idnits 2.17.1 draft-ietf-tsvwg-rfc6040update-shim-07.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 RFC2661, but the abstract doesn't seem to mention this, which it should. -- The draft header indicates that this document updates RFC3931, but the abstract doesn't seem to mention this, which it should. -- The draft header indicates that this document updates RFC4380, but the abstract doesn't seem to mention this, which it should. -- The draft header indicates that this document updates RFC2784, 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 RFC2661, updated by this document, for RFC5378 checks: 1996-09-19) -- 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 (November 12, 2018) is 1985 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(See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-22) exists of draft-ietf-tsvwg-ecn-encap-guidelines-11 == Outdated reference: A later version (-03) exists of draft-eastlake-sfc-nsh-ecn-support-02 == Outdated reference: A later version (-09) exists of draft-ietf-intarea-gue-06 == Outdated reference: A later version (-16) exists of draft-ietf-nvo3-geneve-08 == Outdated reference: A later version (-13) exists of draft-ietf-nvo3-vxlan-gpe-06 -- Obsolete informational reference (is this intentional?): RFC 5996 (Obsoleted by RFC 7296) -- Obsolete informational reference (is this intentional?): RFC 6830 (Obsoleted by RFC 9300, RFC 9301) -- Obsolete informational reference (is this intentional?): RFC 8229 (Obsoleted by RFC 9329) Summary: 0 errors (**), 0 flaws (~~), 6 warnings (==), 9 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Transport Area Working Group B. Briscoe 3 Internet-Draft Independent 4 Updates: 6040, 2661, 2784, 3931, 4380, November 12, 2018 5 7450 (if approved) 6 Intended status: Standards Track 7 Expires: May 16, 2019 9 Propagating Explicit Congestion Notification Across IP Tunnel Headers 10 Separated by a Shim 11 draft-ietf-tsvwg-rfc6040update-shim-07 13 Abstract 15 RFC 6040 on "Tunnelling of Explicit Congestion Notification" made the 16 rules for propagation of ECN consistent for all forms of IP in IP 17 tunnel. This specification updates RFC 6040 to clarify that its 18 scope includes tunnels where two IP headers are separated by at least 19 one shim header that is not sufficient on its own for wide area 20 packet forwarding. It surveys widely deployed IP tunnelling 21 protocols separated by such shim header(s) and updates the 22 specifications of those that do not mention ECN propagation (L2TPv2, 23 L2TPv3, GRE, Teredo and AMT). This specification also updates RFC 24 6040 with configuration requirements needed to make any legacy tunnel 25 ingress safe. 27 Status of This Memo 29 This Internet-Draft is submitted in full conformance with the 30 provisions of BCP 78 and BCP 79. 32 Internet-Drafts are working documents of the Internet Engineering 33 Task Force (IETF). Note that other groups may also distribute 34 working documents as Internet-Drafts. The list of current Internet- 35 Drafts is at https://datatracker.ietf.org/drafts/current/. 37 Internet-Drafts are draft documents valid for a maximum of six months 38 and may be updated, replaced, or obsoleted by other documents at any 39 time. It is inappropriate to use Internet-Drafts as reference 40 material or to cite them other than as "work in progress." 42 This Internet-Draft will expire on May 16, 2019. 44 Copyright Notice 46 Copyright (c) 2018 IETF Trust and the persons identified as the 47 document authors. All rights reserved. 49 This document is subject to BCP 78 and the IETF Trust's Legal 50 Provisions Relating to IETF Documents 51 (https://trustee.ietf.org/license-info) in effect on the date of 52 publication of this document. Please review these documents 53 carefully, as they describe your rights and restrictions with respect 54 to this document. Code Components extracted from this document must 55 include Simplified BSD License text as described in Section 4.e of 56 the Trust Legal Provisions and are provided without warranty as 57 described in the Simplified BSD License. 59 Table of Contents 61 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 62 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3 63 3. Scope of RFC 6040 . . . . . . . . . . . . . . . . . . . . . . 3 64 3.1. Feasibility of ECN Propagation between Tunnel Headers . . 4 65 3.2. Desirability of ECN Propagation between Tunnel Headers . 5 66 4. Making a non-ECN Tunnel Ingress Safe by Configuration . . . . 5 67 5. IP-in-IP Tunnels with Tightly Coupled Shim Headers . . . . . 7 68 5.1. Specific Updates to Protocols under IETF Change Control . 9 69 5.1.1. L2TP (v2 and v3) ECN Extension . . . . . . . . . . . 9 70 5.1.2. GRE . . . . . . . . . . . . . . . . . . . . . . . . . 12 71 5.1.3. Teredo . . . . . . . . . . . . . . . . . . . . . . . 13 72 5.1.4. AMT . . . . . . . . . . . . . . . . . . . . . . . . . 14 73 6. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 16 74 7. Security Considerations . . . . . . . . . . . . . . . . . . . 16 75 8. Comments Solicited . . . . . . . . . . . . . . . . . . . . . 16 76 9. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 16 77 10. References . . . . . . . . . . . . . . . . . . . . . . . . . 17 78 10.1. Normative References . . . . . . . . . . . . . . . . . . 17 79 10.2. Informative References . . . . . . . . . . . . . . . . . 18 80 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 21 82 1. Introduction 84 RFC 6040 on "Tunnelling of Explicit Congestion Notification" 85 [RFC6040] made the rules for propagation of Explicit Congestion 86 Notification (ECN [RFC3168]) consistent for all forms of IP in IP 87 tunnel. 