idnits 2.17.1 draft-ietf-mpls-label-encaps-03.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** Cannot find the required boilerplate sections (Copyright, IPR, etc.) in this document. Expected boilerplate is as follows today (2024-04-16) according to https://trustee.ietf.org/license-info : IETF Trust Legal Provisions of 28-dec-2009, Section 6.a: This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79. IETF Trust Legal Provisions of 28-dec-2009, Section 6.b(i), paragraph 2: Copyright (c) 2024 IETF Trust and the persons identified as the document authors. All rights reserved. IETF Trust Legal Provisions of 28-dec-2009, Section 6.b(i), paragraph 3: This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- ** Missing expiration date. The document expiration date should appear on the first and last page. ** The document seems to lack a 1id_guidelines paragraph about Internet-Drafts being working documents. ** The document seems to lack a 1id_guidelines paragraph about the list of current Internet-Drafts. ** The document seems to lack a 1id_guidelines paragraph about the list of Shadow Directories. == No 'Intended status' indicated for this document; assuming Proposed Standard Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) ** The document seems to lack separate sections for Informative/Normative References. All references will be assumed normative when checking for downward references. ** There are 3 instances of too long lines in the document, the longest one being 2 characters in excess of 72. ** The abstract seems to contain references ([1,2]), which it shouldn't. Please replace those with straight textual mentions of the documents in question. Miscellaneous warnings: ---------------------------------------------------------------------------- -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (September 1998) is 9345 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) -- Possible downref: Non-RFC (?) normative reference: ref. '1' -- Possible downref: Non-RFC (?) normative reference: ref. '2' ** Obsolete normative reference: RFC 1885 (ref. '8') (Obsoleted by RFC 2463) ** Obsolete normative reference: RFC 1981 (ref. '9') (Obsoleted by RFC 8201) -- Possible downref: Non-RFC (?) normative reference: ref. '10' Summary: 11 errors (**), 0 flaws (~~), 1 warning (==), 5 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 1 Network Working Group Eric C. Rosen 2 Internet Draft Yakov Rekhter 3 Expiration Date: March 1999 Daniel Tappan 4 Dino Farinacci 5 Guy Fedorkow 6 Cisco Systems, Inc. 8 Tony Li 9 Juniper Networks, Inc. 11 Alex Conta 12 Lucent Technologies 14 September 1998 16 MPLS Label Stack Encoding 18 draft-ietf-mpls-label-encaps-03.txt 20 Status of this Memo 22 This document is an Internet-Draft. Internet-Drafts are working 23 documents of the Internet Engineering Task Force (IETF), its areas, 24 and its working groups. Note that other groups may also distribute 25 working documents as Internet-Drafts. 27 Internet-Drafts are draft documents valid for a maximum of six months 28 and may be updated, replaced, or obsoleted by other documents at any 29 time. It is inappropriate to use Internet-Drafts as reference 30 material or to cite them other than as "work in progress." 32 To view the entire list of current Internet-Drafts, please check the 33 "1id-abstracts.txt" listing contained in the Internet-Drafts Shadow 34 Directories on ftp.is.co.za (Africa), ftp.nordu.net (Northern 35 Europe), ftp.nis.garr.it (Southern Europe), munnari.oz.au (Pacific 36 Rim), ftp.ietf.org (US East Coast), or ftp.isi.edu (US West Coast). 38 Abstract 40 "Multi-Protocol Label Switching (MPLS)" [1,2] requires a set of 41 procedures for augmenting network layer packets with "label stacks", 42 thereby turning them into "labeled packets". Routers which support 43 MPLS are known as "Label Switching Routers", or "LSRs". In order to 44 transmit a labeled packet on a particular data link, an LSR must 45 support an encoding technique which, given a label stack and a 46 network layer packet, produces a labeled packet. This document 47 specifies the encoding to be used by an LSR in order to transmit 48 labeled packets on PPP data links, on LAN data links, and possibly on 49 other data links as well. On some data links, the label at the top 50 of the stack may be encoded in a different manner, but the techniques 51 described here MUST be used to encode the remainder of the label 52 stack. This document also specifies rules and procedures for 53 processing the various fields of the label stack encoding. 55 Table of Contents 57 1 Introduction ........................................... 3 58 1.1 Specification of Requirements .......................... 3 59 2 The Label Stack ........................................ 3 60 2.1 Encoding the Label Stack ............................... 3 61 2.2 Determining the Network Layer Protocol ................. 6 62 2.3 Generating ICMP Messages for Labeled IP Packets ........ 7 63 2.3.1 Tunneling through a Transit Routing Domain ............. 7 64 2.3.2 Tunneling Private Addresses through a Public Backbone .. 8 65 2.4 Processing the Time to Live Field ...................... 8 66 2.4.1 Definitions ............................................ 8 67 2.4.2 Protocol-independent rules ............................. 