idnits 2.17.1 draft-ietf-mpls-entropy-label-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 draft header indicates that this document updates RFC3031, but the abstract doesn't seem to mention this, which it should. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year (Using the creation date from RFC3031, updated by this document, for RFC5378 checks: 1998-03-17) -- 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 (May 5, 2011) is 4712 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: 'TL0' is mentioned on line 930, but not defined == Missing Reference: 'E' is mentioned on line 930, but not defined == Missing Reference: 'AL' is mentioned on line 752, but not defined == Missing Reference: 'TL1' is mentioned on line 753, but not defined == Missing Reference: '-' is mentioned on line 797, but not defined == Missing Reference: 'VL' is mentioned on line 852, but not defined -- Looks like a reference, but probably isn't: '0' on line 932 == Missing Reference: 'SL1' is mentioned on line 795, but not defined == Missing Reference: 'HL' is mentioned on line 796, but not defined == Missing Reference: 'SL2' is mentioned on line 797, but not defined == Missing Reference: 'CL' is mentioned on line 853, but not defined == Missing Reference: 'PL1' is mentioned on line 931, but not defined == Missing Reference: 'PL2' is mentioned on line 932, but not defined ** Obsolete normative reference: RFC 3107 (Obsoleted by RFC 8277) == Outdated reference: A later version (-07) exists of draft-ietf-pwe3-fat-pw-05 -- Obsolete informational reference (is this intentional?): RFC 4379 (Obsoleted by RFC 8029) -- Obsolete informational reference (is this intentional?): RFC 4447 (Obsoleted by RFC 8077) Summary: 1 error (**), 0 flaws (~~), 14 warnings (==), 6 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group K. Kompella 3 Internet-Draft J. Drake 4 Updates: 3031 (if approved) Juniper Networks 5 Intended status: Standards Track S. Amante 6 Expires: November 6, 2011 Level 3 Communications, LLC 7 W. Henderickx 8 Alcatel-Lucent 9 L. Yong 10 Huawei USA 11 May 5, 2011 13 The Use of Entropy Labels in MPLS Forwarding 14 draft-ietf-mpls-entropy-label-00 16 Abstract 18 Load balancing is a powerful tool for engineering traffic across a 19 network. This memo suggests ways of improving load balancing across 20 MPLS networks using the concept of "entropy labels". It defines the 21 concept, describes why entropy labels are useful, enumerates 22 properties of entropy labels that allow maximal benefit, and shows 23 how they can be signaled and used for various applications. 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 November 6, 2011. 42 Copyright Notice 44 Copyright (c) 2011 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. Conventions used . . . . . . . . . . . . . . . . . . . . . 4 61 1.2. Motivation . . . . . . . . . . . . . . . . . . . . . . . . 5 62 2. Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . 6 63 3. Entropy Labels . . . . . . . . . . . . . . . . . . . . . . . . 7 64 4. Data Plane Processing of Entropy Labels . . . . . . . . . . . 8 65 4.1. Ingress LSR . . . . . . . . . . . . . . . . . . . . . . . 8 66 4.2. Transit LSR . . . . . . . . . . . . . . . . . . . . . . . 9 67 4.3. Egress LSR . . . . . . . . . . . . . . . . . . . . . . . . 9 68 5. Signaling for Entropy Labels . . . . . . . . . . . . . . . . . 9 69 5.1. LDP Signaling . . . . . . . . . . . . . . . . . . . . . . 10 70 5.2. BGP Signaling . . . . . . . . . . . . . . . . . . . . . . 11 71 5.3. RSVP-TE Signaling . . . . . . . . . . . . . . . . . . . . 12 72 6. Operations, Administration, and Maintenance (OAM) and 73 Entropy Labels . . . . . . . . . . . . . . . . . . . . . . . . 13 74 7. MPLS-TP and Entropy Labels . . . . . . . . . . . . . . . . . . 14 75 8. Point-to-Multipoint LSPs and Entropy Labels . . . . . . . . . 15 76 9. Entropy Labels and Applications . . . . . . . . . . . . . . . 15 77 9.1. Tunnels . . . . . . . . . . . . . . . . . . . . . . . . . 15 78 9.2. LDP Pseudowires . . . . . . . . . . . . . . . . . . . . . 17 79 9.3. BGP Applications . . . . . . . . . . . . . . . . . . . . . 18 80 9.3.1. Inter-AS BGP VPNs . . . . . . . . . . . . . . . . . . 19 81 9.4. Multiple Applications . . . . . . . . . . . . . . . . . . 20 82 10. Security Considerations . . . . . . . . . . . . . . . . . . . 21 83 11. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 22 84 11.1. LDP Entropy Label TLV . . . . . . . . . . . . . . . . . . 22 85 11.2. BGP Entropy Label Attribute . . . . . . . . . . . . . . . 22 86 11.3. Attribute Flags for LSP_Attributes Object . . . . . . . . 22 87 11.4. Attributes TLV for LSP_Attributes Object . . . . . . . . . 22 88 12. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 23 89 13. References . . . . . . . . . . . . . . . . . . . . . . . . . . 23 90 13.1. Normative References . . . . . . . . . . . . . . . . . . . 23 91 13.2. Informative References . . . . . . . . . . . . . . . . . . 23 92 Appendix A. Applicability of LDP Entropy Label sub-TLV . . . . . 24 93 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 25 95 1. Introduction 97 Load balancing, or multi-pathing, is an attempt to balance traffic 98 across a network by allowing the traffic to use multiple paths. Load 99 balancing has several benefits: it eases capacity planning; it can 100 help absorb traffic surges by spreading them across multiple paths; 101 it allows better resilience by offering alternate paths in the event 102 of a link or node failure. 104 As providers scale their networks, they use several techniques to 105 achieve greater bandwidth between nodes. Two widely used techniques 106 are: Link Aggregation Group (LAG) and Equal-Cost Multi-Path (ECMP). 107 LAG is used to bond together several physical circuits between two 108 adjacent nodes so they appear to higher-layer protocols as a single, 109 higher bandwidth 'virtual' pipe. ECMP is used between two nodes 110 separated by one or more hops, to allow load balancing over several 111 shortest paths in the network. This is typically obtained by 112 arranging IGP metrics such that there are several equal cost paths 113 between source-destination pairs. Both of these techniques may, and 114 often do, co-exist in various parts of a given provider's network, 115 depending on various choices made by the provider. 117 A very important requirement when load balancing is that packets 118 belonging to a given 'flow' must be mapped to the same path, i.e., 119 the same exact sequence of links across the network. This is to 120 avoid jitter, latency and re-ordering issues for the flow. What 121 constitutes a flow varies considerably. A common example of a flow 122 is a TCP session. Other examples are an L2TP session corresponding 123 to a given broadband user, or traffic within an ATM virtual circuit. 