idnits 2.17.1 draft-ietf-pwe3-fat-pw-04.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (July 9, 2010) is 5034 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) ** Obsolete normative reference: RFC 4379 (Obsoleted by RFC 8029) ** Obsolete normative reference: RFC 4447 (Obsoleted by RFC 8077) == Outdated reference: A later version (-02) exists of draft-kompella-mpls-entropy-label-00 Summary: 2 errors (**), 0 flaws (~~), 2 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 PWE3 S. Bryant, Ed. 3 Internet-Draft C. Filsfils 4 Intended status: Standards Track Cisco Systems 5 Expires: January 10, 2011 U. Drafz 6 Deutsche Telekom 7 V. Kompella 8 J. Regan 9 Alcatel-Lucent 10 S. Amante 11 Level 3 Communications 12 July 9, 2010 14 Flow Aware Transport of Pseudowires over an MPLS PSN 15 draft-ietf-pwe3-fat-pw-04 17 Abstract 19 Where the payload carried over a pseudowire carries a number of 20 identifiable flows it can in some circumstances be desirable to carry 21 those flows over the equal cost multiple paths (ECMPs) that exist in 22 the packet switched network. Most forwarding engines are able to 23 hash based on label stacks and use this to balance flows over ECMPs. 24 This draft describes a method of identifying the flows, or flow 25 groups, to the label switched routers by including an additional 26 label in the label stack. 28 Requirements Language 30 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 31 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 32 document are to be interpreted as described in RFC2119 [RFC2119]. 34 Status of this Memo 36 This Internet-Draft is submitted in full conformance with the 37 provisions of BCP 78 and BCP 79. 39 Internet-Drafts are working documents of the Internet Engineering 40 Task Force (IETF). Note that other groups may also distribute 41 working documents as Internet-Drafts. The list of current Internet- 42 Drafts is at http://datatracker.ietf.org/drafts/current/. 44 Internet-Drafts are draft documents valid for a maximum of six months 45 and may be updated, replaced, or obsoleted by other documents at any 46 time. It is inappropriate to use Internet-Drafts as reference 47 material or to cite them other than as "work in progress." 48 This Internet-Draft will expire on January 10, 2011. 50 Copyright Notice 52 Copyright (c) 2010 IETF Trust and the persons identified as the 53 document authors. All rights reserved. 55 This document is subject to BCP 78 and the IETF Trust's Legal 56 Provisions Relating to IETF Documents 57 (http://trustee.ietf.org/license-info) in effect on the date of 58 publication of this document. Please review these documents 59 carefully, as they describe your rights and restrictions with respect 60 to this document. Code Components extracted from this document must 61 include Simplified BSD License text as described in Section 4.e of 62 the Trust Legal Provisions and are provided without warranty as 63 described in the Simplified BSD License. 65 Table of Contents 67 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 68 1.1. ECMP in Label Switched Routers . . . . . . . . . . . . . . 5 69 1.2. Flow Label . . . . . . . . . . . . . . . . . . . . . . . . 5 70 2. Native Service Processing Function . . . . . . . . . . . . . . 6 71 3. Pseudowire Forwarder . . . . . . . . . . . . . . . . . . . . . 6 72 3.1. Encapsulation . . . . . . . . . . . . . . . . . . . . . . 7 73 4. Signaling the Presence of the Flow Label . . . . . . . . . . . 8 74 4.1. Structure of Flow Label Sub-TLV . . . . . . . . . . . . . 9 75 5. Multi-Segment Pseudowires . . . . . . . . . . . . . . . . . . 10 76 6. OAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 77 7. Applicability of FAT PWs . . . . . . . . . . . . . . . . . . . 11 78 7.1. Equal Cost Multiple Paths . . . . . . . . . . . . . . . . 12 79 7.2. Link Aggregation Groups . . . . . . . . . . . . . . . . . 13 80 7.3. Multiple RSVP-TE Paths . . . . . . . . . . . . . . . . . . 13 81 7.4. The Single Large Flow Case . . . . . . . . . . . . . . . . 13 82 7.5. Applicability to MPLS-TP . . . . . . . . . . . . . . . . . 15 83 7.6. Asymmetric Operation . . . . . . . . . . . . . . . . . . . 15 84 8. Applicability to MPLS . . . . . . . . . . . . . . . . . . . . 15 85 9. Security Considerations . . . . . . . . . . . . . . . . . . . 15 86 10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 16 87 11. Congestion Considerations . . . . . . . . . . . . . . . . . . 16 88 12. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 17 89 13. References . . . . . . . . . . . . . . . . . . . . . . . . . . 17 90 13.1. Normative References . . . . . . . . . . . . . . . . . . . 