9 68 2.4.3 IP-dependent rules ..................................... 9 69 2.4.4 Translating Between Different Encapsulations ........... 10 70 3 Fragmentation and Path MTU Discovery ................... 10 71 3.1 Terminology ............................................ 11 72 3.2 Maximum Initially Labeled IP Datagram Size ............. 13 73 3.3 When are Labeled IP Datagrams Too Big? ................. 14 74 3.4 Processing Labeled IPv4 Datagrams which are Too Big .... 14 75 3.5 Processing Labeled IPv6 Datagrams which are Too Big .... 15 76 3.6 Implications with respect to Path MTU Discovery ........ 16 77 4 Transporting Labeled Packets over PPP .................. 17 78 4.1 Introduction ........................................... 17 79 4.2 A PPP Network Control Protocol for MPLS ................ 17 80 4.3 Sending Labeled Packets ................................ 18 81 4.4 Label Switching Control Protocol Configuration Options . 19 82 5 Transporting Labeled Packets over LAN Media ............ 19 83 6 Security Considerations ................................ 19 84 7 Authors' Addresses ..................................... 20 85 8 References ............................................. 21 87 1. Introduction 89 "Multi-Protocol Label Switching (MPLS)" [1,2] requires a set of 90 procedures for augmenting network layer packets with "label stacks", 91 thereby turning them into "labeled packets". Routers which support 92 MPLS are known as "Label Switching Routers", or "LSRs". In order to 93 transmit a labeled packet on a particular data link, an LSR must 94 support an encoding technique which, given a label stack and a 95 network layer packet, produces a labeled packet. 97 This document specifies the encoding to be used by an LSR in order to 98 transmit labeled packets on PPP data links and on LAN data links. 99 The specified encoding may also be useful for other data links as 100 well. 102 This document also specifies rules and procedures for processing the 103 various fields of the label stack encoding. Since MPLS is 104 independent of any particular network layer protocol, the majority of 105 such procedures are also protocol-independent. A few, however, do 106 differ for different protocols. In this document, we specify the 107 protocol-independent procedures, and we specify the protocol- 108 dependent procedures for IPv4 and IPv6. 110 LSRs that are implemented on certain switching devices (such as ATM 111 switches) may use different encoding techniques for encoding the top 112 one or two entries of the label stack. When the label stack has 113 additional entries, however, the encoding technique described in this 114 document MUST be used for the additional label stack entries. 116 1.1. Specification of Requirements 118 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 119 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 120 document are to be interpreted as described in RFC 2119 [3]. 122 2. The Label Stack 124 2.1. Encoding the Label Stack 126 The label stack is represented as a sequence of "label stack 127 entries". Each label stack entry is represented by 4 octets. This 128 is shown in Figure 1. 130 0 1 2 3 131 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 1 132 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Label 133 | Label | Exp |S| TTL | Stack 134 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Entry 136 Label: Label Value, 20 bits 137 Exp: Experimental Use, 3 bits 138 S: Bottom of Stack, 1 bit 139 TTL: Time to Live, 8 bits 141 Figure 1 143 The label stack entries appear AFTER the data link layer headers, but 144 BEFORE any network layer headers. The top of the label stack appears 145 earliest in the packet, and the bottom appears latest. The network 146 layer packet immediately follows the label stack entry which has the 147 S bit set. 149 Each label stack entry is broken down into the following fields: 151 1. Bottom of Stack (S) 153 This bit is set to one for the last entry in the label stack 154 (i.e., for the bottom of the stack), and zero for all other 155 label stack entries. 157 2. Time to Live (TTL) 159 This eight-bit field is used to encode a time-to-live value. 160 The processing of this field is described in section 2.4. 162 3. Experimental Use 164 This three-bit field is reserved for experimental use. 166 4. Label Value 168 This 20-bit field carries the actual value of the Label. 170 When a labeled packet is received, the label value at the top 171 of the stack is looked up. As a result of a successful lookup 172 one learns: 174 (a) the next hop to which the packet is to be forwarded; 176 (b) the operation to be performed on the label stack before 177 forwarding; this operation may be to replace the top 178 label stack entry with another, or to pop an entry off 179 the label stack, or to replace the top label stack entry 180 and then to push one or more additional entries on the 181 label stack. 183 In addition to learning the next hop and the label stack 184 operation, one may also learn the outgoing data link 185 encapsulation, and possibly other information which is needed 186 in order to properly forward the packet. 188 There are several reserved label values: 190 i. A value of 0 represents the "IPv4 Explicit NULL Label". 