125 To meet this requirement, a node uses certain fields, termed 'keys', 126 within a packet's header as input to a load balancing function 127 (typically a hash function) that selects the path for all packets in 128 a given flow. The keys chosen for the load balancing function depend 129 on the packet type; a typical set (for IP packets) is the IP source 130 and destination addresses, the protocol type, and (for TCP and UDP 131 traffic) the source and destination port numbers. An overly 132 conservative choice of fields may lead to many flows mapping to the 133 same hash value (and consequently poorer load balancing); an overly 134 aggressive choice may map a flow to multiple values, potentially 135 violating the above requirement. 137 For MPLS networks, most of the same principles (and benefits) apply. 138 However, finding useful keys in a packet for the purpose of load 139 balancing can be more of a challenge. In many cases, MPLS 140 encapsulation may require fairly deep inspection of packets to find 141 these keys at transit LSRs. 143 One way to eliminate the need for this deep inspection is to have the 144 ingress LSR of an MPLS Label Switched Path extract the appropriate 145 keys from a given packet, input them to its load balancing function, 146 and place the result in an additional label, termed the 'entropy 147 label', as part of the MPLS label stack it pushes onto that packet. 149 The packet's MPLS entire label stack can then be used by transit LSRs 150 to perform load balancing, as the entropy label introduces the right 151 level of "entropy" into the label stack. 153 There are four key reasons why this is beneficial: 155 1. at the ingress LSR, MPLS encapsulation hasn't yet occurred, so 156 deep inspection is not necessary; 158 2. the ingress LSR has more context and information about incoming 159 packets than transit LSRs; 161 3. ingress LSRs usually operate at lower bandwidths than transit 162 LSRs, allowing them to do more work per packet, and 164 4. transit LSRs do not need to perform deep packet inspection and 165 can load balance effectively using only a packet's MPLS label 166 stack. 168 This memo describes why entropy labels are needed and defines the 169 properties of entropy labels; in particular how they are generated 170 and received, and the expected behavior of transit LSRs. Finally, it 171 describes in general how signaling works and what needs to be 172 signaled, as well as specifics for the signaling of entropy labels 173 for LDP ([RFC5036]), BGP ([RFC3107], [RFC4364]), and RSVP-TE 174 ([RFC3209]). 176 1.1. Conventions used 178 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 179 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 180 document are to be interpreted as described in [RFC2119]. 182 The following acronyms are used: 184 LSR: Label Switching Router; 186 LER: Label Edge Router; 188 PE: Provider Edge router; 189 CE: Customer Edge device; and 191 FEC: Forwarding Equivalence Class. 193 The term ingress (or egress) LSR is used interchangeably with ingress 194 (or egress) LER. The term application throughout the text refers to 195 an MPLS application (such as a VPN or VPLS). 197 A label stack (say of three labels) is denoted by , where 198 L1 is the "outermost" label and L3 the innermost (closest to the 199 payload). Packet flows are depicted left to right, and signaling is 200 shown right to left (unless otherwise indicated). 202 The term 'label' is used both for the entire 32-bit label and the 20- 203 bit label field within a label. It should be clear from the context 204 which is meant. 206 1.2. Motivation 208 MPLS is very successful generic forwarding substrate that transports 209 several dozen types of protocols, most notably: IP, PWE3, VPLS and IP 210 VPNs. Within each type of protocol, there typically exist several 211 variants, each with a different set of load balancing keys, e.g., for 212 IP: IPv4, IPv6, IPv6 in IPv4, etc.; for PWE3: Ethernet, ATM, Frame- 213 Relay, etc. There are also several different types of Ethernet over 214 PW encapsulation, ATM over PW encapsulation, etc. as well. Finally, 215 given the popularity of MPLS, it is likely that it will continue to 216 be extended to transport new protocols. 218 Currently, each transit LSR along the path of a given LSP has to try 219 to infer the underlying protocol within an MPLS packet in order to 220 extract appropriate keys for load balancing. Unfortunately, if the 221 transit LSR is unable to infer the MPLS packet's protocol (as is 222 often the case), it will typically use the topmost (or all) MPLS 223 labels in the label stack as keys for the load balancing function. 224 The result may be an extremely inequitable distribution of traffic 225 across equal-cost paths exiting that LSR. This is because MPLS 226 labels are generally fairly coarse-grained forwarding labels that 227 typically describe a next-hop, or provide some of demultiplexing 228 and/or forwarding function, and do not describe the packet's 229 underlying protocol. 231 On the other hand, an ingress LSR (e.g., a PE router) has detailed 232 knowledge of an packet's contents, typically through a priori 233 configuration of the encapsulation(s) that are expected at a given 234 PE-CE interface, (e.g., IPv4, IPv6, VPLS, etc.). They also have more 235 flexible forwarding hardware. PE routers need this information and 236 these capabilities to: 238 a) apply the required services for the CE; 240 b) discern the packet's CoS forwarding treatment; 242 c) apply filters to forward or block traffic to/from the CE; 244 d) to forward routing/control traffic to an onboard management 245 processor; and, 247 e) load-balance the traffic on its uplinks to transit LSRs (e.g., 248 P routers). 250 By knowing the expected encapsulation types, an ingress LSR router 251 can apply a more specific set of payload parsing routines to extract 252 the keys appropriate for a given protocol. This allows for 253 significantly improved accuracy in determining the appropriate load 254 balancing behavior for each protocol. 256 If the ingress LSR were to capture the flow information so gathered 257 in a convenient form for downstream transit LSRs, transit LSRs could 258 remain completely oblivious to the contents of each MPLS packet, and 259 use only the captured flow information to perform load balancing. In 260 particular, there will be no reason to duplicate an ingress LSR's 261 complex packet/payload parsing functionality in a transit LSR. This 262 will result in less complex transit LSRs, enabling them to more 263 easily scale to higher forwarding rates, larger port density, lower 264 power consumption, etc. The idea in this memo is to capture this 265 flow information as a label, the so-called entropy label. 