17 91 13.2. Informative References . . . . . . . . . . . . . . . . . . 18 92 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 18 94 1. Introduction 96 A pseudowire (PW) [RFC3985] is normally transported over one single 97 network path, even if multiple Equal Cost Multiple Paths (ECMP) exit 98 between the ingress and egress PW provider edge (PE) 99 equipments[RFC4385] [RFC4928]. This is required to preserve the 100 characteristics of the emulated service (e.g. to avoid misordering 101 SAToP pseudowire packets [RFC4553] or subjecting the packets to 102 unusable inter-arrival times ). The use of a single path to preserve 103 order remains the default mode of operation of a pseudowire (PW). 104 The new capability proposed in this document is an OPTIONAL mode 105 which may be used when the use of ECMP paths for is known to be 106 beneficial (and not harmful) to the operation of the PW. 108 Some pseudowires are used to transport large volumes of IP traffic 109 between routers at two locations. One example of this is the use of 110 an Ethernet pseudowire to create a virtual direct link between a pair 111 of routers. Such pseudowire's may carry from hundred's of Mbps to 112 Gbps of traffic. Such pseudowire's do not require strict ordering to 113 be preserved between packets of the pseudowire. They only require 114 ordering to be preserved within the context of each individual 115 transported IP flow. Some operators have requested the ability to 116 explicitly configure such a pseudowire to leverage the availability 117 of multiple ECMP paths. This allows for better capacity planning as 118 the statistical multiplexing of a larger number of smaller flows is 119 more efficient than with a smaller set of larger flows. Although 120 Ethernet is used as an example above, the mechanisms described in 121 this draft are general mechanisms that may be applied to any 122 pseudowire type in which there are identifiable flows, and in which 123 there is no requirement to preserve the order between those flows. 125 Typically, forwarding hardware can deduce that an IP payload is being 126 directly carried by an MPLS label stack, and is capable of looking at 127 some fields in packets to construct hash buckets for conversations or 128 flows. However, an intermediate node has no information on the type 129 pseudowire being carried in the packet. This limits the forwarder at 130 the intermediate node to only being able to make an ECMP choice based 131 on a hash of the label stack. In the case of a pseudowire emulating 132 a high bandwidth trunk, the granularity obtained by hashing the 133 default label stack is inadequate for satisfactory load-balancing. 134 The ingress node, however, is in the special position of being able 135 to look at the un-encapsulated packet and spread flows amongst any 136 available ECMP paths, or even any Loop-Free Alternates [RFC5286] . 137 This draft proposes a method to introduce granularity on the hashing 138 of traffic running over pseudowires by introducing an additional 139 label, chosen by the ingress node, and placed at the bottom of the 140 label stack. 142 In addition to providing an indication of the flow structure for use 143 in ECMP forwarding decisions, the mechanism described in the document 144 may also be used to select flows for distribution over an 802.1ad 145 link aggregation group that has been used in an MPLS network. 147 1.1. ECMP in Label Switched Routers 149 Label switched routers commonly hash the label stack or some elements 150 of the label stack as a method of discriminating between flows, in 151 order to distribute those flows over the available equal cost 152 multiple paths that exist in the network. Since the label at the 153 bottom of stack is usually the label most closely associated with the 154 flow, this normally provides the greatest entropy, and hence is 155 usually included in the hash. This draft describes a method of 156 adding an additional label at the bottom of stack in order to 157 facilitate the load balancing of the flows within a pseudowire over 158 the available ECMPs. A similar design for general MPLS use has also 159 been proposed [I-D.kompella-mpls-entropy-label], however that is 160 outside the scope of this draft. 162 An alternative method of load balancing by creating a number of 163 pseudowires and distributing the flows amongst them was considered, 164 but was rejected because: 166 o It did not introduce as much entropy as the load balance label 167 method. 169 o It required additional pseudowires to be set up and maintained. 