191 This label value is only legal when it is the sole 192 label stack entry. It indicates that the label stack 193 must be popped, and the forwarding of the packet must 194 then be based on the IPv4 header. 196 ii. A value of 1 represents the "Router Alert Label". This 197 label value is legal anywhere in the label stack except 198 at the bottom. When a received packet contains this 199 label value at the top of the label stack, it is 200 delivered to a local software module for processing. 201 The actual forwarding of the packet is determined by 202 the label beneath it in the stack. However, if the 203 packet is forwarded further, the Router Alert Label 204 should be pushed back onto the label stack before 205 forwarding. The use of this label is analogous to the 206 use of the "Router Alert Option" in IP packets [6]. 207 Since this label cannot occur at the bottom of the 208 stack, it is not associated with a particular network 209 layer protocol. 211 iii. A value of 2 represents the "IPv6 Explicit NULL Label". 212 This label value is only legal when it is the sole 213 label stack entry. It indicates that the label stack 214 must be popped, and the forwarding of the packet must 215 then be based on the IPv6 header. 217 iv. A value of 3 represents the "Implicit NULL Label". 218 This is a label that an LSR may assign and distribute, 219 but which never actually appears in the encapsulation. 220 When an LSR would otherwise replace the label at the 221 top of the stack with a new label, but the new label is 222 "Implicit NULL", the LSR will pop the stack instead of 223 doing the replacement. Although this value may never 224 appear in the encapsulation, it needs to be specified 225 in the Label Distribution Protocol, so a value is 226 reserved. 228 v. Values 4-16 are reserved. 230 2.2. Determining the Network Layer Protocol 232 When the last label is popped from a packet's label stack (resulting 233 in the stack being emptied), further processing of the packet is 234 based on the packet's network layer header. The LSR which pops the 235 last label off the stack must therefore be able to identify the 236 packet's network layer protocol. However, the label stack does not 237 contain any field which explicitly identifies the network layer 238 protocol. This means that the identity of the network layer protocol 239 must be inferable from the value of the label which is popped from 240 the bottom of the stack, possibly along with the contents of the 241 network layer header itself. 243 Therefore, when the first label is pushed onto a network layer 244 packet, either the label must be one which is used ONLY for packets 245 of a particular network layer, or the label must be one which is used 246 ONLY for a specified set of network layer protocols, where packets of 247 the specified network layers can be distinguished by inspection of 248 the network layer header. Furthermore, whenever that label is 249 replaced by another label value during a packet's transit, the new 250 value must also be one which meets the same criteria. If these 251 conditions are not met, the LSR which pops the last label off a 252 packet will not be able to identify the packet's network layer 253 protocol. 255 Adherence to these conditions does not necessarily enable 256 intermediate nodes to identify a packet's network layer protocol. 257 Under ordinary conditions, this is not necessary, but there are error 258 conditions under which it is desirable. For instance, if an 259 intermediate LSR determines that a labeled packet is undeliverable, 260 it may be desirable for that LSR to generate error messages which are 261 specific to the packet's network layer. The only means the 262 intermediate LSR has for identifying the network layer is inspection 263 of the top label and the network layer header. So if intermediate 264 nodes are to be able to generate protocol-specific error messages for 265 labeled packets, all labels in the stack must meet the criteria 266 specified above for labels which appear at the bottom of the stack. 268 If a packet cannot be forwarded for some reason (e.g., it exceeds the 269 data link MTU), and either its network layer protocol cannot be 270 identified, or there are no specified protocol-dependent rules for 271 handling the error condition, then the packet MUST be silently 272 discarded. 274 2.3. Generating ICMP Messages for Labeled IP Packets 276 Section 2.4 and section 3 discuss situations in which it is desirable 277 to generate ICMP messages for labeled IP packets. In order for a 278 particular LSR to be able to generate an ICMP packet and have that 279 packet sent to the source of the IP packet, two conditions must hold: 281 1. it must be possible for that LSR to determine that a particular 282 labeled packet is an IP packet; 284 2. it must be possible for that LSR to route to the packet's IP 285 source address. 287 Condition 1 is discussed in section 2.2. The following two 288 subsections discuss condition 2. However, there will be some cases 289 in which condition 2 does not hold at all, and in these cases it will 290 not be possible to generate the ICMP message. 