267 Ingress LSRs can also adapt more readily to new protocols and extract 268 the appropriate keys to use for load balancing packets of those 269 protocols. This means that deploying new protocols or services in 270 edge devices requires fewer concommitant changes in the core, 271 resulting in higher edge service velocity and at the same time more 272 stable core networks. 274 2. Approaches 276 There are two main approaches to encoding load balancing information 277 in the label stack. The first allocates multiple labels for a 278 particular Forwarding Equivalance Class (FEC). These labels are 279 equivalent in terms of forwarding semantics, but having multiple 280 labels allows flexibility in assigning labels to flows belonging to 281 the same FEC. This approach has the advantage that the label stack 282 has the same depth whether or not one uses label-based load 283 balancing; and so, consequently, there is no change to forwarding 284 operations on transit and egress LSRs. However, it has a major 285 drawback in that there is a significant increase in both signaling 286 and forwarding state. 288 The other approach encodes the load balancing information as an 289 additional label in the label stack, thus increasing the depth of the 290 label stack by one. With this approach, there is minimal change to 291 signaling state for a FEC; also, there is no change in forwarding 292 operations in transit LSRs, and no increase of forwarding state in 293 any LSR. The only purpose of the additional label is to increase the 294 entropy in the label stack, so this is called an "entropy label". 295 This memo focuses solely on this approach. 297 3. Entropy Labels 299 An entropy label (as used here) is a label: 301 1. that is not used for forwarding; 303 2. that is not signaled; and 305 3. whose only purpose in the label stack is to provide 'entropy' to 306 improve load balancing. 308 Entropy labels are generated by an ingress LSR, based entirely on 309 load balancing information. However, they MUST NOT have values in 310 the reserved label space (0-15). Entropy labels MUST be at the 311 bottom of the label stack, and thus the 'Bottom of Stack' (S) bit 312 ([RFC3032]) in the label should be set. To ensure that they are not 313 used inadvertently for forwarding, entropy labels SHOULD have a TTL 314 of 0. 316 Since entropy labels are generated by an ingress LSR, an egress LSR 317 MUST be able to tell unambiguously that a given label is an entropy 318 label. If any ambiguity is possible, the label above the entropy 319 label MUST be an 'entropy label indicator' (ELI), which indicates 320 that the following Label is an entropy label. An ELI is typically 321 signaled by an egress LSR and is added to the MPLS label stack along 322 with an entropy label by an ingress LSR. For many applications, the 323 use of entropy labels is unambiguous, and an ELI is not needed. If 324 used, an ELI MUST have S = 0 and SHOULD have a TTL of 0. 326 Applications for MPLS entropy labels include pseudowires ([RFC4447]), 327 Layer 3 VPNs ([RFC4364]), VPLS ([RFC4761], [RFC4762]) and Tunnel LSPs 328 carrying, say, IP traffic. [I-D.ietf-pwe3-fat-pw] explains how 329 entropy labels can be used for RFC 4447-style pseudowires, and thus 330 is complementary to this memo, which focuses on several other 331 applications of entropy labels. 333 4. Data Plane Processing of Entropy Labels 335 4.1. Ingress LSR 337 Suppose that for a particular application (or service or FEC), an 338 ingress LSR X is to push label stack , where TL is the 339 'tunnel label' and AL is the 'application label'. (Note the use of 340 the convention for label stacks described in Section 1.1. The use of 341 a two-label stack is just for illustrative purposes.) Suppose 342 furthermore that the egress LSR Y has told X that it is capable of 343 processing entropy labels for this application. If X can insert 344 entropy labels, it may use a label stack of for this 345 application, where EL is the entropy label. 347 When a packet for this application arrives at X, X does the 348 following: 350 1. X identifies the application to which the packet belongs, 351 identifies the egress LSR as Y, and thereby picks the outgoing 352 label stack to push onto the packet to send to Y; 354 2. X determines which keys that it will use for load balancing; 356 3. X, having kept state that Y can process entropy labels for this 357 application, generates an entropy label EL (based on the output 358 of the load balancing function), and 360 4. X pushes on to the packet before forwarding it to 361 the next LSR on its way to Y. 363 EL is a 'regular' 32-bit label whose S bit MUST be 1 and whose TTL 364 field SHOULD be 0. The load balancing information is encoded in the 365 20-bit label field. If X is told (via signaling) that it must use an 366 entropy label indicator with label value E, then X instead pushes 367 onto the packet, where ELI is a label whose S bit 368 MUST be 0, whose TTL SHOULD be 0, and whose 20-bit label field MUST 369 be E. The CoS fields for EL and ELI can be set to any values. 371 Note that ingress LSR X MUST NOT include an entropy label unless the 372 egress LSR Y for this application has indicated that it is ready to 373 receive entropy labels. Furthermore, if Y has signaled that an ELI 374 is needed, then X MUST include the ELI before the entropy label. 376 Note that the signaling and use of entropy labels in one direction 377 (signaling from Y to X, and data path from X to Y) has no bearing on 378 the behavior in the opposite direction (signaling from X to Y, and 379 data path from Y to X). 381 4.2. Transit LSR 383 Transit LSRs have virtually no change in forwarding behavior. For 384 load balancing, transit LSRs SHOULD use the whole label stack as keys 385 for the load balancing function. Transit LSRs MAY choose to look 386 beyond the label stack for further keys; however, if entropy labels 387 are being used, this may not be very useful. Looking beyond the 388 label stack may be the simplest approach in an environment where some 389 ingress LSRs use entropy labels and others don't, or for backward 390 compatibility. Thus, other than using the full label stack as input 391 to the load balancing function, transit LSRs are almost unaffected by 392 the use of entropy labels. 