171 1.2. Flow Label 173 An additional label is interposed between the pseudowire label and 174 the control word, or if the control word is not present, between the 175 pseudowire label and the pseudowire payload. This additional label 176 is called the flow label. Indivisible flows within the pseudowire 177 MUST be mapped to the same flow label by the ingress PE. The flow 178 label stimulates the correct ECMP load balancing behaviour in the 179 packet switched network (PSN). On receipt of the pseudowire packet 180 at the egress PE (which knows this additional label is present) the 181 flow label is discarded without processing. 183 Note that the flow label MUST NOT be an MPLS reserved label (values 184 in the range 0..15) [RFC3032], but is otherwise unconstrained by the 185 protocol. 187 Considerations of the TTL value are described in the Security section 188 of this document. The flow label can never become the top label in 189 normal operation, and hence the TTL in the flow label is never used 190 to determine whether the packet should be discarded due to TTL 191 expiry. Therefore there are no lower restrictions on the TTL value. 193 2. Native Service Processing Function 195 The Native Service Processing (NSP) function [RFC3985] is a component 196 of a PE that has knowledge of the structure of the emulated service 197 and is able to take action on the service outside the scope of the 198 pseudowire. In this case it is required that the NSP in the ingress 199 PE identify flows, or groups of flows within the service, and 200 indicate the flow (group) identity of each packet as it is passed to 201 the pseudowire forwarder. As an example, where the PW type is an 202 Ethernet, the NSP might parse the ingress Ethernet traffic and 203 consider all of the IP traffic. This traffic could then be 204 categorised into flows by considering all traffic with the same 205 source and destination address pair to be a single indivisible flow. 206 Since this is an NSP function, by definition, the method used to 207 identify a flow is outside the scope of the pseudowire design. 208 Similarly, since the NSP is internal to the PE, the method of flow 209 indication to the pseudowire forwarder is outside the scope of this 210 document. 212 3. Pseudowire Forwarder 214 The pseudowire forwarder must be provided with a method of mapping 215 flows to load balanced paths. 217 The forwarder must generate a label for the flow or group of flows. 218 How the load balance label values are determined is outside the scope 219 of this document, however the load balance label allocated to a flow 220 MUST NOT be an MPLS reserved label and SHOULD remain constant for the 221 life of the flow. It is recommended that the method chosen to 222 generate the load balancing labels introduces a high degree of 223 entropy in their values, to maximise the entropy presented to the 224 ECMP path selection mechanism in the LSRs in the PSN, and hence 225 distribute the flows as evenly as possible over the available PSN 226 ECMP paths. The forwarder at the ingress PE prepends the pseudowire 227 control word (if applicable), and then pushes the flow label, 228 followed by the pseudowire label. 230 The forwarder at the egress PE uses the pseudowire label to identify 231 the pseudowire. From the context associated with the pseudowire 232 label, the egress PE can determine whether a flow label is present. 233 If a flow label is present, the label is discarded. 235 All other pseudowire forwarding operations are unmodified by the 236 inclusion of the flow label. 238 3.1. Encapsulation 240 The PWE3 Protocol Stack Reference Model modified to include flow 241 label is shown in Figure 1 below 243 +-------------+ +-------------+ 244 | Emulated | | Emulated | 245 | Ethernet | | Ethernet | 246 | (including | Emulated Service | (including | 247 | VLAN) |<==============================>| VLAN) | 248 | Services | | Services | 249 +-------------+ +-------------+ 250 | Flow | | Flow | 251 +-------------+ Pseudowire +-------------+ 252 |Demultiplexer|<==============================>|Demultiplexer| 253 +-------------+ +-------------+ 254 | PSN | PSN Tunnel | PSN | 255 | MPLS |<==============================>| MPLS | 256 +-------------+ +-------------+ 257 | Physical | | Physical | 258 +-----+-------+ +-----+-------+ 260 Figure 1: PWE3 Protocol Stack Reference Model 262 The encapsulation of a pseudowire with a flow label is shown in 263 Figure 2 below 264 +-------------------------------+ 265 | | 266 | Payload | 267 | | n octets 268 | | 269 +-------------------------------+ 270 | Optional Control Word | 4 octets 271 +-------------------------------+ 272 | Flow label | 4 octets 273 +-------------------------------+ 274 | PW label | 4 octets 275 +-------------------------------+ 