference: 'RFC2328' is defined on line 518, but no explicit reference was found in the text == Unused Reference: 'RFC3031' is defined on line 520, but no explicit reference was found in the text Summary: 1 error (**), 0 flaws (~~), 7 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 1 Network Working Group S. Giacalone 2 Internet Draft Thomson Reuters 3 Intended status: Proposed Standard 4 Expires: May 2012 D. Ward 5 Juniper Networks 7 J. Drake 8 Juniper Networks 10 A. Atlas 11 Juniper Networks 13 S. Previdi 14 Cisco Systems 16 November 1, 2011 18 OSPF Traffic Engineering (TE) Metric Extensions 19 draft-ietf-ospf-te-metric-extensions-00.txt 21 Abstract 23 In certain networks, such as, but not limited to, financial 24 information networks (e.g. stock market data providers), network 25 performance criteria (e.g. latency) are becoming as critical to data 26 path selection as other metrics. 28 This document describes extensions to OSPF TE [RFC3630] such that 29 network performance information can be distributed and collected in a 30 scalable fashion. The information distributed using OSPF TE Express 31 Path can then be used to make path selection decisions based on 32 network performance. 34 Note that this document only covers the mechanisms with which network 35 performance information is distributed. The mechanisms for measuring 36 network performance or acting on that information, once distributed, 37 are outside the scope of this document. 39 Status of this Memo 41 This Internet-Draft is submitted in full conformance with the 42 provisions of BCP 78 and BCP 79. 44 Internet-Drafts are working documents of the Internet Engineering 45 Task Force (IETF), its areas, and its working groups. Note that 46 other groups may also distribute working documents as Internet- 47 Drafts. 49 Internet-Drafts are draft documents valid for a maximum of six months 50 and may be updated, replaced, or obsoleted by other documents at any 51 time. It is inappropriate to use Internet-Drafts as reference 52 material or to cite them other than as "work in progress." 54 The list of current Internet-Drafts can be accessed at 55 http://www.ietf.org/ietf/1id-abstracts.txt 57 The list of Internet-Draft Shadow Directories can be accessed at 58 http://www.ietf.org/shadow.html 60 This Internet-Draft will expire on May 1, 2012. 62 Copyright Notice 64 Copyright (c) 2011 IETF Trust and the persons identified as the 65 document authors. All rights reserved. 67 This document is subject to BCP 78 and the IETF Trust's Legal 68 Provisions Relating to IETF Documents 69 (http://trustee.ietf.org/license-info) in effect on the date of 70 publication of this document. Please review these documents 71 carefully, as they describe your rights and restrictions with respect 72 to this document. Code Components extracted from this document must 73 include Simplified BSD License text as described in Section 4.e of 74 the Trust Legal Provisions and are provided without warranty as 75 described in the Simplified BSD License. 77 Table of Contents 79 1. Introduction...................................................3 80 2. Conventions used in this document..............................4 81 3. Express Path Extensions to OSPF TE.............................4 82 4. Sub TLV Details................................................6 83 4.1. Unidirectional Link Delay Sub-TLV.........................6 84 4.1.1. Type.................................................6 85 4.1.2. Length...............................................6 86 4.1.3. A bit................................................7 87 4.1.4. Reserved.............................................7 88 4.1.5. Delay Value..........................................7 89 4.2. Unidirectional Delay Variation Sub-TLV....................7 90 4.2.1. Type.................................................7 91 4.2.2. Length...............................................7 92 4.2.3. Reserved.............................................8 93 4.2.4. Delay Variation......................................8 94 4.3. Unidirectional Link Loss Sub-TLV..........................8 95 4.3.1. Type.................................................8 96 4.3.2. Length...............................................8 97 4.3.3. A bit................................................8 98 4.3.4. Reserved.............................................9 99 4.3.5. Link Loss............................................9 100 4.4. Unidirectional Residual Bandwidth Sub-TLV.................9 101 4.4.1. Type.................................................9 102 4.4.2. Length..............................................10 103 4.4.3. Residual