89 A common pattern for many tunnelling protocols is to encapsulate an 90 inner IP header (v4 or v6) with shim header(s) then an outer IP 91 header (v4 or v6). Some of these shim headers are designed as 92 generic encapsulations, so they do not necessarily directly 93 encapsulate an inner IP header. Instead they can encapsulate headers 94 such as link-layer (L2) protocols that in turn often encapsulate IP. 96 To clear up confusion, this specification clarifies that the scope of 97 RFC 6040 includes any IP-in-IP tunnel, including those with shim 98 header(s) and other encapsulations between the IP headers. Where 99 necessary, it updates the specifications of the relevant 100 encapsulation protocols with the specific text necessary to comply 101 with RFC 6040. 103 This specification also updates RFC 6040 to state how operators ought 104 to configure a legacy tunnel ingress to avoid unsafe system 105 configurations. 107 2. Terminology 109 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 110 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 111 document are to be interpreted as described in RFC 2119 [RFC2119] 112 when, and only when, they appear in all capitals, as shown here. 114 This specification uses the terminology defined in RFC 6040 115 [RFC6040]. 117 3. Scope of RFC 6040 119 In section 1.1 of RFC 6040, its scope is defined as: 121 "...ECN field processing at encapsulation and decapsulation for 122 any IP-in-IP tunnelling, whether IPsec or non-IPsec tunnels. It 123 applies irrespective of whether IPv4 or IPv6 is used for either 124 the inner or outer headers. ..." 126 This was intended to include cases where shim header(s) sit between 127 the IP headers. Many tunnelling implementers have interpreted the 128 scope of RFC 6040 as it was intended, but it is ambiguous. 129 Therefore, this specification updates RFC 6040 by adding the 130 following scoping text after the sentences quoted above: 132 It applies in cases where an outer IP header encapsulates an inner 133 IP header either directly or indirectly by encapsulating other 134 headers that in turn encapsulate (or might encapsulate) an inner 135 IP header. 137 There is another problem with the scope of RFC 6040. Like many IETF 138 specifications, RFC 6040 is written as a specification that 139 implementations can choose to claim compliance with. This means it 140 does not cover two important cases: 142 1. those cases where it is infeasible for an implementation to 143 access an inner IP header when adding or removing an outer IP 144 header; 146 2. those implementations that choose not to propagate ECN between IP 147 headers. 149 However, the ECN field is a non-optional part of the IP header (v4 150 and v6). So any implementation that creates an outer IP header has 151 to give the ECN field some value. There is only one safe value a 152 tunnel ingress can use if it does not know whether the egress 153 supports propagation of the ECN field; it has to clear the ECN field 154 in any outer IP header to 0b00. 156 However, an RFC has no jurisdiction over implementations that choose 157 not to comply with it or cannot comply with it, including all those 158 implementations that pre-dated the RFC. Therefore it would have been 159 unreasonable to add such a requirement to RFC 6040. Nonetheless, to 160 ensure safe propagation of the ECN field over tunnels, it is 161 reasonable to add requirements on operators, to ensure they configure 162 their tunnels safely (where possible). Before stating these 163 configuration requirements in Section 4, the factors that determine 164 whether propagating ECN is feasible or desirable will be briefly 165 introduced. 167 3.1. Feasibility of ECN Propagation between Tunnel Headers 169 In many cases shim header(s) and an outer IP header are always added 170 to (or removed from) an inner IP packet as part of the same 171 procedure. We call this a tightly coupled shim header. Processing 172 the shim and outer together is often necessary because the shim(s) 173 are not sufficient for packet forwarding in their own right; not 174 unless complemented by an outer header. In these cases it will often 175 be feasible for an implementation to propagate the ECN field between 176 the IP headers. 178 In some cases a tunnel adds an outer IP header and a tightly coupled 179 shim header to an inner header that is not an IP header, but that in 180 turn encapsulates an IP header (or might encapsulate an IP header). 181 For instance an inner Ethernet (or other link layer) header might 182 encapsulate an inner IP header as its payload. We call this a 183 tightly coupled shim over an encapsulating header. 185 Digging to arbitrary depths to find an inner IP header within an 186 encapsulation is strictly a layering violation so it cannot be a 187 required behaviour. Nonetheless, some tunnel endpoints already look 188 within a L2 header for an IP header, for instance to map the Diffserv 189 codepoint between an encapsulated IP header and an outer IP header 191 [RFC2983]. In such cases at least, it should be feasible to also 192 (independently) propagate the ECN field between the same IP headers. 193 Thus, access to the ECN field within an encapsulating header can be a 194 useful and benign optimization. The guidelines in section 5 of 195 [I-D.ietf-tsvwg-ecn-encap-guidelines] give the conditions for this 196 layering violation to be benign. 