292 2.3.1. Tunneling through a Transit Routing Domain 294 Suppose one is using MPLS to "tunnel" through a transit routing 295 domain, where the external routes are not leaked into the domain's 296 interior routers. For example, the interior routers may be running 297 OSPF, and may only know how to reach destinations within that OSPF 298 domain. The domain might contain several Autonomous System Border 299 Routers (ASBRs), which talk BGP to each other. However, in this 300 example the routes from BGP are not distributed into OSPF, and the 301 LSRs which are not ASBRs do not run BGP. 303 In this example, only an ASBR will know how to route to the source of 304 some arbitrary packet. If an interior router needs to send an ICMP 305 message to the source of an IP packet, it will not know how to route 306 the ICMP message. 308 One solution is to have one or more of the ASBRs inject "default" 309 into the IGP. (N.B.: this does NOT require that there be a "default" 310 carried by BGP.) This would then ensure that any unlabeled packet 311 which must leave the domain (such as an ICMP packet) gets sent to a 312 router which has full routing information. The routers with full 313 routing information will label the packets before sending them back 314 through the transit domain, so the use of default routing within the 315 transit domain does not cause any loops. 317 This solution only works for packets which have globally unique 318 addresses, and for networks in which all the ASBRs have complete 319 routing information. The next subsection describes a solution which 320 works when these conditions do not hold. 322 2.3.2. Tunneling Private Addresses through a Public Backbone 324 In some cases where MPLS is used to tunnel through a routing domain, 325 it may not be possible to route to the source address of a fragmented 326 packet at all. This would be the case, for example, if the IP 327 addresses carried in the packet were private (i.e., not globally 328 unique) addresses, and MPLS were being used to tunnel those packets 329 through a public backbone. Default routing to an ASBR will not work 330 in this environment. 332 In this environment, in order to send an ICMP message to the source 333 of a packet, one can copy the label stack from the original packet to 334 the ICMP message, and then label switch the ICMP message. This will 335 cause the message to proceed in the direction of the original 336 packet's destination, rather than its source. Unless the message is 337 label switched all the way to the destination host, it will end up, 338 unlabeled, in a router which does know how to route to the source of 339 original packet, at which point the message will be sent in the 340 proper direction. 342 2.4. Processing the Time to Live Field 344 2.4.1. Definitions 346 The "incoming TTL" of a labeled packet is defined to be the value of 347 the TTL field of the top label stack entry when the packet is 348 received. 350 The "outgoing TTL" of a labeled packet is defined to be the larger 351 of: 353 (a) one less than the incoming TTL, 354 (b) zero. 356 2.4.2. Protocol-independent rules 358 If the outgoing TTL of a labeled packet is 0, then the labeled packet 359 MUST NOT be further forwarded; nor may the label stack be stripped 360 off and the packet forwarded as an unlabeled packet. The packet's 361 lifetime in the network is considered to have expired. 363 Depending on the label value in the label stack entry, the packet MAY 364 be simply discarded, or it may be passed to the appropriate 365 "ordinary" network layer for error processing (e.g., for the 366 generation of an ICMP error message, see section 2.3). 368 When a labeled packet is forwarded, the TTL field of the label stack 369 entry at the top of the label stack MUST be set to the outgoing TTL 370 value. 372 Note that the outgoing TTL value is a function solely of the incoming 373 TTL value, and is independent of whether any labels are pushed or 374 popped before forwarding. There is no significance to the value of 375 the TTL field in any label stack entry which is not at the top of the 376 stack. 378 2.4.3. IP-dependent rules 380 We define the "IP TTL" field to be the value of the IPv4 TTL field, 381 or the value of the IPv6 Hop Limit field, whichever is applicable. 383 When an IP packet is first labeled, the TTL field of the label stack 384 entry MUST BE set to the value of the IP TTL field. (If the IP TTL 385 field needs to be decremented, as part of the IP processing, it is 386 assumed that this has already been done.) 388 When a label is popped, and the resulting label stack is empty, then 389 the value of the IP TTL field SHOULD BE replaced with the outgoing 390 TTL value, as defined above. In IPv4 this also requires modification 391 of the IP header checksum. 393 It is recognized that there may be situations where a network 394 administration prefers to decrement the IPv4 TTL by one as it 395 traverses an MPLS domain, instead of decrementing the IPv4 TTL by the 396 number of LSP hops within the domain. 398 2.4.4. Translating Between Different Encapsulations 400 Sometimes an LSR may receive a labeled packet over, say, a label 401 switching controlled ATM (LC-ATM) interface [10], and may need to 402 send it out over a PPP or LAN link. Then the incoming packet will 403 not be received using the encapsulation specified in this document, 404 but the outgoing packet will be sent using the encapsulation 405 specified in this document. 