394 4.3. Egress LSR 396 If egresss LSR Y signals that it is capable of processing entropy 397 labels without an ELI for an application, then when Y receives a 398 packet with the application label, then Y looks to see if the S bit 399 is set. If so, Y applies its usual processing rules to the packet, 400 including popping the application label. If the S bit is not set, Y 401 assumes that the label below the application label is an entropy 402 label and pops both the application label and the entropy label. Y 403 SHOULD ensure that the entropy label has its S bit set. Y then 404 processes the packet as usual. Implementations may choose the order 405 in which they apply these operations, but the net result should be as 406 specified. 408 If Y signals that it is capable of processing entropy labels but that 409 an ELI is necessary for a given application, then when Y receives a 410 packet with the application label, Y processes the application label 411 as usual, then pops it. Y then checks whether the S bit on the 412 application label is set. If not, Y looks to see if the label below 413 the application label is the ELI. If so, Y further pops both the ELI 414 and the label below (which should be the entropy label). Y SHOULD 415 ensure that the ELI has its S bit unset, and that the entropy label 416 has its S bit set. If the S bit of the application label is set, or 417 the label below is not the ELI, Y processes the packet as usual 418 (there is no entropy label). 420 5. Signaling for Entropy Labels 422 An egress LSR Y may signal to ingress LSR(s) its ability to process 423 entropy labels on a per-application (or per-FEC) basis. As part of 424 this signaling, Y also signals the ELI to use, if any. 426 In cases where an application label is used and must be the 427 bottommost label in the label stack, Y MAY signal that no ELI is 428 needed for that application. 430 In cases where no application label exists, or where the application 431 label may not be the bottommost label in the label stack, Y MUST 432 signal a valid ELI to be used in conjunction with the entropy label 433 for this FEC. In this case, an ingress LSR will either not add an 434 entropy label, or push the ELI before the entropy label. This makes 435 the use or non-use of an entropy label by the ingress LSR 436 unambiguous. Valid ELI label values are strictly greater than 15. 438 It should be noted that egress LSR Y may use the same ELI value for 439 all applications for which an ELI is needed. The ELI MUST be a label 440 that does not conflict with any other labels that Y has advertised to 441 other LSRs for other applications. Furthermore, it should be noted 442 that the ability to process entropy labels (and the corresponding 443 ELI) may be asymmetric: an LSR X may be willing to process entropy 444 labels, whereas LSR Y may not be willing to process entropy labels. 445 The signaling extensions below allow for this asymmetry. 447 For an illustration of signaling and forwarding with entropy labels, 448 see Figure 9. 450 5.1. LDP Signaling 452 When using LDP for signaling tunnel labels ([RFC5036]), a Label 453 Mapping Message sub-TLV (Entropy Label sub-TLV) is used to signal an 454 egress LSR's ability to process entropy labels. 456 The presence of the Entropy Label sub-TLV in the Label Mapping 457 Message indicates to ingress LSRs that the egress LSR can process an 458 entropy label. In addition, the Entropy Label sub-TLV contains a 459 label value for the ELI. If the ELI is zero, this indicates the 460 egress doesn't need an ELI for the signaled application; if not, the 461 egress requires the given ELI with entropy labels. An example where 462 an ELI is needed is when the signaled application is an LSP that can 463 carry IP traffic. 465 The structure of the Entropy Label sub-TLV is shown below. 467 0 1 2 3 468 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 1 469 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 470 |U|F| Type (TBD) | Length (8) | 471 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 472 | Value | Must Be Zero | 473 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 475 Figure 1: Entropy Label sub-TLV 477 where: 479 U: Unknown bit. This bit MUST be set to 1. If the Entropy Label 480 sub-TLV is not understood, then the TLV is not known to the 481 receiver and MUST be ignored. 483 F: Forward bit. This bit MUST be set be set to 1. Since this 484 sub-TLV is going to be propagated hop-by-hop, the sub-TLV should 485 be forwarded even by nodes that may not understand it. 487 Type: sub-TLV Type field, as specified by IANA. 489 Length: sub-TLV Length field. This field specifies the total 490 length in octets of the Entropy Label sub-TLV. 492 Value: value of the Entropy Label Indicator Label. 494 5.2. BGP Signaling 496 When BGP [RFC4271] is used for distributing Network Layer 497 Reachability Information (NLRI) as described in, for example, 498 [RFC3107], [RFC4364] and [RFC4761], the BGP UPDATE message may 499 include the Entropy Label attribute. This is an optional, transitive 500 BGP attribute of type TBD. The inclusion of this attribute with an 501 NLRI indicates that the advertising BGP router can process entropy 502 labels as an egress LSR for that NLRI. If the attribute length is 503 less than three octets, this indicates that the egress doesn't need 504 an ELI for the signaled application. If the attribute length is at 505 least three octets, the first three octets encode an ELI label value 506 as the high order 20 bits; the egress requires this ELI with entropy 507 labels. An example where an ELI is needed is when the NLRI contains 508 unlabeled IP prefixes. 510 A BGP speaker S that originates an UPDATE should only include the 511 Entropy Label attribute if both of the following are true: 513 A1: S sets the BGP NEXT_HOP attribute to itself; AND 515 A2: S can process entropy labels for the given application. 517 If both A1 and A2 are true, and S needs an ELI to recognize entropy 518 labels, then S MUST include the ELI label value as part of the 519 Entropy Label attribute. An UPDATE SHOULD contain at most one 520 Entropy Label attribute. 522 Suppose a BGP speaker T receives an UPDATE U with the Entropy Label 523 attribute ELA. T has two choices. T can simply re-advertise U with 524 the same ELA if either of the following is true: 526 B1: T does not change the NEXT_HOP attribute; OR 528 B2: T simply swaps labels without popping the entire label stack and 529 processing the payload below. 531 An example of the use of B1 is Route Reflectors; an example of the 532 use of B2 is illustrated in Section 9.3.1.2. 534 However, if T changes the NEXT_HOP attribute for U and in the data 535 plane pops the entire label stack to process the payload, T MUST 536 remove ELA. T MAY include a new Entropy Label attribute ELA' for 537 UPDATE U' if both of the following are true: 539 C1: T sets the NEXT_HOP attribute of U' to itself; AND 541 C2: T can process entropy labels for the given application. 