198 3.2. Desirability of ECN Propagation between Tunnel Headers 200 Developers and network operators are encouraged to implement and 201 deploy tunnel endpoints compliant with RFC 6040 (as updated by the 202 present specification) in order to provide the benefits of wider ECN 203 deployment [RFC8087]. Nonetheless, propagation of ECN between IP 204 headers, whether separated by shim headers or not, has to be optional 205 to implement and to use, because: 207 o Legacy implementations of tunnels without any ECN support already 208 exist 210 o A network might be designed so that there is usually no bottleneck 211 within the tunnel 213 o If the tunnel endpoints would have to search within an L2 header 214 to find an encapsulated IP header, it might not be worth the 215 potential performance hit 217 4. Making a non-ECN Tunnel Ingress Safe by Configuration 219 Even when no specific attempt has been made to implement propagation 220 of the ECN field at a tunnel ingress, it ought to be possible for the 221 operator to render a tunnel ingress safe by configuration. The main 222 safety concern is to disable (clear to zero) the ECN capability in 223 the outer IP header at the ingress if the egress of the tunnel does 224 not implement ECN logic to propagate any ECN markings into the packet 225 forwarded beyond the tunnel. Otherwise the non-ECN egress could 226 discard any ECN marking introduced within the tunnel, which would 227 break all the ECN-based control loops that regulate the traffic load 228 over the tunnel. 230 Therefore this specification updates RFC 6040 by inserting the 231 following text at the end of section 4.3: 233 " 234 Whether or not an ingress implementation claims compliance with 235 RFC 6040, RFC 4301 or RFC3168, when the outer tunnel header is IP 236 (v4 or v6), if possible, the operator MUST configure the ingress 237 to zero the outer ECN field in any of the following cases: 239 * if it is known that the tunnel egress does not support 240 propagation of the ECN field (RFC 6040, RFC 4301 or the full 241 functionality mode of RFC 3168) 243 * or if the behaviour of the egress is not known or an egress 244 with unknown behaviour might be dynamically paired with the 245 ingress. 247 * or if an IP header might be encapsulated within a non-IP header 248 that the tunnel ingress is encapsulating, but the ingress does 249 not inspect within the encapsulation. 251 For the avoidance of doubt, the above only concerns the outer IP 252 header. The ingress MUST NOT alter the ECN field of the arriving 253 IP header that will become the inner IP header. 255 In order that the network operator can comply with the above 256 safety rules, even if an implementation of a tunnel ingress does 257 not claim to support RFC 6040, RFC 4301 or the full functionality 258 mode of RFC 3168: 260 * it MUST make propagation of the ECN field between inner and 261 outer IP headers independent of any configuration of Diffserv 262 codepoint propagation; 264 * it SHOULD be able to be configured to zero the outer ECN field. 266 " 268 There might be concern that the above "MUST" makes compliant 269 equipment non-compliant at a stroke. However, any equipment that is 270 still treating the former ToS octet (IPv4) or the former Traffic 271 Class octet (IPv6) as a single 8-bit field is already non-compliant, 272 and has been since 1998 when the upper 6 bits were separated off for 273 the Diffserv field [RFC2474], [RFC3260]. For instance, copying the 274 ECN field as a side-effect of copying the DSCP is a seriously unsafe 275 bug that risks breaking the feedback loops that regulate load on a 276 tunnel. 278 Permanently zeroing the outer ECN field is safe, but it is not 279 sufficient to claim compliance with RFC 6040 because it does not meet 280 the aim of introducing ECN support to tunnels (see Section 4.3 of 281 [RFC6040]). 283 5. IP-in-IP Tunnels with Tightly Coupled Shim Headers 285 There follows a list of specifications of encapsulations with tightly 286 coupled shim header(s), in rough chronological order. The list is 287 confined to standards track or widely deployed protocols. The list 288 is not necessarily exhaustive so, for the avoidance of doubt, the 289 scope of RFC 6040 is defined in Section 3 and is not limited to this 290 list. 292 o PPTP (Point-to-Point Tunneling Protocol) [RFC2637]; 294 o L2TP (Layer 2 Tunnelling Protocol), specifically L2TPv2 [RFC2661] 295 and L2TPv3 [RFC3931], which not only includes all the L2-specific 296 specializations of L2TP, but also derivatives such as the Keyed 297 IPv6 Tunnel [RFC8159]; 299 o GRE (Generic Routing Encapsulation) [RFC2784] and NVGRE (Network 300 Virtualization using GRE) [RFC7637]; 302 o GTP (GPRS Tunnelling Protocol), specifically GTPv1 [GTPv1], GTP v1 303 User Plane [GTPv1-U], GTP v2 Control Plane [GTPv2-C]; 305 o Teredo [RFC4380]; 307 o CAPWAP (Control And Provisioning of Wireless Access Points) 308 [RFC5415]; 310 o LISP (Locator/Identifier Separation Protocol) [RFC6830]; 312 o AMT (Automatic Multicast Tunneling) [RFC7450]; 314 o VXLAN (Virtual eXtensible Local Area Network) [RFC7348] and VXLAN- 315 GPE [I-D.ietf-nvo3-vxlan-gpe]; 317 o The Network Service Header (NSH [RFC8300]) for Service Function 318 Chaining (SFC); 320 o Geneve [I-D.ietf-nvo3-geneve]; 322 o GUE (Generic UDP Encapsulation) [I-D.ietf-intarea-gue]; 324 o Direct tunnelling of an IP packet within a UDP/IP datagram (see 325 Section 3.1.11 of [RFC8085]); 327 o TCP Encapsulation of IKE and IPsec Packets (see Section 12.5 of 328 [RFC8229]). 