276 | MPLS Tunnel label(s) | n*4 octets (four octets per label) 277 +-------------------------------+ 279 Figure 2: Encapsulation of a pseudowire with a pseudowire load 280 balancing label 282 4. Signaling the Presence of the Flow Label 284 When using the signalling procedures in [RFC4447], a Pseudowire 285 Interface Parameter Flow Label Sub-TLV (FL Sub-TLV) type is used to 286 synchronise the flow label states between the ingress and egress PEs. 288 The absence of a FL Sub-TLV indicates that the PE is unable process 289 flow labels. A PE that is using PW signalling and that does not send 290 a FL Sub-TLV MUST NOT include a flow label in the PW packet. A PE 291 that is using PW signalling and which does not receive a FL Sub-TLV 292 from its peer MUST NOT include a flow label in the PW packet. This 293 preserves backwards compatibility with existing PW specifications. 295 A PE that wishes to send a flow label in a PW packet includes in its 296 label mapping message a FL Sub-TLV with T = 1 (see Section 4.1). 298 A PE that is able to receive a flow label includes in its label 299 mapping message a FL Sub-TLV with R = 1 (see Section 4.1). 301 A PE that receives a label mapping message a FL Sub-TLV with R = 0 302 MUST NOT include a flow label in the PW packet. 304 Thus a PE sending a FL Sub-TLV with T = 1 and receiving a FL Sub-TLV 305 with R = 1 MUST include a flow label in the PW packet. Under all 306 other combinations of FL Sub-TLV signalling a PE MUST NOT include a 307 flow label in the PW packet. 309 The signalling procedures in [RFC4447] state that "Processing of the 310 interface parameters should continue when unknown interface 311 parameters are encountered, and they MUST be silently ignored." The 312 signalling procedure described here is therefore backwards compatible 313 with existing implementations. 315 If PWE3 signalling [RFC4447] is not in use for a pseudowire, then 316 whether the flow label is used MUST be identically provisioned in 317 both PEs at the pseudowire endpoints. If there is no provisioning 318 support for this option, the default behaviour is not to include the 319 flow label. 321 Note that what is signalled is the desire to include the flow label 322 in the label stack. The value of the label is a local matter for the 323 ingress PE, and the label value itself is not signalled. 325 4.1. Structure of Flow Label Sub-TLV 327 The structure of the flow label TLV is shown in Figure 3. 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 | FL | Length |T|R| Reserved | 333 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 335 Figure 3: Flow Label Sub-TLV 337 Where: 339 o FL is the flow label sub-TLV identifier assigned by IANA. 341 o Length is the length of the TLV in octets and is 4. 343 o When T=1 the PE is requesting the ability to send a PW packet that 344 includes a flow label. When T= 0, the PE is indicating that it 345 will not send a PW packet containing a flow label. 347 o When R=1 the PE is able to receive a PW packet with a flow label 348 present. When R=0 the PE is unable to receive a PW packet with 349 the flow label present. 351 o Reserved bits MUST be zero on transmit and MUST be ignored on 352 receive. 354 5. Multi-Segment Pseudowires 356 The flow label mechanism described in this document works on multi- 357 segment PWs without requiring modification to the Switching PEs 358 (S-PEs). This is because the flow label is transparent to the label 359 swap operation, and because interface parameter Sub-TLV signalling is 360 transitive. 362 6. OAM 364 The following OAM considerations apply to this method of load 365 balancing. 367 Where the OAM is only to be used to perform a basic test that the 368 pseudowires have been configured at the PEs, VCCV [RFC5085] messages 369 may be sent using any load balance pseudowire path, i.e. using any 370 value for the flow label. 372 Where it is required to verify that a pseudowire is fully functional 373 for all flows, VCCV [RFC5085] connection verification message MUST be 374 sent over each ECMP path to the pseudowire egress PE. This problem 375 is difficult to solve and scales poorly. We believe that this 376 problem is addressed by the following two methods: 378 1. If a failure occurs within the PSN, this failure will normally be 379 detected by the PSN's Interior Gateway protocol (IGP) link/node 380 failure detection mechanism (loss of light, bidirectional 381 forwarding detection [RFC5880] or IGP hello detection), and the 382 IGP convergence will naturally modify the ECMP set of network 383 paths between the Ingress and Egress PE's. Hence the PW is only 384 impacted during the normal IGP convergence time. 386 2. If the failure is related to the individual corruption of an 387 Label Forwarding Information database (LFIB) entry in a router, 388 then only the network path using that specific entry is impacted. 