543 Again, if both C1 and C2 are true, and T needs an ELI to recognize 544 entropy labels, then T MUST include the ELI label value as part of 545 the Entropy Label attribute. 547 5.3. RSVP-TE Signaling 549 Entropy Label support is signaled in RSVP-TE [RFC3209] using an 550 Entropy Label Attribute TLV (Type TBD) of the LSP_ATTRIBUTES object 551 [RFC5420]. The presence of this attribute indicates that the 552 signaler (the egress in the downstream direction using Resv messages; 553 the ingress in the upstream direction using Path messages) can 554 process entropy labels. The Entropy Label Attribute contains a value 555 for the ELI. If the ELI is zero, this indicates that the signaler 556 doesn't need an ELI for this application; if not, then the signaler 557 requires the given ELI with entropy labels. An example where an ELI 558 is needed is when the signaled LSP can carry IP traffic. 560 The format of the Entropy Label Attribute is as follows: 562 0 1 2 3 563 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 1 564 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 565 | Entropy Label Attribute | Length (4) | 566 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 567 | ELI Label | MBZ | 568 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 570 An egress LSR includes the Entropy Label Attribute in a Resv message 571 to indicate that it can process entropy labels in the downstream 572 direction of the signaled LSP. 574 An ingress LSR includes the Entropy Label Attribute in a Path message 575 for a bi-directional LSP to indicate that it can process entropy 576 labels in the upstream direction of the signaled LSP. If the 577 signaled LSP is not bidirectional, the Entropy Label Attribute SHOULD 578 NOT be included in the Path message, and egress LSR(s) SHOULD ignore 579 the attribute, if any. 581 As described in Section 8, there is also the need to distribute an 582 ELI from the ingress (upstream label allocation). In the case of 583 RSVP-TE, this is accomplished using the Upstream ELI Attribute TLV of 584 the LSP_ATTRIBUTES object, as shown below: 586 0 1 2 3 587 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 1 588 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 589 | Upstream ELI Attribute | Length (4) | 590 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 591 | ELI Label | MBZ | 592 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 594 6. Operations, Administration, and Maintenance (OAM) and Entropy Labels 596 Generally OAM comprises a set of functions operating in the data 597 plane to allow a network operator to monitor its network 598 infrastructure and to implement mechanisms in order to enhance the 599 general behavior and the level of performance of its network, e.g., 600 the efficient and automatic detection, localization, diagnosis and 601 handling of defects. 603 Currently defined OAM mechanisms for MPLS include LSP Ping/Traceroute 604 [RFC4379] and Bidirectional Failure Detection (BFD) for MPLS 605 [RFC5884]. The latter provides connectivity verification between the 606 endpoints of an LSP, and recommends establishing a separate BFD 607 session for every path between the endpoints. 609 The LSP traceroute procedures of [RFC4379] allow an ingress LSR to 610 obtain label ranges that can be used to send packets on every path to 611 the egress LSR. It works by having ingress LSR sequentially ask the 612 transit LSRs along a particular path to a given egress LSR to return 613 a label range such that the inclusion of a label in that range in a 614 packet will cause the replying transit LSR to send that packet out 615 the egress interface for that path. The ingress provides the label 616 range returned by transit LSR N to transit LSR N + 1, which returns a 617 label range which is less than or equal in span to the range provided 618 to it. This process iterates until the penultimate transit LSR 619 replies to the ingress LSR with a label range that is acceptable to 620 it and to all LSRs along path preceding it for forwarding a packet 621 along the path. 623 However, the LSP traceroute procedures do not specify where in the 624 label stack the value from the label range is to be placed, whether 625 deep packet inspection is allowed and if so, which keys and key 626 values are to be used. 628 This memo updates LSP traceroute by specifying that the value from 629 the label range is to be placed in the entropy label. Deep packet 630 inspection is thus not necessary, although an LSR may use it, 631 provided it do so consistently, i.e., if the label range to go to a 632 given downstream LSR is computed with deep packet inspection, then 633 the data path should use the same approach and the same keys. 635 In order to have a BFD session on a given path, a value from the 636 label range for that path should be used as the EL value for BFD 637 packets sent on that path. 639 As part of the MPLS-TP work, an in-band OAM channel is defined in 640 [RFC5586]. Packets sent in this channel are identified with a 641 reserved label, the Generic Associated Channel Label (GAL) placed at 642 the bottom of the MPLS label stack. In order to use the inband OAM 643 channel with entropy labels, this memo relaxes the restriction that 644 the GAL must be at the bottom of the MPLS label stack. Rather, the 645 GAL is placed in the MPLS label stack above the entropy label so that 646 it effectively functions as an application label. 648 7. MPLS-TP and Entropy Labels 650 Since MPLS-TP does not use ECMP, entropy labels are not applicable to 651 an MPLS-TP deployment. 653 8. Point-to-Multipoint LSPs and Entropy Labels 655 Point-to-Multipoint (P2MP) LSPs [RFC4875] typically do not use ECMP 656 for load balancing, as the combination of replication and 657 multipathing can lead to duplicate traffic delivery. However, P2MP 658 LSPs can traverse Bundled Links [RFC4201] and LAGs. In both these 659 cases, load balancing is useful, and hence entropy labels can be of 660 some value for P2MP LSPs. 662 There are two potential complications with the use of entropy labels 663 in the context of P2MP LSPs, both a consequence of the fact that the 664 entire label stack below the P2MP label must be the same for all 665 egress LSRs. First, all egress LSRs must be willing to receive 666 entropy labels; if even one egress LSR is not willing, then entropy 667 labels MUST NOT be used for this P2MP LSP. Second, if an ELI is 668 required, all egress LSRs must agree to the same value of ELI. This 669 can be achieved by upstream allocation of the ELI; in particular, for 670 RSVP-TE P2MP LSPs, the ingress LSR distributes the ELI value using 671 the Upstream ELI Attribute TLV of the LSP_ATTRIBUTES object, defined 672 in Section 5.3. 