330 Some of the listed protocols enable encapsulation of a variety of 331 network layer protocols as inner and/or outer. This specification 332 applies in the cases where there is an inner and outer IP header as 333 described in Section 3. Otherwise 334 [I-D.ietf-tsvwg-ecn-encap-guidelines] gives guidance on how to design 335 propagation of ECN into other protocols that might encapsulate IP. 337 Where protocols in the above list need to be updated to specify ECN 338 propagation and they are under IETF change control, update text is 339 given in the following subsections. For those not under IETF 340 control, it is RECOMMENDED that implementations of encapsulation and 341 decapsulation comply with RFC 6040. It is also RECOMMENDED that 342 their specifications are updated to add a requirement to comply with 343 RFC 6040 (as updated by the present document). 345 PPTP is not under the change control of the IETF, but it has been 346 documented in an informational RFC [RFC2637]. However, there is no 347 need for the present specification to update PPTP because L2TP has 348 been developed as a standardized replacement. 350 NVGRE is not under the change control of the IETF, but it has been 351 documented in an informational RFC [RFC7637]. NVGRE is a specific 352 use-case of GRE (it re-purposes the key field from the initial 353 specification of GRE [RFC1701] as a Virtual Subnet ID). Therefore 354 the text that updates GRE in Section 5.1.2 below is also intended to 355 update NVGRE. 357 Although the definition of the various GTP shim headers is under the 358 control of the 3GPP, it is hard to determine whether the 3GPP or the 359 IETF controls standardization of the _process_ of adding both a GTP 360 and an IP header to an inner IP header. Nonetheless, the present 361 specification is provided so that the 3GPP can refer to it from any 362 of its own specifications of GTP and IP header processing. 364 The specification of CAPWAP already specifies RFC 3168 ECN 365 propagation and ECN capability negotiation. Without modification the 366 CAPWAP specification already interworks with the backward compatible 367 updates to RFC 3168 in RFC 6040. 369 LISP made the ECN propagation procedures in RFC 3168 mandatory from 370 the start. RFC 3168 has since been updated by RFC 6040, but the 371 changes are backwards compatible so there is still no need for LISP 372 tunnel endpoints to negotiate their ECN capabilities. 374 VXLAN is not under the change control of the IETF but it has been 375 documented in an informational RFC. In contrast, VXLAN-GPE (Generic 376 Protocol Extension) is being documented under IETF change control. 377 It is RECOMMENDED that VXLAN and VXLAN-GPE implementations comply 378 with RFC 6040 when the VXLAN header is inserted between (or removed 379 from between) IP headers. The authors of any future update to these 380 specifications are encouraged to add a requirement to comply with RFC 381 6040 as updated by the present specification. 383 The Network Service Header (NSH [RFC8300]) has been defined as a 384 shim-based encapsulation to identify the Service Function Path (SFP) 385 in the Service Function Chaining (SFC) architecture [RFC7665]. A 386 proposal has been made for the processing of ECN when handling 387 transport encapsulation [I-D.eastlake-sfc-nsh-ecn-support]. 389 The specifications of Geneve and GUE already refer to RFC 6040 for 390 ECN encapsulation. 392 Section 3.1.11 of the UDP usage guidelines [RFC8085] already explains 393 that a tunnel that encapsulates an IP header directly within a UDP/IP 394 datagram needs to follow RFC 6040 when propagating the ECN field 395 between inner and outer IP headers. The requirements in Section 4 396 update RFC 6040 so, by reference, they automatically update RFC 8085 397 to add the important but previously unstated requirement that, if the 398 UDP tunnel egress does not, or might not, support ECN propagation, a 399 legacy UDP tunnel ingress has to clear the outer IP ECN field to 400 0b00, e.g. by configuration. 402 Section 12.5 of TCP Encapsulation of IKE and IPsec Packets [RFC8229] 403 already recommends the compatibility mode of RFC 6040 in this case, 404 because there is not a one-to-one mapping between inner and outer 405 packets. 407 5.1. Specific Updates to Protocols under IETF Change Control 409 5.1.1. L2TP (v2 and v3) ECN Extension 411 The L2TP terminology used here is defined in [RFC2661] and [RFC3931]. 413 L2TPv3 [RFC3931] is used as a shim header between any packet-switched 414 network (PSN) header (e.g. IPv4, IPv6, MPLS) and many types of layer 415 2 (L2) header. The L2TPv3 shim header encapsulates an L2-specific 416 sub-layer then an L2 header that is likely to contain an inner IP 417 header (v4 or v6). Then this whole stack of headers can be 418 encapsulated optionally within an outer UDP header then an outer PSN 419 header that is typically IP (v4 or v6). 421 L2TPv2 is used as a shim header between any PSN header and a PPP 422 header, which is in turn likely to encapsulate an IP header. 424 Even though these shims are rather fat (particularly in the case of 425 L2TPv3), they still fit the definition of a tightly coupled shim 426 header over an encapsulating header (Section 3.1), because all the 427 headers encapsulating the L2 header are added (or removed) together. 428 L2TPv2 and L2TPv3 are therefore within the scope of RFC 6040, as 429 updated by Section 3 above. 