Bandwidth..................................10 104 4.5. Unidirectional Available Bandwidth Sub-TLV...............10 105 4.4.4. Type................................................10 106 4.4.5. Length..............................................11 107 4.4.6. Available Bandwidth.................................11 108 5. Announcement Thresholds and Filters...........................11 109 6. Announcement Suppression......................................11 110 7. Network Stability and Announcement Periodicity................12 111 8. Compatibility.................................................12 112 9. Security Considerations.......................................12 113 10. IANA Considerations..........................................12 114 11. References...................................................12 115 11.1. Normative References....................................12 116 11.2. Informative References..................................13 117 12. Acknowledgments..............................................13 118 13. Author's Addresses...........................................14 120 1. Introduction 122 In certain networks, such as, but not limited to, financial 123 information networks (e.g. stock market data providers), network 124 performance information (e.g. latency) is becoming as critical to 125 data path selection as other metrics. 127 In these networks, extremely large amounts of money rest on the 128 ability to access market data in "real time" and to predictably make 129 trades faster than the competition. Because of this, using metrics 130 such as hop count or cost as routing metrics is becoming only 131 tangentially important. Rather, it would be beneficial to be able to 132 make path selection decisions based on performance data (such as 133 latency) in a cost-effective and scalable way. 135 This document describes extensions to OSPF TE (hereafter called "OSPF 136 TE Express Path"), that can be used to distribute network performance 137 information (such as link delay, delay variation, packet loss, 138 residual bandwidth, and available bandwidth). 140 The data distributed by OSPF TE OSPF TE Express Path is meant to be 141 used as part of the operation of the routing protocol (e.g. by 142 replacing cost with latency or considering bandwidth as well as 143 cost), by enhancing CSPF, or for other uses such as supplementing the 144 data used by an Alto server [Alto]. With respect to CSPF, the data 145 distributed by OSPF TE Express Path can be used to setup, fail over, 146 and fail back data paths using protocols such as RSVP-TE [RFC3209]. 148 Note that the mechanisms described in this document only disseminate 149 performance information. The methods for initially gathering that 150 performance information, such as [Frost], or acting on it once it is 151 distributed are outside the scope of this document. 153 2. Conventions used in this document 155 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 156 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 157 document are to be interpreted as described in RFC-2119 [RFC2119]. 159 In this document, these words will appear with that interpretation 160 only when in ALL CAPS. Lower case uses of these words are not to be 161 interpreted as carrying RFC-2119 significance. 163 3. Express Path Extensions to OSPF TE 165 This document proposes new OSPF TE sub-TLVs that can be announced in 166 OSPF TE LSAs to distribute network performance information. The 167 extensions in this document build on the ones provided in OSPF TE 168 [RFC3630] and GMPLS [RFC4203]. 170 OSPF TE LSAs [RFC3630] are opaque LSAs [RFC5250] with area flooding 171 scope. Each TLV has one or more nested sub-TLVs which permit the TE 172 LSA to be readily extended. There are two main types of OSPF TE LSA; 173 the Router Address or Link TE LSA. Like the extensions in GMPLS 174 (RFC4203), this document proposes several additional sub-TLVs for 175 the Link TE LSA: 177 Type Length Value 179 TBD1 4 Unidirectional Link Delay 181 TBD2 4 Unidirectional Delay Variation 183 TBD3 4 Unidirectional Packet Loss 185 TBD4 4 Unidirectional Residual Bandwidth Sub TLV 187 TBD5 4 Unidirectional Available Bandwidth Sub TLV 189 As can be seen in the list above, the sub-TLVs described in this 190 document carry different types of network performance information. 191 Many (but not all) of the sub-TLVs include a bit called the Anomalous 192 (or "A") bit. When the A bit is clear (or when the sub-TLV does not 193 include an A bit), the sub-TLV describes steady state link 194 performance. This information could conceivably be used to construct 195 a steady state performance topology for initial tunnel path 196 computation, or to verify alternative failover paths. 