389 If the PW is load balanced over multiple network paths, then this 390 failure can only be detected if, by chance, the transported OAM 391 flow is mapped onto the impacted network path, or all paths are 392 tested. This type of error may be better solved be solved by 393 other means such as LSP self test [I-D.ietf-mpls-lsr-self-test]. 395 To troubleshoot the MPLS PSN, including multiple paths, the 396 techniques described in [RFC4378] and [RFC4379] can be used. 398 Where the pseudowire OAM is carried out of band (VCCV Type 2) 399 [RFC5085] it is necessary to insert an "MPLS Router Alert Label" in 400 the label stack. The resultant label stack is a follows: 402 +-------------------------------+ 403 | | 404 | Payload | 405 | | n octets 406 | | 407 +-------------------------------+ 408 | Optional Control Word | 4 octets 409 +-------------------------------+ 410 | Flow label | 4 octets 411 +-------------------------------+ 412 | PW label | 4 octets 413 +-------------------------------+ 414 | Router Alert label | 4 octets 415 +-------------------------------+ 416 | MPLS Tunnel label(s) | n*4 octets (four octets per label) 417 +-------------------------------+ 419 Figure 4: Use of Router Alert LAbel 421 7. Applicability of FAT PWs 423 A node within the PSN is not able to perform deep-packet-inspection 424 (DPI) of the PW as the PW technology is not self-describing: the 425 structure of the PW payload is only known to the ingress and egress 426 PE devices. The method proposed in this document provides a 427 statistical mitigation of the problem of load balance in those cased 428 where a PE is able to discern flows embedded in the traffic received 429 on the attachment circuit. 431 The methods describe in this document are transparent to the PSN and 432 as such do not require any new capability from the PSN. 434 The requirement to load-balance over multiple PSN paths occurs when 435 the ratio between the PW access speed and the PSN's core link 436 bandwidth is large (e.g. >= 10%). ATM and FR are unlikely to meet 437 this property. Ethernet may have this property, and for that reason 438 this document focuses on Ethernet. Applications for other high- 439 access-bandwidth PW's (e.g. Fibre Channel) may be defined in the 440 future. 442 This design applies to MPLS pseudowires where it is meaningful to de- 443 construct the packets presented to the ingress PE into flows. The 444 mechanism described in this document promotes the distribution of 445 flows within the pseudowire over different network paths. This in 446 turn means that whilst packets within a flow are delivered in order 447 (subject to normal IP delivery perturbations due to topology 448 variation), order is not maintained amongst packets of different 449 flows. It is not proposed to associate a different sequence number 450 with each flow. If sequence number support is required this 451 mechanism is not applicable. 453 Where it is known that the traffic carried by the Ethernet pseudowire 454 is IP the method of identifying the flows are well known and can be 455 applied. Such methods typically include hashing on the source and 456 destination addresses, the protocol ID and higher-layer flow- 457 dependent fields such as TCP/UDP ports, L2TPv3 Session ID's etc. 459 Where it is known that the traffic carried by the Ethernet pseudowire 460 is non-IP, techniques used for link bundling between Ethernet 461 switches may be reused. In this case however the latency 462 distribution would be larger than is found in the link bundle case. 463 The acceptability of the increased latency is for further study. Of 464 particular importance the Ethernet control frames SHOULD always be 465 mapped to the same PSN path to ensure in-order delivery. 467 7.1. Equal Cost Multiple Paths 469 ECMP in packet switched networks is statistical in nature. The 470 mapping of flows to a particular path does not take into account the 471 bandwidth of the flow being mapped or the current bandwidth usage of 472 the members of the ECMP set. This simplification works well when the 473 distribution of flows is evenly spread over the ECMP set and there 474 are a large number of flows that have low bandwidth relative to the 475 paths. The random allocation of a flow to a path provides a good 476 approximation to an even spread of flows, provided that polarisation 477 effects are avoided. The method proposed in this document has the 478 same statistical properties as an IP PSN. 480 ECMP is a load-sharing mechanism that is based on sharing the load 481 over a number of layer 3 paths through the PSN. Often however 482 multiple links exist between a pair of LSRs that are considered by 483 the IGP to be a single link. These are known as link bundles. The 484 mechanism described in this document can also be used to distribute 485 the flows within a pseudowire over the members of the link bundle by 486 using the flow label value to identify candidate flows. How that 487 mapping takes place is outside the scope of this specification. 