431 L2TP maintainers are RECOMMENDED to implement the ECN extension to 432 L2TPv2 and L2TPv3 defined in Section 5.1.1.2 below, in order to 433 provide the benefits of ECN [RFC8087], whenever a node within an L2TP 434 tunnel becomes the bottleneck for an end-to-end traffic flow. 436 5.1.1.1. Safe Configuration of a 'Non-ECN' Ingress LCCE 438 The following text is appended to both Section 5.3 of [RFC2661] and 439 Section 4.5 of [RFC3931] as an update to the base L2TPv2 and L2TPv3 440 specifications: 442 The operator of an LCCE that does not support the ECN Extension in 443 Section 5.1.1.2 of RFCXXXX MUST follow the configuration 444 requirements in Section 4 of RFCXXXX to ensure it clears the outer 445 IP ECN field to 0b00 when the outer PSN header is IP (v4 or v6). 446 {RFCXXXX refers to the present document so it will need to be 447 inserted by the RFC Editor} 449 In particular, for an LCCE implementation that does not support the 450 ECN Extension, this means that configuration of how it propagates the 451 ECN field between inner and outer IP headers MUST be independent of 452 any configuration of the Diffserv extension of L2TP [RFC3308]. 454 5.1.1.2. ECN Extension for L2TP (v2 or v3) 456 When the outer PSN header and the payload inside the L2 header are 457 both IP (v4 or v6), to comply with RFC 6040, an LCCE will follow the 458 rules for propagation of the ECN field at ingress and egress in 459 Section 4 of RFC 6040 [RFC6040]. 461 Before encapsulating any data packets, RFC 6040 requires an ingress 462 LCCE to check that the egress LCCE supports ECN propagation as 463 defined in RFC 6040 or one of its compatible predecessors ([RFC4301] 464 or the full functionality mode of [RFC3168]). If the egress supports 465 ECN propagation, the ingress LCCE can use the normal mode of 466 encapsulation (copying the ECN field from inner to outer). 467 Otherwise, the ingress LCCE has to use compatibility mode [RFC6040] 468 (clearing the outer IP ECN field to 0b00). 470 An LCCE can determine the remote LCCE's support for ECN either 471 statically (by configuration) or by dynamic discovery during setup of 472 each control connection between the LCCEs, using the Capability AVP 473 defined in Section 5.1.1.2.1 below. 475 Where the outer PSN header is some protocol other than IP that 476 supports ECN, the appropriate ECN propagation specification will need 477 to be followed, e.g. "Explicit Congestion Marking in MPLS" 478 [RFC5129]. Where no specification exists for ECN propagation by a 479 particular PSN, [I-D.ietf-tsvwg-ecn-encap-guidelines] gives general 480 guidance on how to design ECN propagation into a protocol that 481 encapsulates IP. 483 5.1.1.2.1. LCCE Capability AVP for ECN Capability Negotiation 485 The LCCE Capability Attribute-Value Pair (AVP) defined here has 486 Attribute Type ZZ. The Attribute Value field for this AVP is a bit- 487 mask with the following 16-bit format: 489 0 1 490 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 491 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 492 |X X X X X X X X X X X X X X X E| 493 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 495 Figure 1: Value Field for the LCCE Capability Attribute 497 This AVP MAY be present in the following message types: SCCRQ and 498 SCCRP (Start-Control-Connection-Request and Start-Control-Connection- 499 Reply). This AVP MAY be hidden (the H-bit set to 0 or 1) and is 500 optional (M-bit not set). The length (before hiding) of this AVP 501 MUST be 8 octets. The Vendor ID is the IETF Vendor ID of 0. 503 Bit 15 of the Value field of the LCCE Capability AVP is defined as 504 the ECN Capability flag (E). When the ECN Capability flag is set to 505 1, it indicates that the sender supports ECN propagation. When the 506 ECN Capability flag is cleared to zero, or when no LCCE Capabiliy AVP 507 is present, it indicates that the sender does not support ECN 508 propagation. All the other bits are reserved. They MUST be cleared 509 to zero when sent and ignored when received or forwarded. 511 An LCCE initiating a control connection will send a Start-Control- 512 Connection-Request (SCCRQ) containing an LCCE Capability AVP with the 513 ECN Capability flag set to 1. If the tunnel terminator supports ECN, 514 it will return a Start-Control-Connection-Reply (SCCRP) that also 515 includes an LCCE Capability AVP with the ECN Capability flag set to 516 1. Then, for any sessions created by that control connection, both 517 ends of the tunnel can use the normal mode of RFC 6040, i.e. it can 518 copy the IP ECN field from inner to outer when encapsulating data 519 packets. 521 If, on the other hand, the tunnel terminator does not support ECN it 522 will ignore the ECN flag in the LCCE Capability AVP and send an SCCRP 523 to the tunnel initiator without a Capability AVP (or with a 524 Capability AVP but with the ECN Capability flag cleared to zero). 525 The tunnel initiator interprets the absence of the ECN Capability 526 flag in the SCCRP as an indication that the tunnel terminator is 527 incapable of supporting ECN. When encapsulating data packets for any 528 sessions created by that control connection, the tunnel initiator 529 will then use the compatibility mode of RFC 6040 to clear the ECN 530 field of the outer IP header to 0b00. 532 If the tunnel terminator does not support this ECN extension, the 533 network operator is still expected to configure it to comply with the 534 safety provisions set out in Section 5.1.1.1 above, when it acts as 535 an ingress LCCE. 537 5.1.2. GRE 539 The GRE terminology used here is defined in [RFC2784]. GRE is often 540 used as a tightly coupled shim header between IP headers. Sometimes 541 the GRE shim header encapsulates an L2 header, which might in turn 542 encapsulate an IP header. Therefore GRE is within the scope of RFC 543 6040 as updated by Section 3 above. 