674 With regard to the first issue, the ingress LSR MUST keep track of 675 the ability of each egress LSR to process entropy labels, especially 676 since the set of egress LSRs of a given P2MP LSP may change over 677 time. Whenever an existing egress LSR leaves, or a new egress LSR 678 joins the P2MP LSP, the ingress MUST re-evaluate whether or not to 679 include entropy labels for the P2MP LSP. 681 In some cases, it may be feasible to deploy two P2MP LSPs, one to 682 entropy label capable egress LSRs, and the other to the remaining 683 egress LSRs. However, this requires more state in the network, more 684 bandwidth, and more operational overhead (tracking EL-capable LSRs, 685 and provisioning P2MP LSPs accordingly). Furthermore, this approach 686 may not work for some applications (such mVPNs and VPLS) which 687 automatically create and/or use P2MP LSPs for their multicast 688 requirements. 690 9. Entropy Labels and Applications 692 This section describes the usage of entropy labels in various 693 scenarios with different applications. 695 9.1. Tunnels 697 Tunnel LSPs, signaled with either LDP or RSVP-TE, typically carry 698 other MPLS applications such as VPNs or pseudowires. This being the 699 case, if the egress LSR of a tunnel LSP is willing to process entropy 700 labels, it would signal the need for an Entropy Label Indicator to 701 distinguish between entropy labels and other application labels. 703 In the figures below, the following convention is used to depict 704 information signaled between X and Y: 706 X ---------- ... ---------- Y 707 app: <--- [label L, ELI value] 709 This means Y signals to X label L for application app. The ELI value 710 can be one of: 712 -: meaning entropy labels are NOT accepted; 714 0: meaning entropy labels are accepted, no ELI is needed; or 716 E: entropy labels are accepted, ELI label E is required. 718 The following illustrates a simple intra-AS tunnel LSP. 720 X -------- A --- ... --- B -------- Y 721 tunnel LSP L: [TL, E] <--- ... <--- [TL0, E] 723 IP pkt: push ---------------> 725 Figure 2: Tunnel LSPs and Entropy Labels 727 Tunnel LSPs may cross Autonomous System (AS) boundaries, usually 728 using BGP ([RFC3107]). In this case, the AS Border Routers (ASBRs) 729 MAY simply propagate the egress LSR's ability to process entropy 730 labels, or they MAY declare that entropy labels may not be used. If 731 an ASBR (say A2 below) chooses to propagate the egress LSR Y's 732 ability to process entropy labels, A2 MUST also propagate Y's choice 733 of ELI. 735 X ---- ... ---- A1 ------- A2 ---- ... ---- Y 736 intra-AS LSP A2-Y: <--- [TL0, E] 737 inter-AS LSP A1-A2: [AL, E] 738 intra-AS LSP X-A1: <--- [TL1, E] 740 IP pkt: push 742 Here, ASBR A2 chooses to propagate Y's ability to process entropy 743 labels, by "translating" Y's signaling of entropy label capability 744 (say using LDP) to BGP; and A1 translate A2's BGP signaling to (say) 745 RSVP-TE. The end-to-end tunnel (X to Y) will have entropy labels if 746 X chooses to insert them. 748 Figure 3: Inter-AS Tunnel LSP with Entropy Labels 750 X ---- ... ---- A1 ------- A2 ---- ... ---- Y 751 intra-AS LSP A2-Y: <--- [TL0, E] 752 inter-AS LSP A1-A2: [AL, E] 753 intra-AS LSP X-A1: <--- [TL1, -] 755 IP pkt: push --> 757 Here, ASBR A1 decided that entropy labels are not to be used; thus, 758 the end-to-end tunnel cannot have entropy labels, even though both X 759 and Y may be capable of inserting and processing entropy labels. 761 Figure 4: Inter-AS Tunnel LSP with no Entropy Labels 763 9.2. LDP Pseudowires 765 [I-D.ietf-pwe3-fat-pw] describes the signaling and use of entropy 766 labels in the context of RFC 4447 pseudowires, so this will not be 767 described further here. 769 [RFC4762] specifies the use of LDP for signaling VPLS pseudowires. 770 An egress VPLS PE that can process entropy labels can indicate this 771 by adding the Entropy Label sub-TLV in the LDP message it sends to 772 other PEs. An ELI is not required. An ingress PE must maintain 773 state per egress PE as to whether it can process entropy labels. 775 X -------- A --- ... --- B -------- Y 776 tunnel LSP L: [TL, E] <--- ... <--- [TL0, E] 777 VPLS label: <------------------------ [VL, 0] 779 VPLS pkt: push --------------> 781 Figure 5: Entropy Labels with LDP VPLS 783 Note that although the underlying tunnel LSP signaling indicated the 784 need for an ELI, VPLS packets don't need an ELI, and thus the label 785 stack pushed by X do not have one. 787 [RFC4762] also describes the notion of "hierarchical VPLS" (H-VPLS). 788 In H-VPLS, 'hub PEs' remove the label stack and process VPLS packets; 789 thus, they must make their own decisions on the use of entropy 790 labels, independent of other hub PEs or spoke PEs with which they 791 exchange signaling. In the example below, spoke PEs X and Y and hub 792 PE B can process entropy labels, but hub PE A cannot. 794 X ---- ... ---- A ---- ... ---- B ---- ... ---- Y 795 spoke PW1: <--- [SL1, 0] 796 hub-hub PW: <---- [HL, 0] 797 spoke PW2: <--- [SL2, -] 799 SPW2 pkt: push 800 H-H PW pkt: push 801 SPW1 pkt: push 803 Figure 6: Entropy Labels with H-VPLS 805 9.3. BGP Applications 807 Section 9.1 described a BGP application for the creation of inter-AS 808 tunnel LSPs. This section describes two other BGP applications, IP 809 VPNs ([RFC4364]) and BGP VPLS ([RFC4761]). An egress PE for either 810 of these applications indicates its ability to process entropy labels 811 by adding the Entropy Label attribute to its BGP UPDATE message. 812 Again, ingress PEs must maintain per-egress PE state regarding its 813 ability to process entropy labels. In this section, both of these 814 applications will be referred to as VPNs. 816 In the intra-AS case, PEs signal application labels and entropy label 817 capability to each other, either directly, or via Route Reflectors 818 (RRs). If RRs are used, they must not change the BGP NEXT_HOP 819 attribute in the UPDATE messages; furthermore, they can simply pass 820 on the Entropy Label attribute as is. 822 X -------- A --- ... --- B -------- Y 823 tunnel LSP L: [TL, E] <--- ... <--- [TL0, E] 824 BGP VPN label: <------------------------ [VL, 0] 826 BGP VPN pkt: push --------------> 828 Figure 7: Entropy Labels with Intra-AS BGP apps 830 For BGP VPLS, the application label is at the bottom of stack, so no 831 ELI is needed. For BGP IP VPNs, the application label is usually at 832 the bottom of stack, so again no ELI is needed. However, in the case 833 of Carrier's Carrier (CsC) VPNs, the BGP VPN label may not be at the 834 bottom of stack. In this case, an ELI is necessary for CsC VPN 835 packets with entropy labels to distinguish them from nested VPN 836 packets. In the example below, the nested VPN signaling is not 837 shown; the egress PE for the nested VPN (not shown) must signal 838 whether or not it can process egress labels, and the ingress nested 839 VPN PE may insert an entropy label if so. 841 Three cases are shown: a plain BGP VPN packet, a CsC VPN packet 842 originating from X, and a transit nested VPN packet originating from 843 a nested VPN ingress PE (conceptually to the left of X). It is 844 assumed that the nested VPN packet arrives at X with label stack where ZL is the tunnel label (to be swapped with ) and 846 CVL is the nested VPN label. Note that Y can use the same ELI for 847 the tunnel LSP and the CsC VPN (and any other application that needs 848 an ELI). 