488 Similar considerations apply to link aggregation groups. 490 In the ECMP case and the link bundling case the NSP may attempt to 491 take bandwidth into consideration when allocating groups of flows to 492 a common path. That is permitted, but it must be borne in mind that 493 the semantics of a label stack entry (LSE) as defined by [RFC3032] 494 cannot be modified, the value of the flow label cannot be modified at 495 any point on the LSP, and the interpretation of bit patterns in, or 496 values of, the flow label by an LSR are undefined. 498 A different type of load balancing is the desire to carry a 499 pseudowire over a set of PSN links in which the bandwidth of members 500 of the link set is less than the bandwidth of the pseudowire. This 501 problem is addressed in [I-D.stein-pwe3-pwbonding]. Such a mechanism 502 can be considered complementary to this mechanism. 504 7.2. Link Aggregation Groups 506 A Link Aggregation Group (LAG) is used to bond together several 507 physical circuits between two adjacent nodes so they appear to 508 higher-layer protocols as a single, higher bandwidth "virtual" pipe. 509 These may co-exist in various parts of a given network. An advantage 510 of LAGs is that they reduce the number of routing and signalling 511 protocol adjacencies between devices, reducing control plane 512 processing overhead. As with ECMP, the key problem related to LAGs 513 is that due to inefficiencies in LAG load-distribution algorithms, a 514 particular component of a LAG may experience congestion. The 515 mechanism proposed here may be able to assist in producing a more 516 uniform flow distribution. 518 The same considerations requiring a flow to go over a single member 519 of an ECMP path set apply to a member of a LAG. 521 7.3. Multiple RSVP-TE Paths 523 In some networks it is desirable for a Label Edge Router (LER) to be 524 able to load balance a PW across multiple RSVP-TE tunnels. The flow 525 label mechanism described in this document may be used to provide the 526 LER with the required flow information, and necessary entropy to 527 provide this type of load balancing. An example of such a case is 528 the of the flow label mechanism in networks using a link bundle with 529 the all ones component [RFC4201]. 531 Methods by which the LER is configured to apply this type of ECMP is 532 outside the scope of this document. 534 7.4. The Single Large Flow Case 536 Clearly the operator should make sure that the service offered using 537 PW technology and the method described in this document does not 538 exceed the maximum planned link capacity, unless it can be guaranteed 539 that it conforms to the Internet traffic profile of a very large 540 number of small flows. 542 If the payload on a PW is made of a single inner flow (i.e. an 543 encrypted connection between two routers), or the flow identifiers 544 are too deeply buried in the packet, then the functionality described 545 in this document does not give any benefits, though neither does it 546 cause harm relative to the existing situation. The most common case 547 where a single flow dominated the traffic on a PW is when it is used 548 to transport enterprise traffic. Enterprise traffic may well consist 549 of a large single TCP flows, or encrypted flows that cannot be 550 handled by the methods described in this document. 552 An operator has six options under these circumstances: 554 1. The operator can do nothing and the system will work as it does 555 without the flow label. 557 2. The operator can make the customer aware that the service 558 offering has a restriction on flow bandwidth and police flows to 559 that restriction. This would allow customers offering multiple 560 flows to use a larger fraction their access bandwidth, whilst 561 preventing an single flow from consuming a fraction of internal 562 link bandwidth that the operator considered excessive. 