407 In this case, the value of the "incoming TTL" is determined by the 408 procedures used for carrying labeled packets on, e.g., LC-ATM 409 interfaces. TTL processing then proceeds as described above. 411 Sometimes an LSR may receive a labeled packet over a PPP or a LAN 412 link, and may need to send it out, say, an LC-ATM interface. Then 413 the incoming packet will be received using the encapsulation 414 specified in this document, but the outgoing packet will not be sent 415 using the encapsulation specified in this document. In this case, 416 the procedure for carrying the value of the "outgoing TTL" is 417 determined by the procedures used for carrying labeled packets on, 418 e.g., LC-ATM interfaces. 420 3. Fragmentation and Path MTU Discovery 422 Just as it is possible to receive an unlabeled IP datagram which is 423 too large to be transmitted on its output link, it is possible to 424 receive a labeled packet which is too large to be transmitted on its 425 output link. 427 It is also possible that a received packet (labeled or unlabeled) 428 which was originally small enough to be transmitted on that link 429 becomes too large by virtue of having one or more additional labels 430 pushed onto its label stack. In label switching, a packet may grow 431 in size if additional labels get pushed on. Thus if one receives a 432 labeled packet with a 1500-byte frame payload, and pushes on an 433 additional label, one needs to forward it as frame with a 1504-byte 434 payload. 436 This section specifies the rules for processing labeled packets which 437 are "too large". In particular, it provides rules which ensure that 438 hosts implementing Path MTU Discovery [5], and hosts using IPv6 439 [8,9], will be able to generate IP datagrams that do not need 440 fragmentation, even if those datagrams get labeled as they traverse 441 the network. 443 In general, IPv4 hosts which do not implement Path MTU Discovery [5] 444 send IP datagrams which contain no more than 576 bytes. Since the 445 MTUs in use on most data links today are 1500 bytes or more, the 446 probability that such datagrams will need to get fragmented, even if 447 they get labeled, is very small. 449 Some hosts that do not implement Path MTU Discovery [5] will generate 450 IP datagrams containing 1500 bytes, as long as the IP Source and 451 Destination addresses are on the same subnet. These datagrams will 452 not pass through routers, and hence will not get fragmented. 454 Unfortunately, some hosts will generate IP datagrams containing 1500 455 bytes, as long the IP Source and Destination addresses do not have 456 the same classful network number. This is the one case in which 457 there is any risk of fragmentation when such datagrams get labeled. 458 (Even so, fragmentation is not likely unless the packet must traverse 459 an ethernet of some sort between the time it first gets labeled and 460 the time it gets unlabeled.) 462 This document specifies procedures which allow one to configure the 463 network so that large datagrams from hosts which do not implement 464 Path MTU Discovery get fragmented just once, when they are first 465 labeled. These procedures make it possible (assuming suitable 466 configuration) to avoid any need to fragment packets which have 467 already been labeled. 469 3.1. Terminology 471 With respect to a particular data link, we can use the following 472 terms: 474 - Frame Payload: 476 The contents of a data link frame, excluding any data link layer 477 headers or trailers (e.g., MAC headers, LLC headers, 802.1Q 478 headers, PPP header, frame check sequences, etc.). 480 When a frame is carrying an unlabeled IP datagram, the Frame 481 Payload is just the IP datagram itself. When a frame is carrying 482 a labeled IP datagram, the Frame Payload consists of the label 483 stack entries and the IP datagram. 485 - Conventional Maximum Frame Payload Size: 487 The maximum Frame Payload size allowed by data link standards. 488 For example, the Conventional Maximum Frame Payload Size for 489 ethernet is 1500 bytes. 491 - True Maximum Frame Payload Size: 493 The maximum size frame payload which can be sent and received 494 properly by the interface hardware attached to the data link. 496 On ethernet and 802.3 networks, it is believed that the True 497 Maximum Frame Payload Size is 4-8 bytes larger than the 498 Conventional Maximum Frame Payload Size (as long as neither an 499 802.1Q header nor an 802.1p header is present, and as long as 500 neither can be added by a switch or bridge while a packet is in 501 transit to its next hop). For example, it is believed that most 502 ethernet equipment could correctly send and receive packets 503 carrying a payload of 1504 or perhaps even 1508 bytes, at least, 504 as long as the ethernet header does not have an 802.1Q or 802.1p 505 field. 507 On PPP links, the True Maximum Frame Payload Size may be 508 virtually unbounded. 