850 X -------- A --- ... --- B -------- Y 851 tunnel LSP L: [TL, E] <--- ... <--- [TL0, E] 852 BGP VPN label: <------------------------ [VL, 0] 853 BGP CsC VPN label: <------------------------ [CL, E] 855 BGP VPN pkt: push --------------> 856 CsC VPN pkt: push -----------> 857 nested VPN pkt: swap with --------> 859 Figure 8: Entropy Labels with CoC VPN 861 9.3.1. Inter-AS BGP VPNs 863 There are three commonly used options for inter-AS IP VPNs and BGP 864 VPLS, known informally as "Option A", "Option B" and "Option C". 865 This section describes how entropy labels can be used in these 866 options. 868 9.3.1.1. Option A Inter-AS VPNs 870 In option A, an ASBR pops the full label stack of a VPN packet 871 exiting an AS, processes the payload header (IP or Ethernet), and 872 forwards the packet natively (i.e., as IP or Ethernet, but not as 873 MPLS) to the peer ASBR. Thus, entropy label signaling and insertion 874 are completely local to each AS. The inter-AS paths do not use 875 entropy labels, as they do not use a label stack. 877 9.3.1.2. Option B Inter-AS VPNs 879 The ASBRs in option B inter-AS VPNs have a choice (usually determined 880 by configuration) of whether to just swap labels (from within the AS 881 to the neighbor AS or vice versa), or to pop the full label stack and 882 process the packet natively. This choice occurs at each ASBR in each 883 direction. In the case of native packet processing at an ASBR, 884 entropy label signaling and insertion is local to each AS and to the 885 inter-AS paths (which, unlike option A, do have labeled packets). 887 In the case of simple label swapping at an ASBR, the ASBR can 888 propagate received entropy label signaling onward. That is, if a PE 889 signals to its ASBR that it can process entropy labels (via an 890 Entropy Label attribute), the ASBR can propagate that attribute to 891 its peer ASBR; if a peer ASBR signals that it can process entropy 892 labels, the ASBR can propagate that to all PEs within its AS). Note 893 that this is the case even though ASBRs change the BGP NEXT_HOP 894 attribute to "self", because of clause B2 in Section 5.2. 896 9.3.1.3. Option C Inter-AS VPNs 898 In Option C inter-AS VPNs, the ASBRs are not involved in signaling; 899 they do not have VPN state; they simply swap labels of inter-AS 900 tunnels. Signaling is PE to PE, usually via Route Reflectors; 901 however, if RRs are used, the RRs do not change the BGP NEXT_HOP 902 attribute. Thus, entropy label signaling and insertion are on a PE- 903 pair basis, and the intermediate routers, ASBRs and RRs do not play a 904 role. 906 9.4. Multiple Applications 908 It has been mentioned earlier that an ingress PE must keep state per 909 egress PE with regard to its ability to process entropy labels. An 910 ingress PE must also keep state per application, as entropy label 911 processing must be based on the application context in which a packet 912 is received (and of course, the corresponding entropy label 913 signaling). 915 In the example below, an egress LSR Y signals a tunnel LSP L, and is 916 prepared to receive entropy labels on L, but requires an ELI. 917 Furthermore, Y signals two pseudowires PW1 and PW2 with labels PL1 918 and PL2, respectively, and indicates that it can receive entropy 919 labels for both pseudowires without the need of an ELI; and finally, 920 Y signals a L3 VPN with label VL, but Y does not indicate that it can 921 receive entropy labels for the L3 VPN. Ingress LSR X chooses to send 922 native IP packets to Y over L with entropy labels, thus X must 923 include the given ELI (yielding a label stack of ). X 924 chooses to add entropy labels on PW1 packets to Y, with a label stack 925 of , but chooses not to do so for PW2 packets. X must 926 not send entropy labels on L3 VPN packets to Y, i.e., the label stack 927 must be . 929 X -------- A --- ... --- B -------- Y 930 tunnel LSP L: [TL, E] <--- ... <--- [TL0, E] 931 PW1 label: <----------------------- [PL1, 0] 932 PW2 label: <----------------------- [PL2, 0] 933 VPN label: <----------------------- [VL, -] 935 IP pkt: push -------------> 936 PW1 pkt: push -------------> 937 PW2 pkt: push -----------------> 938 VPN pkt: push ------------------> 940 Figure 9: Entropy Labels for Multiple Applications 942 10. Security Considerations 944 This document describes advertisement of the capability to support 945 receipt of entropy-labels and an Entropy Label Indicator that an 946 ingress LSR may apply to MPLS packets in order to allow transit LSRs 947 to attain better load-balancing across LAG and/or ECMP paths in the 948 network. 950 This document does not introduce new security vulnerabilities to LDP. 951 Please refer to the Security Considerations section of LDP 952 ([RFC5036]) for security mechanisms applicable to LDP. 954 Given that there is no end-user control over the values used for 955 entropy labels, there is little risk of Entropy Label forgery which 956 could cause uneven load-balancing in the network. 958 If Entropy Label Capability is not signaled from an egress PE to an 959 ingress PE, due to, for example, malicious configuration activity on 960 the egress PE, then the PE's will fall back to not using entropy 961 labels for load-balancing traffic over LAG or ECMP paths which, in 962 some cases, in no worse than the behavior observed in current 963 production networks. That said, operators are recommended to monitor 964 changes to PE configurations and, more importantly, the fairness of 965 load distribution over equal-cost LAG or ECMP paths. If the fairness 966 of load distribution over a set of paths changes that could indicate 967 a misconfiguration, bug or other non-optimal behavior on their PE's 968 and they should take corrective action. 