198 When network performance violates configurable link-local thresholds 199 a sub-TLV with the A bit set is advertised. These sub-TLVs could be 200 used by the receiving node to determine whether to fail traffic to a 201 backup path, or whether to calculate an entirely new path. From an 202 MPLS perspective, the intent of the A bit is to permit LSP ingress 203 nodes to: 205 A) Determine whether the link referenced in the sub-TLV affects any 206 of the LSPs for which it is ingress. If there are, then: 208 B) Determine whether those LSPs still meet end-to-end performance 209 objectives. If not, then: 211 C) The node could then conceivably move affected traffic to a pre- 212 established protection LSP or establish a new LSP and place the 213 traffic in it. 215 If link performance then improves beyond a configurable minimum 216 value (reuse threshold), that sub-TLV can be re-advertised with the 217 Anomalous bit cleared. In this case, a receiving node can 218 conceivably do whatever re-optimization (or failback) it wishes to 219 do (including nothing). 221 Note that when a sub-TLV does not include the A bit, that sub-TLV 222 cannot be used for failover purposes. The A bit was intentionally 223 omitted from some sub-TLVs to help mitigate oscillations. See section 224 7. 1. for more information. 226 Consistent with existing OSPF TE specifications (RFC3630), the 227 bandwidth advertisements defined in this draft MUST be encoded as 228 IEEE floating point values. The delay and delay variation 229 advertisements defined in this draft MUST be encoded as integer 230 values. Delay values MUST be quantified in units of microseconds, 231 packet loss MUST be quantified as a percentage of packets sent, and 232 bandwidth MUST be sent as bytes per second. All values (except 233 residual bandwidth) MUST be calculated as rolling averages where the 234 averaging period MUST be a configurable period of time. See section 235 5. for more information. 237 4. Sub TLV Details 239 4.1. Unidirectional Link Delay Sub-TLV 241 This sub-TLV advertises the average link delay between two directly 242 connected OSPF neighbors. The delay advertised by this sub-TLV MUST 243 be the delay from the local neighbor to the remote one (i.e. the 244 forward path latency). The format of this sub-TLV is shown in the 245 following diagram: 247 0 1 2 3 248 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 1 249 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 250 | TBD1 | 4 | 251 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 252 |A| RESERVED | Delay | 253 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 255 4.1.1. Type 257 This sub-TLV has a type of TBD1. 259 4.1.2. Length 261 The length is 4. 263 4.1.3. A bit 265 This field represents the Anomalous (A) bit. The A bit is set when 266 the measured value of this parameter exceeds its configured maximum 267 threshold. The A bit is cleared when the measured value falls below 268 its configured reuse threshold. If the A bit is clear, the sub-TLV 269 represents steady state link performance. 271 4.1.4. Reserved 273 This field is reserved for future use. It MUST be set to 0 when sent 274 and MUST be ignored when received. 276 4.1.5. Delay Value 278 This 24-bit field carries the average link delay over a configurable 279 interval in micro-seconds, encoded as an integer value. When set to 280 0, it has not been measured. When set to the maximum value 16,777,215 281 (16.777215 sec), then the delay is at least that value and may be 282 larger. 284 4.2. Unidirectional Delay Variation Sub-TLV 286 This sub-TLV advertises the average link delay variation between two 287 directly connected OSPF neighbors. The delay variation advertised by 288 this sub-TLV MUST be the delay from the local neighbor to the remote 289 one (i.e. the forward path latency). The format of this sub-TLV is 290 shown in the following diagram: 292 0 1 2 3 293 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 1 294 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 295 | TBD2 | 4 | 296 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 297 | RESERVED | Delay Variation | 298 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 300 4.2.1. Type 302 This sub-TLV has a type of TBD2. 304 4.2.2. Length 306 The length is 4. 308 4.2.3. Reserved 310 This field is reserved for future use. It MUST be set to 0 when sent 311 and MUST be ignored when received. 313 4.2.4. Delay Variation 315 This 24-bit field carries the average link delay variation over a 316 configurable interval in micro-seconds, encoded as an integer value. 