510 - Effective Maximum Frame Payload Size for Labeled Packets: 512 This is either the Conventional Maximum Frame Payload Size or the 513 True Maximum Frame Payload Size, depending on the capabilities of 514 the equipment on the data link and the size of the data link 515 header being used. 517 - Initially Labeled IP Datagram: 519 Suppose that an unlabeled IP datagram is received at a particular 520 LSR, and that the the LSR pushes on a label before forwarding the 521 datagram. Such a datagram will be called an Initially Labeled IP 522 Datagram at that LSR. 524 - Previously Labeled IP Datagram: 526 An IP datagram which had already been labeled before it was 527 received by a particular LSR. 529 3.2. Maximum Initially Labeled IP Datagram Size 531 Every LSR which is capable of 533 (a) receiving an unlabeled IP datagram, 534 (b) adding a label stack to the datagram, and 535 (c) forwarding the resulting labeled packet, 537 SHOULD support a configuration parameter known as the "Maximum 538 Initially Labeled IP Datagram Size", which can be set to a non- 539 negative value. 541 If this configuration parameter is set to zero, it has no effect. 543 If it is set to a positive value, it is used in the following way. 544 If: 545 (a) an unlabeled IP datagram is received, and 546 (b) that datagram does not have the DF bit set in its IP header, 547 and 548 (c) that datagram needs to be labeled before being forwarded, and 549 (d) the size of the datagram (before labeling) exceeds the value 550 of the parameter, 551 then 552 (a) the datagram must be broken into fragments, each of whose size 553 is no greater than the value of the parameter, and 554 (b) each fragment must be labeled and then forwarded. 556 For example, if this configuration parameter is set to a value of 557 1488, then any unlabeled IP datagram containing more than 1488 bytes 558 will be fragmented before being labeled. Each fragment will be 559 capable of being carried on a 1500-byte data link, without further 560 fragmentation, even if as many as three labels are pushed onto its 561 label stack. 563 In other words, setting this parameter to a non-zero value allows one 564 to eliminate all fragmentation of Previously Labeled IP Datagrams, 565 but it may cause some unnecessary fragmentation of Initially Labeled 566 IP Datagrams. 568 Note that the setting of this parameter does not affect the 569 processing of IP datagrams that have the DF bit set; hence the result 570 of Path MTU discovery is unaffected by the setting of this parameter. 572 3.3. When are Labeled IP Datagrams Too Big? 574 A labeled IP datagram whose size exceeds the Conventional Maximum 575 Frame Payload Size of the data link over which it is to be forwarded 576 MAY be considered to be "too big". 578 A labeled IP datagram whose size exceeds the True Maximum Frame 579 Payload Size of the data link over which it is to be forwarded MUST 580 be considered to be "too big". 582 A labeled IP datagram which is not "too big" MUST be transmitted 583 without fragmentation. 585 3.4. Processing Labeled IPv4 Datagrams which are Too Big 587 If a labeled IPv4 datagram is "too big", and the DF bit is not set in 588 its IP header, then the LSR MAY silently discard the datagram. 590 Note that discarding such datagrams is a sensible procedure only if 591 the "Maximum Initially Labeled IP Datagram Size" is set to a non-zero 592 value in every LSR in the network which is capable of adding a label 593 stack to an unlabeled IP datagram. 595 If the LSR chooses not to discard a labeled IPv4 datagram which is 596 too big, or if the DF bit is set in that datagram, then it MUST 597 execute the following algorithm: 599 1. Strip off the label stack entries to obtain the IP datagram. 601 2. Let N be the number of bytes in the label stack (i.e, 4 times 602 the number of label stack entries). 604 3. If the IP datagram does NOT have the "Don't Fragment" bit set 605 in its IP header: 607 a. convert it into fragments, each of which MUST be at least 608 N bytes less than the Effective Maximum Frame Payload 609 Size. 611 b. Prepend each fragment with the same label header that 612 would have been on the original datagram had 613 fragmentation not been necessary. 615 c. Forward the fragments 617 4. If the IP datagram has the "Don't Fragment" bit set in its IP 618 header: 620 a. the datagram MUST NOT be forwarded 622 b. Create an ICMP Destination Unreachable Message: 624 i. set its Code field [4] to "Fragmentation Required 625 and DF Set", 627 ii. set its Next-Hop MTU field [5] to the difference 628 between the Effective Maximum Frame Payload Size 629 and the value of N 631 c. If possible, transmit the ICMP Destination Unreachable 632 Message to the source of the of the discarded datagram. 634 3.5. Processing Labeled IPv6 Datagrams which are Too Big 636 To process a labeled IPv6 datagram which is too big, an LSR MUST 637 execute the following algorithm: 639 1. Strip off the label stack entries to obtain the IP datagram. 641 2. Let N be the number of bytes in the label stack (i.e, 4 times 642 the number of label stack entries). 644 3. If the IP datagram contains more than 1280 bytes (not counting 645 the label stack entries), then: 647 a. Create an ICMP Packet Too Big Message, and set its Next- 648 Hop MTU field to the difference between the Effective 649 Maximum Frame Payload Size and the value of N 651 b. If possible, transmit the ICMP Packet Too Big Message to 652 the source of the datagram. 654 c. discard the labeled IPv6 datagram. 