564 3. The operator could configure the ingress PE to assign a constant 565 flow label to all high bandwidth flows so that only one path was 566 affected by these flows, 568 4. The operator could configure the ingress PE to assign a random 569 flow label to all high bandwidth flows so as to minimise the 570 disruption to the network as a cost of out of order traffic to 571 the user. 573 5. The operator could configure the ingress to assign a label of 574 special significance (such as a reserved label) to all high 575 bandwidth flows so that some other action (not specified in this 576 document) could be taken on the flow. 578 The issues described above are mitigated by the following two 579 factors: 581 o Firstly, the customer of a high-bandwidth PW service has an 582 incentive to get the best transport service because an inefficient 583 use of the PSN leads to jitter and eventually to loss to the PW's 584 payload. 586 o Secondly, the customer is usually able to tailor their 587 applications to generate many flows in the PSN. A well-known 588 example is massive data transport between servers which use many 589 parallel TCP sessions. This same technique can be used by any 590 transport protocol: multiple UDP ports, multiple L2TPv3 Session 591 ID's, multiple GRE keys may be used to decompose a large flow into 592 smaller components. This approach may be applied to IPsec 593 [RFC4301] where multiple Security Parameters Indexes (SPI's) may 594 be allocated to the same security association. 596 7.5. Applicability to MPLS-TP 598 The MPLS Transport Profile (MPLS-TP) [RFC5654] requirement 44 states 599 that "MPLS-TP MUST support mechanisms that ensure the integrity of 600 the transported customer's service traffic as required by its 601 associated SLA. Loss of integrity may be defined as packet 602 corruption, reordering, or loss during normal network conditions. " 603 The flow aware transport of a PW reorders packets, therefore MUST NOT 604 be deployed in a network conforming to the MPLS-TP unless these 605 integrity requirements specified in the SLA can be satisfied. 607 7.6. Asymmetric Operation 609 The protocol defined in this document supports the asymmetric 610 inclusion of the FAT label. Asymmetric operation can be expected 611 when there is asymmetry in the bandwidth requirements making it 612 unprofitable for one PE to perform the flow classification, or when 613 that PE is otherwise unable to perform the classification but is able 614 to receive flow labeled packet from its peer. Asymmetric operation 615 of the PW may also be required when one PE has a high transmission 616 bandwidth requirement, but has a need to receive the entire PW on a 617 single interface in order to perform a processing operation that 618 requires the context of the complete PW (for example policing of the 619 egress traffic). 621 8. Applicability to MPLS 623 A further application of this technique would be to create a basis 624 for hash diversity without having to peek below the label stack for 625 IP traffic carried over LDP LSPs. Work on the generalisation of this 626 to MPLS has been described in [I-D.kompella-mpls-entropy-label]. 627 This is can be regarded as a complementary, but distinct, approach 628 since although similar consideration may apply to the identification 629 of flows and the allocation of flow label values, the flow labels are 630 imposed by different network components, and the associated 631 signalling mechanisms are different. 633 9. Security Considerations 635 The pseudowire generic security considerations described in [RFC3985] 636 and the security considerations applicable to a specific pseudowire 637 type (for example, in the case of an Ethernet pseudowire [RFC4448] 638 apply. 640 The ingress PE SHOULD take steps to ensure that the load-balance 641 label is not used as a covert channel. 643 It is useful to give consideration to the choice of TTL value in the 644 flow label stack entry [RFC3032]. The flow label is at the bottom of 645 label stack. Therefore, even when penultimate hop popping is 646 employed, it will always be will preceded by the PW label on arrival 647 at the PE. The flow label TTL should therefore never be considered 648 by the forwarder, and hence SHOULD be set to a value of 1. This will 649 prevent the packet being inadvertently forwarded based on the value 650 of the flow label. Note that this may be a departure from 651 considerations that apply to the general MPLS case. 653 10. IANA Considerations 655 IANA is requested to allocate the next available values from the IETF 656 Consensus range in the Pseudowire Interface Parameters Sub-TLV type 657 Registry as a Flow Label indicator. 