545 GRE tunnel endpoint maintainers are RECOMMENDED to support [RFC6040] 546 as updated by the present specification, in order to provide the 547 benefits of ECN [RFC8087] whenever a node within a GRE tunnel becomes 548 the bottleneck for an end-to-end IP traffic flow tunnelled over GRE 549 using IP as the delivery protocol (outer header). 551 GRE itself does not support dynamic set-up and configuration of 552 tunnels. However, control plane protocols such as Mobile IPv4 (MIP4) 553 [RFC5944], Mobile IPv6 (MIP6) [RFC6275], Proxy Mobile IP (PMIP) 554 [RFC5845] and IKEv2 [RFC5996] are sometimes used to set up GRE 555 tunnels dynamically. 557 When these control protocols set up IP-in-IP or IPSec tunnels, it is 558 likely that they propagate the ECN field as defined in RFC 6040 or 559 one of its compatible predecessors (RFC 4301 or the full 560 functionality mode of RFC 3168). However, if they use a GRE 561 encapsulation, this presumption is less sound. 563 Therefore, If the outer delivery protocol is IP (v4 or v6) the 564 operator is obliged to follow the safe configuration requirements in 565 Section 4 above. Section 5.1.2.1 below updates the base GRE 566 specification with this requirement, to emphasize its importance. 568 Where the delivery protocol is some protocol other than IP that 569 supports ECN, the appropriate ECN propagation specification will need 570 to be followed, e.g Explicit Congestion Marking in MPLS [RFC5129]. 572 Where no specification exists for ECN propagation by a particular 573 PSN, [I-D.ietf-tsvwg-ecn-encap-guidelines] gives more general 574 guidance on how to propagate ECN to and from protocols that 575 encapsulate IP. 577 5.1.2.1. Safe Configuration of a 'Non-ECN' GRE Ingress 579 The following text is appended to Section 3 of [RFC2784] as an update 580 to the base GRE specification: 582 The operator of a GRE tunnel ingress MUST follow the configuration 583 requirements in Section 4 of RFCXXXX when the outer delivery 584 protocol is IP (v4 or v6). {RFCXXXX refers to the present document 585 so it will need to be inserted by the RFC Editor} 587 5.1.3. Teredo 589 Teredo [RFC4380] provides a way to tunnel IPv6 over an IPv4 network, 590 with a UDP-based shim header between the two. 592 For Teredo tunnel endpoints to provide the benefits of ECN, the 593 Teredo specification would have to be updated to include negotiation 594 of the ECN capability between Teredo tunnel endpoints. Otherwise it 595 would be unsafe for a Teredo tunnel ingress to copy the ECN field to 596 the IPv6 outer. 598 It is believed that current implementations do not support 599 propagation of ECN, but that they do safely zero the ECN field in the 600 outer IPv6 header. However the specification does not mention 601 anything about this. 603 To make existing Teredo deployments safe, it would be possible to add 604 ECN capability negotiation to those that are subject to remote OS 605 update. However, for those implementations not subject to remote OS 606 update, it will not be feasible to require them to be configured 607 correctly, because Teredo tunnel endpoints are generally deployed on 608 hosts. 610 Therefore, until ECN support is added to the specification of Teredo, 611 the only feasible further safety precaution available here is to 612 update the specification of Teredo implementations with the following 613 text, as a new section 5.1.3: 615 "5.1.3 Safe 'Non-ECN' Teredo Encapsulation 617 A Teredo tunnel ingress implementation that does not support ECN 618 propagation as defined in RFC 6040 or one of its compatible 619 predecessors (RFC 4301 or the full functionality mode of RFC 3168) 620 MUST zero the ECN field in the outer IPv6 header." 622 5.1.4. AMT 624 Automatic Multicast Tunneling (AMT [RFC7450]) is a tightly coupled 625 shim header that encapsulates an IP packet and is itself encapsulated 626 within a UDP/IP datagram. Therefore AMT is within the scope of RFC 627 6040 as updated by Section 3 above. 629 AMT tunnel endpoint maintainers are RECOMMENDED to support [RFC6040] 630 as updated by the present specification, in order to provide the 631 benefits of ECN [RFC8087] whenever a node within an AMT tunnel 632 becomes the bottleneck for an IP traffic flow tunnelled over AMT. 634 To comply with RFC 6040, an AMT relay and gateway will follow the 635 rules for propagation of the ECN field at ingress and egress 636 respectively, as described in Section 4 of RFC 6040 [RFC6040]. 638 Before encapsulating any data packets, RFC 6040 requires an ingress 639 AMT relay to check that the egress AMT gateway supports ECN 640 propagation as defined in RFC 6040 or one of its compatible 641 predecessors (RFC 4301 or the full functionality mode of RFC 3168). 642 If the egress gateway supports ECN, the ingress relay can use the 643 normal mode of encapsulation (copying the IP ECN field from inner to 644 outer). Otherwise, the ingress relay has to use compatibility mode, 645 which means it has to clear the outer ECN field to zero [RFC6040]. 647 An AMT tunnel is created dynamically (not manually), so the relay 648 will need to determine the remote gateway's support for ECN using the 649 ECN capability declaration defined in Section 5.1.4.2 below. 