656 4. If the IP datagram is not larger than 1280 octets, then 658 a. Convert it into fragments, each of which MUST be at least 659 N bytes less than the Effective Maximum Frame Payload 660 Size. 662 b. Prepend each fragment with the same label header that 663 would have been on the original datagram had 664 fragmentation not been necessary. 666 c. Forward the fragments. 668 Reassembly of the fragments will be done at the destination 669 host. 671 3.6. Implications with respect to Path MTU Discovery 673 The procedures described above for handling datagrams which have the 674 DF bit set, but which are "too large", have an impact on the Path MTU 675 Discovery procedures of RFC 1191 [5]. Hosts which implement these 676 procedures will discover an MTU which is small enough to allow n 677 labels to be pushed on the datagrams, without need for fragmentation, 678 where n is the number of labels that actually get pushed on along the 679 path currently in use. 681 In other words, datagrams from hosts that use Path MTU Discovery will 682 never need to be fragmented due to the need to put on a label header, 683 or to add new labels to an existing label header. (Also, datagrams 684 from hosts that use Path MTU Discovery generally have the DF bit set, 685 and so will never get fragmented anyway.) 687 Note that Path MTU Discovery will only work properly if, at the point 688 where a labeled IP Datagram's fragmentation needs to occur, it is 689 possible to cause an ICMP Destination Unreachable message to be 690 routed to the packet's source address. See section 2.3. 692 If it is not possible to forward an ICMP message from within an MPLS 693 "tunnel" to a packet's source address, but the network configuration 694 makes it possible for the LSR at the transmitting end of the tunnel 695 to receive packets that must go through the tunnel, but are too large 696 to pass through the tunnel unfragmented, then: 698 - The LSR at the transmitting end of the tunnel MUST be able to 699 determine the MTU of the tunnel as a whole. It MAY do this by 700 sending packets through the tunnel to the tunnel's receiving 701 endpoint, and performing Path MTU Discovery with those packets. 703 - Any time the transmitting endpoint of the tunnel needs to send a 704 packet into the tunnel, and that packet has the DF bit set, and 705 it exceeds the tunnel MTU, the transmitting endpoint of the 706 tunnel MUST send the ICMP Destination Unreachable message to the 707 source, with code "Fragmentation Required and DF Set", and the 708 Next-Hop MTU Field set as described above. 710 4. Transporting Labeled Packets over PPP 712 The Point-to-Point Protocol (PPP) [7] provides a standard method for 713 transporting multi-protocol datagrams over point-to-point links. PPP 714 defines an extensible Link Control Protocol, and proposes a family of 715 Network Control Protocols for establishing and configuring different 716 network-layer protocols. 718 This section defines the Network Control Protocol for establishing 719 and configuring label Switching over PPP. 721 4.1. Introduction 723 PPP has three main components: 725 1. A method for encapsulating multi-protocol datagrams. 727 2. A Link Control Protocol (LCP) for establishing, configuring, 728 and testing the data-link connection. 730 3. A family of Network Control Protocols for establishing and 731 configuring different network-layer protocols. 733 In order to establish communications over a point-to-point link, each 734 end of the PPP link must first send LCP packets to configure and test 735 the data link. After the link has been established and optional 736 facilities have been negotiated as needed by the LCP, PPP must send 737 "MPLS Control Protocol" packets to enable the transmission of labeled 738 packets. Once the "MPLS Control Protocol" has reached the Opened 739 state, labeled packets can be sent over the link. 741 The link will remain configured for communications until explicit LCP 742 or MPLS Control Protocol packets close the link down, or until some 743 external event occurs (an inactivity timer expires or network 744 administrator intervention). 746 4.2. A PPP Network Control Protocol for MPLS 748 The MPLS Control Protocol (MPLSCP) is responsible for enabling and 749 disabling the use of label switching on a PPP link. It uses the same 750 packet exchange mechanism as the Link Control Protocol (LCP). MPLSCP 751 packets may not be exchanged until PPP has reached the Network-Layer 752 Protocol phase. MPLSCP packets received before this phase is reached 753 should be silently discarded. 755 The MPLS Control Protocol is exactly the same as the Link Control 756 Protocol [7] with the following exceptions: 758 1. Frame Modifications 760 The packet may utilize any modifications to the basic frame 761 format which have been negotiated during the Link Establishment 762 phase. 764 2. Data Link Layer Protocol Field 766 Exactly one MPLSCP packet is encapsulated in the PPP 767 Information field, where the PPP Protocol field indicates type 768 hex 8281 (MPLS). 770 3. Code field 772 Only Codes 1 through 7 (Configure-Request, Configure-Ack, 773 Configure-Nak, Configure-Reject, Terminate-Request, Terminate- 774 Ack and Code-Reject) are used. Other Codes should be treated 775 as unrecognized and should result in Code-Rejects. 