317 When set to 0, it has not been measured. When set to the maximum 318 value 16,777,215 (16.777215 sec), then the delay is at least that 319 value and may be larger. 321 4.3. Unidirectional Link Loss Sub-TLV 323 This sub-TLV advertises the loss (as a packet percentage) between two 324 directly connected OSPF neighbors. The link loss advertised by this 325 sub-TLV MUST be the packet loss from the local neighbor to the remote 326 one (i.e. the forward path loss). The format of this sub-TLV is shown 327 in the following diagram: 329 0 1 2 3 330 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 1 331 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 332 | TBD3 | 4 | 333 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 334 |A| RESERVED | Link Loss | 335 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 337 4.3.1. Type 339 This sub-TLV has a type of TBD3 341 4.3.2. Length 343 The length is 4 345 4.3.3. A bit 347 This field represents the Anomalous (A) bit. The A bit is set when 348 the measured value of this parameter exceeds its configured maximum 349 threshold. The A bit is cleared when the measured value falls below 350 its configured reuse threshold. If the A bit is clear, the sub-TLV 351 represents steady state link performance. 353 4.3.4. Reserved 355 This field is reserved for future use. It MUST be set to 0 when sent 356 and MUST be ignored when received. 358 4.3.5. Link Loss 360 This 24-bit field carries link packet loss as a percentage of the 361 total traffic sent over a configurable interval. The basic unit is 362 0.000003%, where (2^24 - 2) is 50.331642%. This value is the highest 363 packet loss percentage that can be expressed (the assumption being 364 that precision is more important on high speed links than the ability 365 to advertise loss rates greater than this, and that high speed links 366 with over 50% loss are unusable). Therefore, measured values that are 367 larger than the field maximum SHOULD be encoded as the maximum value. 368 When set to a value of all 1s (2^24 - 1), the link packet loss has 369 not been measured. 371 4.4. Unidirectional Residual Bandwidth Sub-TLV 373 This TLV advertises the residual bandwidth (defined in section 4.4.3. 374 between two directly connected OSPF neighbors. The residual bandwidth 375 advertised by this sub-TLV MUST be the residual bandwidth from the 376 system originating the LSA to its neighbor. 378 The format of this sub-TLV is shown in the following diagram: 380 0 1 2 3 381 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 1 382 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 383 | TBD4 | 4 | 384 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 385 | Residual Bandwidth | 386 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 388 4.4.1. Type 390 This sub-TLV has a type of TBD4. 392 4.4.2. Length 394 The length is 4. 396 4.4.3. Residual Bandwidth 398 This field carries the residual bandwidth on a link, forwarding 399 adjacency [RFC4206], or bundled link in IEEE floating point format 400 with units of bytes per second. For a link or forwarding adjacency, 401 residual bandwidth is defined to be Maximum Bandwidth [RFC3630] minus 402 the bandwidth currently allocated to RSVP-TE LSPs. For a bundled 403 link, residual bandwidth is defined to be the sum of the component 404 link residual bandwidths. 406 Note that although it may seem possible to calculate Residual 407 Bandwidth using the existing sub-TLVs in RFC 3630, this is not a 408 consistently reliable approach and hence the Residual Bandwidth sub- 409 TLV has been added here. For example, because the Maximum Reservable 410 Bandwidth [RFC3630] can be larger than the capacity of the link, 411 using it as part of an algorithm to determine the value of the 412 Maximum Bandwidth [RFC3630] minus the bandwidth currently allocated 413 to RSVP-TE LSPs cannot be considered reliably accurate. 415 4.5. Unidirectional Available Bandwidth Sub-TLV 417 This TLV advertises the available bandwidth (defined in section 418 4.4.6. ) between two directly connected OSPF neighbors. The available 419 bandwidth advertised by this sub-TLV MUST be the available bandwidth 420 from the system originating the LSA to its neighbor. The format of 421 this sub-TLV is shown in the following diagram: 423 0 1 2 3 424 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 1 425 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 426 | TBD5 | 4 | 427 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 428 | Available Bandwidth | 429 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 431 4.4.4. Type 433 This sub-TLV has a type of TBD5. 