659 Parameter Length Description 660 ID 662 TBD 4 Flow Label 664 11. Congestion Considerations 666 The congestion considerations applicable to pseudowires as described 667 in [RFC3985] and any additional congestion considerations developed 668 at the time of publication apply to this design. 670 The ability to explicitly configure a PW to leverage the availability 671 of multiple ECMP paths is beneficial to capacity planning as, all 672 other parameters being constant, the statistical multiplexing of a 673 larger number of smaller flows is more efficient than with a smaller 674 number of larger flows. 676 Note that if the classification into flows is only performed on IP 677 packets the behaviour of those flows in the face of congestion will 678 be as already defined by the IETF for packets of that type and no 679 additional congestion processing is required. 681 Where flows that are not IP are classified pseudowire congestion 682 avoidance must be applied to each non-IP load balance group. 684 12. Acknowledgements 686 The authors wish to thank Eric Grey, Kireeti Kompella, Joerg 687 Kuechemann, Wilfried Maas, Luca Martini, Mark Townsley, and Lucy Yong 688 for valuable comments on this document. 690 13. References 692 13.1. Normative References 694 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 695 Requirement Levels", BCP 14, RFC 2119, March 1997. 697 [RFC3032] Rosen, E., Tappan, D., Fedorkow, G., Rekhter, Y., 698 Farinacci, D., Li, T., and A. Conta, "MPLS Label Stack 699 Encoding", RFC 3032, January 2001. 701 [RFC4379] Kompella, K. and G. Swallow, "Detecting Multi-Protocol 702 Label Switched (MPLS) Data Plane Failures", RFC 4379, 703 February 2006. 705 [RFC4385] Bryant, S., Swallow, G., Martini, L., and D. McPherson, 706 "Pseudowire Emulation Edge-to-Edge (PWE3) Control Word for 707 Use over an MPLS PSN", RFC 4385, February 2006. 709 [RFC4447] Martini, L., Rosen, E., El-Aawar, N., Smith, T., and G. 710 Heron, "Pseudowire Setup and Maintenance Using the Label 711 Distribution Protocol (LDP)", RFC 4447, April 2006. 713 [RFC4448] Martini, L., Rosen, E., El-Aawar, N., and G. Heron, 714 "Encapsulation Methods for Transport of Ethernet over MPLS 715 Networks", RFC 4448, April 2006. 717 [RFC4553] Vainshtein, A. and YJ. Stein, "Structure-Agnostic Time 718 Division Multiplexing (TDM) over Packet (SAToP)", 719 RFC 4553, June 2006. 721 [RFC4928] Swallow, G., Bryant, S., and L. Andersson, "Avoiding Equal 722 Cost Multipath Treatment in MPLS Networks", BCP 128, 723 RFC 4928, June 2007. 725 [RFC5085] Nadeau, T. and C. Pignataro, "Pseudowire Virtual Circuit 726 Connectivity Verification (VCCV): A Control Channel for 727 Pseudowires", RFC 5085, December 2007. 729 13.2. Informative References 731 [I-D.ietf-mpls-lsr-self-test] 732 Swallow, G., "Label Switching Router Self-Test", 733 draft-ietf-mpls-lsr-self-test-07 (work in progress), 734 May 2007. 736 [I-D.kompella-mpls-entropy-label] 737 Kompella, K. and S. Amante, "The Use of Entropy Labels in 738 MPLS Forwarding", draft-kompella-mpls-entropy-label-00 739 (work in progress), July 2008. 741 [I-D.stein-pwe3-pwbonding] 742 Stein, Y., Mendelsohn, I., and R. Insler, "PW Bonding", 743 draft-stein-pwe3-pwbonding-01 (work in progress), 744 November 2008. 746 [RFC3985] Bryant, S. and P. Pate, "Pseudo Wire Emulation Edge-to- 747 Edge (PWE3) Architecture", RFC 3985, March 2005. 749 [RFC4201] Kompella, K., Rekhter, Y., and L. Berger, "Link Bundling 750 in MPLS Traffic Engineering (TE)", RFC 4201, October 2005. 752 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 753 Internet Protocol", RFC 4301, December 2005. 755 [RFC4378] Allan, D. and T. Nadeau, "A Framework for Multi-Protocol 756 Label Switching (MPLS) Operations and Management (OAM)", 757 RFC 4378, February 2006. 759 [RFC5286] Atlas, A. and A. Zinin, "Basic Specification for IP Fast 760 Reroute: Loop-Free Alternates", RFC 5286, September 2008. 762 [RFC5654] Niven-Jenkins, B., Brungard, D., Betts, M., Sprecher, N., 763 and S. Ueno, "Requirements of an MPLS Transport Profile", 764 RFC 5654, September 2009. 766 [RFC5880] Katz, D. and D. Ward, "Bidirectional Forwarding Detection 767 (BFD)", RFC 5880, June 2010. 769 Authors' Addresses 771 Stewart Bryant (editor) 772 Cisco Systems 773 250 Longwater Ave 774 Reading RG2 6GB 775 United Kingdom 777 Phone: +44-208-824-8828 778 Email: stbryant@cisco.com 780 Clarence Filsfils 781 Cisco Systems 782 Brussels 783 Belgium 785 Email: cfilsfil@cisco.com 787 Ulrich Drafz 788 Deutsche Telekom 789 Muenster 790 Germany 792 Email: Ulrich.Drafz@t-com.net 794 Vach Kompella 795 Alcatel-Lucent 797 Email: Alcatel-Lucent vach.kompella@alcatel-lucent.com 799 Joe Regan 800 Alcatel-Lucent 802 Email: joe.regan@alcatel-lucent.comRegan 804 Shane Amante 805 Level 3 Communications 807 Email: shane@castlepoint.net