777 4. Timeouts 779 MPLSCP packets may not be exchanged until PPP has reached the 780 Network-Layer Protocol phase. An implementation should be 781 prepared to wait for Authentication and Link Quality 782 Determination to finish before timing out waiting for a 783 Configure-Ack or other response. It is suggested that an 784 implementation give up only after user intervention or a 785 configurable amount of time. 787 5. Configuration Option Types 789 None. 791 4.3. Sending Labeled Packets 793 Before any labeled packets may be communicated, PPP must reach the 794 Network-Layer Protocol phase, and the MPLS Control Protocol must 795 reach the Opened state. 797 Exactly one labeled packet is encapsulated in the PPP Information 798 field, where the PPP Protocol field indicates either type hex 0281 799 (MPLS Unicast) or type hex 0283 (MPLS Multicast). The maximum length 800 of a labeled packet transmitted over a PPP link is the same as the 801 maximum length of the Information field of a PPP encapsulated packet. 803 The format of the Information field itself is as defined in section 804 2. 806 Note that two codepoints are defined for labeled packets; one for 807 multicast and one for unicast. Once the MPLSCP has reached the 808 Opened state, both label switched multicasts and label switched 809 unicasts can be sent over the PPP link. 811 4.4. Label Switching Control Protocol Configuration Options 813 There are no configuration options. 815 5. Transporting Labeled Packets over LAN Media 817 Exactly one labeled packet is carried in each frame. 819 The label stack entries immediately precede the network layer header, 820 and follow any data link layer headers, including, e.g., any 802.1Q 821 headers that may exist. 823 The ethertype value 8847 hex is used to indicate that a frame is 824 carrying an MPLS unicast packet. 826 The ethertype value 8848 hex is used to indicate that a frame is 827 carrying an MPLS multicast packet. 829 These ethertype values can be used with either the ethernet 830 encapsulation or the 802.3 LLC/SNAP encapsulation to carry labeled 831 packets. 833 6. Security Considerations 835 The MPLS encapsulation that is specified herein does not raise any 836 security issues that are not already present in either the MPLS 837 architecture [1] or in the architecture of the network layer protocol 838 contained within the encapsulation. 840 There are two security considerations inherited from the MPLS 841 architecture which may be pointed out here: 843 - Some routers may implement security procedures which depend on 844 the network layer header being in a fixed place relative to the 845 data link layer header. These procedures will not work when the 846 MPLS encapsulation is used, because that encapsulation is of a 847 variable size. 849 - An MPLS label has its meaning by virtue of an agreement between 850 the LSR that puts the label in the label stack (the "label 851 writer") , and the LSR that interprets that label (the "label 852 reader"). However, the label stack does not provide any means of 853 determining who the label writer was for any particular label. 854 If labeled packets are accepted from untrusted sources, the 855 result may be that packets are routed in an illegitimate manner. 857 7. Authors' Addresses 859 Eric C. Rosen 860 Cisco Systems, Inc. 861 250 Apollo Drive 862 Chelmsford, MA, 01824 863 E-mail: erosen@cisco.com 865 Dan Tappan 866 Cisco Systems, Inc. 867 250 Apollo Drive 868 Chelmsford, MA, 01824 869 E-mail: tappan@cisco.com 871 Dino Farinacci 872 Cisco Systems, Inc. 873 170 Tasman Drive 874 San Jose, CA, 95134 875 E-mail: dino@cisco.com 877 Yakov Rekhter 878 Cisco Systems, Inc. 879 170 Tasman Drive 880 San Jose, CA, 95134 881 E-mail: yakov@cisco.com 883 Guy Fedorkow 884 Cisco Systems, Inc. 885 250 Apollo Drive 886 Chelmsford, MA, 01824 887 E-mail: fedorkow@cisco.com 889 Tony Li 890 Juniper Networks 891 385 Ravendale Dr. 892 Mountain View, CA, 94043 893 E-mail: tli@juniper.net 894 Alex Conta 895 Lucent Technologies 896 300 Baker Avenue 897 Concord, MA, 01742 898 E-mail: aconta@lucent.com 900 8. References 902 [1] Rosen, E., Viswanathan, A., and Callon, R., "Multiprotocol Label 903 Switching Architecture", Work in Progress, July, 1998 905 [2] Callon, R., Doolan, P., Feldman, N., Fredette, A., Swallow, G., 906 Viswanathan, A., "A Framework for Multiprotocol Label Switching", 907 Work in Progress, November 1997. 909 [3] Bradner, S., "Key words for use in RFCs to Indicate Requirement 910 Levels", RFC 2119, BCP 14, March 1997. 912 [4] Postel, J., "Internet Control Message Protocol", RFC 792, 913 September 1981. 915 [5] Mogul, J. and Deering S., "Path MTU Discovery", RFC 1191, 916 November 1990. 918 [6] Katz, D., "IP Router Alert Option", RFC 2113, February 1997. 920 [7] Simpson, W., Editor, "The Point-to-Point Protocol (PPP)", RFC 921 1661, STD 51, July 1994. 923 [8] Conta, A. and Deering, S., "Internet Control Message Protocol 924 (ICMPv6) for the Internet Protocol Version 6 (IPv6) Specification", 925 RFC 1885, December 1995. 927 [9] McCann, J., Deering, S. and Mogul, J., "Path MTU Discovery for IP 928 version 6", RFC 1981, August 1996. 930 [10] Davie, B., Lawrence, J., McCloghrie, K., Rekhter, Y., Rosen, E. 931 and Swallow G., "Use of Label Switching with ATM", Work in Progress, 932 July 1998.