435 4.4.5. Length 437 The length is 4. 439 4.4.6. Available Bandwidth 441 This field carries the available bandwidth on a link, forwarding 442 adjacency, or bundled link in IEEE floating point format with units 443 of bytes per second. For a link or forwarding adjacency, available 444 bandwidth is defined to be residual bandwidth (see section 4.4. ) 445 minus the measured bandwidth used for the actual forwarding of non- 446 RSVP-TE LSP packets. For a bundled link, available bandwidth is 447 defined to be the sum of the component link available bandwidths. 449 5. Announcement Thresholds and Filters 451 The values advertised in all sub-TLVs MUST be controlled using an 452 exponential filter (i.e. a rolling average) with a configurable 453 measurement interval and filter coefficient. 455 Implementations are expected to provide separately configurable 456 advertisement thresholds. All thresholds MUST be configurable on a 457 per sub-TLV basis. 459 The announcement of all sub-TLVs that do not include the A bit SHOULD 460 be controlled by variation thresholds that govern when they are sent. 462 Sub-TLV that include the A bit are governed by several thresholds. 463 Firstly, a threshold SHOULD be implemented to govern the announcement 464 of sub-TLVs that advertise a change in performance, but not an SLA 465 violation (i.e. when the A bit is not set). Secondly, implementations 466 MUST provide configurable thresholds that govern the announcement of 467 sub-TLVs with the A bit set (for the indication of a performance 468 violation). Thirdly, implementations SHOULD provide reuse 469 thresholds. These thresholds govern sub-TLV re-announcement with the 470 A bit cleared to permit fail back. 472 6. Announcement Suppression 474 When link performance average values change, but fall under the 475 threshold that would cause the announcement of a sub-TLV with the A 476 bit set, implementations MAY suppress or throttle sub-TLV 477 announcements. All suppression features and thresholds SHOULD be 478 configurable. 480 7. Network Stability and Announcement Periodicity 482 To mitigate concerns about stability, all values (except residual 483 bandwidth) MUST be calculated as rolling averages where the averaging 484 period MUST be a configurable period of time, rather than 485 instantaneous measurements. 487 Announcements MUST also be able to be throttled using configurable 488 inter-update throttle timers. The minimum announcement periodicity is 489 1 announcement per second. 491 8. Compatibility 493 As per (RFC3630), unrecognized TLVs should be silently ignored 495 9. Security Considerations 497 This document does not introduce security issues beyond those 498 discussed in [RFC3630] and [RFC5329]. 500 10. IANA Considerations 502 IANA maintains the registry for the sub-TLVs. OSPF TE Express Path 503 will require one new type code per sub-TLV defined in this document. 505 11. References 507 11.1. Normative References 509 [RFC2119]Bradner, S., "Key words for use in RFCs to Indicate 510 Requirement Levels", BCP 14, RFC 2119, March 1997. 512 [RFC3630] Katz, D., Kompella, K., Yeung, D., "Traffic 513 Engineering (TE) Extensions to OSPF Version 2", RFC 3630, 514 September 2003. 516 11.2. Informative References 518 [RFC2328] Moy, J, "OSPF Version 2", RFC 2328, April 1998 520 [RFC3031] Rosen, E., Viswanathan, A., Callon, R., "Multiprotocol 521 Label Switching Architecture", January 2001 523 [RFC3209] Awduche, D., Berger, L., Gan, D., Li, T., Srinivasan, 524 V., and G. Swallow, "RSVP-TE: Extensions to RSVP for LSP 525 Tunnels", RFC 3209, December 2001. 527 [RFC5250] Berger, L., Bryskin I., Zinin, A., Coltun, R., "The OSPF 528 Opaque LSA Option", RFC 5250, July 2008. 530 [Frost] D. Frost, S. Bryant"A Packet Loss and Delay Measurement 531 Profile for MPLS-based Transport Networks" 533 [Alto] R. Alimi R. Penno Y. Yang, "ALTO Protocol" 535 12. Acknowledgments 537 The authors would like to recognize Ayman Soliman for his 538 contributions. 540 This document was prepared using 2-Word-v2.0.template.dot. 542 13. Author's Addresses 544 Spencer Giacalone 545 Thomson Reuters 546 195 Broadway 547 New York NY 10007, USA 549 Email: Spencer.giacalone@thomsonreuters.com 551 Dave Ward 552 Juniper Networks 553 1194 N. Mathilda Ave. 554 Sunnyvale, CA 94089, USA 556 Email: dward@juniper.net 558 John Drake 559 Juniper Networks 560 1194 N. Mathilda Ave. 561 Sunnyvale, CA 94089, USA 563 Email: jdrake@juniper.net 565 Alia Atlas 566 Juniper Networks 567 1194 N. Mathilda Ave. 568 Sunnyvale, CA 94089, USA 570 Email: akatlas@juniper.net 572 Stefano Previdi 573 Cisco Systems 574 Via Del Serafico 200 575 00142 Rome 576 Italy 578 Email: sprevidi@cisco.com