970 Given that most applications already signal an Application Label, 971 e.g.: IPVPNs, LDP VPLS, BGP VPLS, whose Bottom of Stack bit is being 972 re-used to signal entropy label capability, there is little to no 973 additional risk that traffic could be misdirected into an 974 inappropriate IPVPN VRF or VPLS VSI at the egress PE. 976 In the context of downstream-signaled entropy labels that require the 977 use of an Entropy Label Indicator (ELI), there should be little to no 978 additional risk because the egress PE is solely responsible for 979 allocating an ELI value and ensuring that ELI label value DOES NOT 980 conflict with other MPLS labels it has previously allocated. On the 981 other hand, for upstream-signaled entropy labels, e.g.: RSVP-TE 982 point-to-point or point-to-multipoint LSP's or Multicast LDP (mLDP) 983 point-to-multipoint or multipoint-to-multipoint LSP's, there is a 984 risk that the head-end MPLS LER may choose an ELI value that is 985 already in use by a downstream LSR or LER. In this case, it is the 986 responsibility of the downstream LSR or LER to ensure that it MUST 987 NOT accept signaling for an ELI value that conflicts with MPLS 988 label(s) that are already in use. 990 11. IANA Considerations 992 11.1. LDP Entropy Label TLV 994 IANA is requested to allocate the next available value from the IETF 995 Consensus range in the LDP TLV Type Name Space Registry as the 996 "Entropy Label TLV". 998 11.2. BGP Entropy Label Attribute 1000 IANA is requested to allocate the next available Path Attribute Type 1001 Code from the "BGP Path Attributes" registry as the "BGP Entropy 1002 Label Attribute". 1004 11.3. Attribute Flags for LSP_Attributes Object 1006 IANA is requested to allocate a new bit from the "Attribute Flags" 1007 sub-registry of the "RSVP TE Parameters" registry. 1009 Bit | Name | Attribute | Attribute | RRO 1010 No | | Flags Path | Flags Resv | 1011 ----+----------------------+------------+------------+----- 1012 TBD Entropy Label LSP Yes Yes No 1014 11.4. Attributes TLV for LSP_Attributes Object 1016 IANA is requested to allocate the next available value from the 1017 "Attributes TLV" sub-registry of the "RSVP TE Parameters" registry. 1019 12. Acknowledgments 1021 We wish to thank Ulrich Drafz for his contributions, as well as the 1022 entire 'hash label' team for their valuable comments and discussion. 1024 13. References 1026 13.1. Normative References 1028 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1029 Requirement Levels", BCP 14, RFC 2119, March 1997. 1031 [RFC3032] Rosen, E., Tappan, D., Fedorkow, G., Rekhter, Y., 1032 Farinacci, D., Li, T., and A. Conta, "MPLS Label Stack 1033 Encoding", RFC 3032, January 2001. 1035 [RFC3107] Rekhter, Y. and E. Rosen, "Carrying Label Information in 1036 BGP-4", RFC 3107, May 2001. 1038 [RFC3209] Awduche, D., Berger, L., Gan, D., Li, T., Srinivasan, V., 1039 and G. Swallow, "RSVP-TE: Extensions to RSVP for LSP 1040 Tunnels", RFC 3209, December 2001. 1042 [RFC5420] Farrel, A., Papadimitriou, D., Vasseur, JP., and A. 1043 Ayyangarps, "Encoding of Attributes for MPLS LSP 1044 Establishment Using Resource Reservation Protocol Traffic 1045 Engineering (RSVP-TE)", RFC 5420, February 2009. 1047 13.2. Informative References 1049 [I-D.ietf-pwe3-fat-pw] 1050 Bryant, S., Filsfils, C., Drafz, U., Kompella, V., Regan, 1051 J., and S. Amante, "Flow Aware Transport of Pseudowires 1052 over an MPLS PSN", draft-ietf-pwe3-fat-pw-05 (work in 1053 progress), October 2010. 1055 [RFC4201] Kompella, K., Rekhter, Y., and L. Berger, "Link Bundling 1056 in MPLS Traffic Engineering (TE)", RFC 4201, October 2005. 1058 [RFC4271] Rekhter, Y., Li, T., and S. Hares, "A Border Gateway 1059 Protocol 4 (BGP-4)", RFC 4271, January 2006. 1061 [RFC4364] Rosen, E. and Y. Rekhter, "BGP/MPLS IP Virtual Private 1062 Networks (VPNs)", RFC 4364, February 2006. 1064 [RFC4379] Kompella, K. and G. Swallow, "Detecting Multi-Protocol 1065 Label Switched (MPLS) Data Plane Failures", RFC 4379, 1066 February 2006. 1068 [RFC4447] Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G. 1069 Heron, "Pseudowire Setup and Maintenance Using the Label 1070 Distribution Protocol (LDP)", RFC 4447, April 2006. 1072 [RFC4761] Kompella, K. and Y. Rekhter, "Virtual Private LAN Service 1073 (VPLS) Using BGP for Auto-Discovery and Signaling", 1074 RFC 4761, January 2007. 1076 [RFC4762] Lasserre, M. and V. Kompella, "Virtual Private LAN Service 1077 (VPLS) Using Label Distribution Protocol (LDP) Signaling", 1078 RFC 4762, January 2007. 1080 [RFC4875] Aggarwal, R., Papadimitriou, D., and S. Yasukawa, 1081 "Extensions to Resource Reservation Protocol - Traffic 1082 Engineering (RSVP-TE) for Point-to-Multipoint TE Label 1083 Switched Paths (LSPs)", RFC 4875, May 2007. 1085 [RFC5036] Andersson, L., Minei, I., and B. Thomas, "LDP 1086 Specification", RFC 5036, October 2007. 1088 [RFC5586] Bocci, M., Vigoureux, M., and S. Bryant, "MPLS Generic 1089 Associated Channel", RFC 5586, June 2009. 1091 [RFC5884] Aggarwal, R., Kompella, K., Nadeau, T., and G. Swallow, 1092 "Bidirectional Forwarding Detection (BFD) for MPLS Label 1093 Switched Paths (LSPs)", RFC 5884, June 2010. 1095 Appendix A. Applicability of LDP Entropy Label sub-TLV 1097 In the case of unlabeled IPv4 (Internet) traffic, the Best Current 1098 Practice is for an egress LSR to propagate eBGP learned routes within 1099 a SP's Autonomous System after resetting the BGP next-hop attribute 1100 to one of its Loopback IP addresses. That Loopback IP address is 1101 injected into the Service Provider's IGP and, concurrently, a label 1102 assigned to it via LDP. Thus, when an ingress LSR is performing a 1103 forwarding lookup for a BGP destination it recursively resolves the 1104 associated next-hop to a Loopback IP address and associated LDP label 1105 of the egress LSR. 1107 Thus, in the context of unlabeled IPv4 traffic, the LDP Entropy Label 1108 sub-TLV will typically be applied only to the FEC for the Loopback IP 1109 address of the egress LSR and the egress LSR will not announce an 1110 entropy label capability for the eBGP learned route. 1112 Authors' Addresses 1114 Kireeti Kompella 1115 Juniper Networks 1116 1194 N. Mathilda Ave. 1117 Sunnyvale, CA 94089 1118 US 1120 Email: kireeti@juniper.net 1122 John Drake 1123 Juniper Networks 1124 1194 N. Mathilda Ave. 1125 Sunnyvale, CA 94089 1126 US 1128 Email: jdrake@juniper.net 1130 Shane Amante 1131 Level 3 Communications, LLC 1132 1025 Eldorado Blvd 1133 Broomfield, CO 80021 1134 US 1136 Email: shane@level3.net 1138 Wim Henderickx 1139 Alcatel-Lucent 1140 Copernicuslaan 50 1141 2018 Antwerp 1142 Belgium 1144 Email: wim.henderickx@alcatel-lucent.com 1146 Lucy Yong 1147 Huawei USA 1148 1700 Alma Dr. Suite 500 1149 Plano, TX 75075 1150 US 1152 Email: lucyyong@huawei.com