651 5.1.4.1. Safe Configuration of a 'Non-ECN' Ingress AMT Relay 653 The following text is appended to Section 4.2.2 of [RFC7450] as an 654 update to the AMT specification: 656 The operator of an AMT relay that does not support RFC 6040 or one 657 of its compatible predecessors (RFC 4301 or the full functionality 658 mode of RFC 3168) MUST follow the configuration requirements in 659 Section 4 of RFCXXXX to ensure it clears the outer IP ECN field to 660 zero. {RFCXXXX refers to the present document so it will need to 661 be inserted by the RFC Editor} 663 5.1.4.2. ECN Capability Declaration of an AMT Gateway 665 0 1 2 3 666 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 1 667 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 668 | V=0 |Type=3 | Reserved |E|P| Reserved | 669 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 670 | Request Nonce | 671 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 673 Figure 2: Updated AMT Request Message Format 675 Bit 14 of the AMT Request Message counting from 0 (or bit 7 of the 676 Reserved field counting from 1) is defined here as the AMT Gateway 677 ECN Capability flag (E), as shown in Figure 2. The definitions of 678 all other fields in the AMT Request Message are unchanged from RFC 679 7450. 681 When the E flag is set to 1, it indicates that the sender of the 682 message supports RFC 6040 ECN propagation. When it is cleared to 683 zero, it indicates the sender of the message does not support RFC 684 6040 ECN propagation. An AMT gateway "that supports RFC 6040 ECN 685 propagation" means one that propagates the ECN field to the forwarded 686 data packet based on the combination of arriving inner and outer ECN 687 fields, as defined in Section 4 of RFC 6040. 689 The other bits of the Reserved field remain reserved. They will 690 continue to be cleared to zero when sent and ignored when either 691 received or forwarded, as specified in Section 5.1.3.3. of RFC 7450. 693 An AMT gateway that does not support RFC 6040 MUST NOT set the E flag 694 of its Request Message to 1. 696 An AMT gateway that supports RFC 6040 ECN propagation MUST set the E 697 flag of its Relay Discovery Message to 1. 699 The action of the corresponding AMT relay that receives a Request 700 message with the E flag set to 1 depends on whether the relay itself 701 supports RFC 6040 ECN propagation: 703 o If the relay supports RFC 6040 ECN propagation, it will store the 704 ECN capability of the gateway along with its address. Then 705 whenever it tunnels datagrams towards this gateway, it MUST use 706 the normal mode of RFC 6040 to propagate the ECN field when 707 encapsulating datagrams (i.e. it copies the IP ECN field from 708 inner to outer). 710 o If the discovered AMT relay does not support RFC 6040 ECN 711 propagation, it will ignore the E flag in the Reserved field, as 712 per section 5.1.3.3. of RFC 7450. 714 If the AMT relay does not support RFC 6040 ECN propagation, the 715 network operator is still expected to configure it to comply with 716 the safety provisions set out in Section 5.1.4.1 above. 718 6. IANA Considerations 720 IANA is requested to assign the following L2TP Control Message 721 Attribute Value Pair: 723 +----------------+----------------+-----------+ 724 | Attribute Type | Description | Reference | 725 +----------------+----------------+-----------+ 726 | ZZ | ECN Capability | RFCXXXX | 727 +----------------+----------------+-----------+ 729 [TO BE REMOVED: This registration should take place at the following 730 location: https://www.iana.org/assignments/l2tp-parameters/l2tp- 731 parameters.xhtml ] 733 7. Security Considerations 735 The Security Considerations in [RFC6040] and 736 [I-D.ietf-tsvwg-ecn-encap-guidelines] apply equally to the scope 737 defined for the present specification. 739 8. Comments Solicited 741 Comments and questions are encouraged and very welcome. They can be 742 addressed to the IETF Transport Area working group mailing list 743 , and/or to the authors. 745 9. Acknowledgements 747 Thanks to Ing-jyh (Inton) Tsang for initial discussions on the need 748 for ECN propagation in L2TP and its applicability. Thanks also to 749 Carlos Pignataro, Tom Herbert, Ignacio Goyret, Alia Atlas, Praveen 750 Balasubramanian, Joe Touch, Mohamed Boucadair, David Black, Jake 751 Holland and Sri Gundavelli for helpful advice and comments. "A 752 Comparison of IPv6-over-IPv4 Tunnel Mechanisms" [RFC7059] helped to 753 identify a number of tunnelling protocols to include within the scope 754 of this document. 756 Bob Briscoe was part-funded by the Research Council of Norway through 757 the TimeIn project. The views expressed here are solely those of the 758 authors. 760 10. References 762 10.1. Normative References 764 [I-D.ietf-tsvwg-ecn-encap-guidelines] 765 Briscoe, B., Kaippallimalil, J., and P. Thaler, 766 "Guidelines for Adding Congestion Notification to 767 Protocols that Encapsulate IP", draft-ietf-tsvwg-ecn- 768 encap-guidelines-11 (work in progress), Novemeber 2018. 770 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 771 Requirement Levels", BCP 14, RFC 2119, 772 DOI 10.17487/RFC2119, March 1997, 773 . 775 [RFC2474] Nichols, K., Blake, S., Baker, F., and D. Black, 776 "Definition of the Differentiated Services Field (DS 777 Field) in the IPv4 and IPv6 Headers", RFC 2474, 778 DOI 10.17487/RFC2474, December 1998, 779 . 781 [RFC2661] Townsley, W., Valencia, A., Rubens, A., Pall, G., Zorn, 782 G., and B. Palter, "Layer Two Tunneling Protocol "L2TP"", 783 RFC 2661, DOI 10.17487/RFC2661, August 1999, 784 . 786 [RFC2784] Farinacci, D., Li, T., Hanks, S., Meyer, D., and P. 787 Traina, "Generic Routing Encapsulation (GRE)", RFC 2784, 788 DOI 10.17487/RFC2784, March 2000, 789 . 791 [RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 792 of Explicit Congestion Notification (ECN) to IP", 793 RFC 3168, DOI 10.17487/RFC3168, September 2001, 794 . 796 [RFC3931] Lau, J., Ed., Townsley, M., Ed., and I. Goyret, Ed., 797 "Layer Two Tunneling Protocol - Version 3 (L2TPv3)", 798 RFC 3931, DOI 10.17487/RFC3931, March 2005, 799 . 801 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 802 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 803 December 2005, . 805 [RFC4380] Huitema, C., "Teredo: Tunneling IPv6 over UDP through 806 Network Address Translations (NATs)", RFC 4380, 807 DOI 10.17487/RFC4380, February 2006, 808 . 810 [RFC5129] Davie, B., Briscoe, B., and J. Tay, "Explicit Congestion 811 Marking in MPLS", RFC 5129, DOI 10.17487/RFC5129, January 812 2008, . 814 [RFC6040] Briscoe, B., "Tunnelling of Explicit Congestion 815 Notification", RFC 6040, DOI 10.17487/RFC6040, November 816 2010, . 818 10.2. Informative References 820 [GTPv1] 3GPP, "GPRS Tunnelling Protocol (GTP) across the Gn and Gp 821 interface", Technical Specification TS 29.060. 823 [GTPv1-U] 3GPP, "General Packet Radio System (GPRS) Tunnelling 824 Protocol User Plane (GTPv1-U)", Technical Specification TS 825 29.281. 827 [GTPv2-C] 3GPP, "Evolved General Packet Radio Service (GPRS) 828 Tunnelling Protocol for Control plane (GTPv2-C)", 829 Technical Specification TS 29.274. 831 [I-D.eastlake-sfc-nsh-ecn-support] 832 Eastlake, D., Briscoe, B., and A. Malis, "Explicit 833 Congestion Notification (ECN) and Congestion Feedback 834 Using the Network Service Header (NSH)", draft-eastlake- 835 sfc-nsh-ecn-support-02 (work in progress), October 2018. 837 [I-D.ietf-intarea-gue] 838 Herbert, T. and O. Zia, "Generic UDP Encapsulation", 839 draft-ietf-intarea-gue-06 (work in progress), August 2018. 841 [I-D.ietf-nvo3-geneve] 842 Gross, J., Ganga, I., and T. Sridhar, "Geneve: Generic 843 Network Virtualization Encapsulation", draft-ietf- 844 nvo3-geneve-08 (work in progress), October 2018. 846 [I-D.ietf-nvo3-vxlan-gpe] 847 Maino, F., Kreeger, L., and U. Elzur, "Generic Protocol 848 Extension for VXLAN", draft-ietf-nvo3-vxlan-gpe-06 (work 849 in progress), April 2018. 851 [RFC1701] Hanks, S., Li, T., Farinacci, D., and P. Traina, "Generic 852 Routing Encapsulation (GRE)", RFC 1701, 853 DOI 10.17487/RFC1701, October 1994, 854 . 856 [RFC2637] Hamzeh, K., Pall, G., Verthein, W., Taarud, J., Little, 857 W., and G. Zorn, "Point-to-Point Tunneling Protocol 858 (PPTP)", RFC 2637, DOI 10.17487/RFC2637, July 1999, 859 . 861 [RFC2983] Black, D., "Differentiated Services and Tunnels", 862 RFC 2983, DOI 10.17487/RFC2983, October 2000, 863 . 865 [RFC3260] Grossman, D., "New Terminology and Clarifications for 866 Diffserv", RFC 3260, DOI 10.17487/RFC3260, April 2002, 867 . 869 [RFC3308] Calhoun, P., Luo, W., McPherson, D., and K. Peirce, "Layer 870 Two Tunneling Protocol (L2TP) Differentiated Services 871 Extension", RFC 3308, DOI 10.17487/RFC3308, November 2002, 872 . 874 [RFC5415] Calhoun, P., Ed., Montemurro, M., Ed., and D. Stanley, 875 Ed., "Control And Provisioning of Wireless Access Points 876 (CAPWAP) Protocol Specification", RFC 5415, 877 DOI 10.17487/RFC5415, March 2009, 878 . 880 [RFC5845] Muhanna, A., Khalil, M., Gundavelli, S., and K. Leung, 881 "Generic Routing Encapsulation (GRE) Key Option for Proxy 882 Mobile IPv6", RFC 5845, DOI 10.17487/RFC5845, June 2010, 883 . 885 [RFC5944] Perkins, C., Ed., "IP Mobility Support for IPv4, Revised", 886 RFC 5944, DOI 10.17487/RFC5944, November 2010, 887 . 889 [RFC5996] Kaufman, C., Hoffman, P., Nir, Y., and P. Eronen, 890 "Internet Key Exchange Protocol Version 2 (IKEv2)", 891 RFC 5996, DOI 10.17487/RFC5996, September 2010, 892 . 894 [RFC6275] Perkins, C., Ed., Johnson, D., and J. Arkko, "Mobility 895 Support in IPv6", RFC 6275, DOI 10.17487/RFC6275, July 896 2011, . 898 [RFC6830] Farinacci, D., Fuller, V., Meyer, D., and D. Lewis, "The 899 Locator/ID Separation Protocol (LISP)", RFC 6830, 900 DOI 10.17487/RFC6830, January 2013, 901 . 903 [RFC7059] Steffann, S., van Beijnum, I., and R. van Rein, "A 904 Comparison of IPv6-over-IPv4 Tunnel Mechanisms", RFC 7059, 905 DOI 10.17487/RFC7059, November 2013, 906 . 908 [RFC7348] Mahalingam, M., Dutt, D., Duda, K., Agarwal, P., Kreeger, 909 L., Sridhar, T., Bursell, M., and C. Wright, "Virtual 910 eXtensible Local Area Network (VXLAN): A Framework for 911 Overlaying Virtualized Layer 2 Networks over Layer 3 912 Networks", RFC 7348, DOI 10.17487/RFC7348, August 2014, 913 . 915 [RFC7450] Bumgardner, G., "Automatic Multicast Tunneling", RFC 7450, 916 DOI 10.17487/RFC7450, February 2015, 917 . 919 [RFC7637] Garg, P., Ed. and Y. Wang, Ed., "NVGRE: Network 920 Virtualization Using Generic Routing Encapsulation", 921 RFC 7637, DOI 10.17487/RFC7637, September 2015, 922 . 924 [RFC7665] Halpern, J., Ed. and C. Pignataro, Ed., "Service Function 925 Chaining (SFC) Architecture", RFC 7665, 926 DOI 10.17487/RFC7665, October 2015, 927 . 929 [RFC8085] Eggert, L., Fairhurst, G., and G. Shepherd, "UDP Usage 930 Guidelines", BCP 145, RFC 8085, DOI 10.17487/RFC8085, 931 March 2017, . 933 [RFC8087] Fairhurst, G. and M. Welzl, "The Benefits of Using 934 Explicit Congestion Notification (ECN)", RFC 8087, 935 DOI 10.17487/RFC8087, March 2017, 936 . 938 [RFC8159] Konstantynowicz, M., Ed., Heron, G., Ed., Schatzmayr, R., 939 and W. Henderickx, "Keyed IPv6 Tunnel", RFC 8159, 940 DOI 10.17487/RFC8159, May 2017, 941 . 943 [RFC8229] Pauly, T., Touati, S., and R. Mantha, "TCP Encapsulation 944 of IKE and IPsec Packets", RFC 8229, DOI 10.17487/RFC8229, 945 August 2017, . 947 [RFC8300] Quinn, P., Ed., Elzur, U., Ed., and C. Pignataro, Ed., 948 "Network Service Header (NSH)", RFC 8300, 949 DOI 10.17487/RFC8300, January 2018, 950 . 952 Author's Address 954 Bob Briscoe 955 Independent 956 UK 958 EMail: ietf@bobbriscoe.net 959 URI: http://bobbriscoe.net/