idnits 2.17.1 draft-ietf-ipsecme-iptfs-05.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 == Line 1191 has weird spacing: '...4 any any...' == Line 1207 has weird spacing: '...4 any any...' == Using lowercase 'not' together with uppercase 'MUST', 'SHALL', 'SHOULD', or 'RECOMMENDED' is not an accepted usage according to RFC 2119. Please use uppercase 'NOT' together with RFC 2119 keywords (if that is what you mean). Found 'MUST not' in this paragraph: In order for the sender to estimate it's "RTT" value, the sender places a timestamp value in the "TVal" header field. On first receipt of this "TVal", the receiver records the new "TVal" value along with the time it arrived locally, subsequent receipt of the same "TVal" MUST not update the recorded time. When the receiver sends it's CC header it places this latest recorded value in the "TEcho" header field, along with 2 delay values, "Echo Delay" and "Transmit Delay". The "Echo Delay" value is the time delta from the recorded arrival time of "TVal" and the current clock in microseconds. The second value, "Transmit Delay", is the receiver's current transmission delay on the tunnel (i.e., the average time between sending packets on it's half of the IP-TFS tunnel). When the sender receives back it's "TVal" in the "TEcho" header field it calculates 2 RTT estimates. The first is the actual delay found by subtracting the "TEcho" value from it's current clock and then subtracting "Echo Delay" as well. The second RTT estimate is found by adding the received "Transmit Delay" header value to the senders own transmission delay (i.e., the average time between sending packets on it's half of the IP-TFS tunnel). The larger of these 2 RTT estimates SHOULD be used as the "RTT" value. The two estimates are required to handle different combinations of faster or slower tunnel packet paths with faster or slower fixed tunnel rates. Choosing the larger of the two values guarantees that the "RTT" is never considered faster than the aggregate transmission delay based on the IP-TFS tunnel rate (the second estimate), as well as never being considered faster than the actual RTT along the tunnel packet path (the first estimate). -- The document date (December 20, 2020) is 1216 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Missing Reference: '--800--' is mentioned on line 1023, but not defined -- Looks like a reference, but probably isn't: '60' on line 1023 == Missing Reference: '-240-' is mentioned on line 1023, but not defined == Missing Reference: '--4000----------------------' is mentioned on line 1023, but not defined Summary: 0 errors (**), 0 flaws (~~), 7 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group C. Hopps 3 Internet-Draft LabN Consulting, L.L.C. 4 Intended status: Standards Track December 20, 2020 5 Expires: June 23, 2021 7 IP Traffic Flow Security Using Aggregation and Fragmentation 8 draft-ietf-ipsecme-iptfs-05 10 Abstract 12 This document describes a mechanism to enhance IPsec traffic flow 13 security by adding traffic flow confidentiality to encrypted IP 14 encapsulated traffic. Traffic flow confidentiality is provided by 15 obscuring the size and frequency of IP traffic using a fixed-sized, 16 constant-send-rate IPsec tunnel. The solution allows for congestion 17 control as well as non-constant send-rate usage. 19 Status of This Memo 21 This Internet-Draft is submitted in full conformance with the 22 provisions of BCP 78 and BCP 79. 24 Internet-Drafts are working documents of the Internet Engineering 25 Task Force (IETF). Note that other groups may also distribute 26 working documents as Internet-Drafts. The list of current Internet- 27 Drafts is at https://datatracker.ietf.org/drafts/current/. 29 Internet-Drafts are draft documents valid for a maximum of six months 30 and may be updated, replaced, or obsoleted by other documents at any 31 time. It is inappropriate to use Internet-Drafts as reference 32 material or to cite them other than as "work in progress." 34 This Internet-Draft will expire on June 23, 2021. 36 Copyright Notice 38 Copyright (c) 2020 IETF Trust and the persons identified as the 39 document authors. All rights reserved. 41 This document is subject to BCP 78 and the IETF Trust's Legal 42 Provisions Relating to IETF Documents 43 (https://trustee.ietf.org/license-info) in effect on the date of 44 publication of this document. Please review these documents 45 carefully, as they describe your rights and restrictions with respect 46 to this document. Code Components extracted from this document must 47 include Simplified BSD License text as described in Section 4.e of 48 the Trust Legal Provisions and are provided without warranty as 49 described in the Simplified BSD License. 51 Table of Contents 53 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 54 1.1. Terminology & Concepts . . . . . . . . . . . . . . . . . 3 55 2. The IP-TFS Tunnel . . . . . . . . . . . . . . . . . . . . . . 4 56 2.1. Tunnel Content . . . . . . . . . . . . . . . . . . . . . 4 57 2.2. Payload Content . . . . . . . . . . . . . . . . . . . . . 5 58 2.2.1. Data Blocks . . . . . . . . . . . . . . . . . . . . . 6 59 2.2.2. No Implicit End Padding Required . . . . . . . . . . 6 60 2.2.3. Fragmentation, Sequence Numbers and All-Pad Payloads 6 61 2.2.4. Empty Payload . . . . . . . . . . . . . . . . . . . . 7 62 2.2.5. IP Header Value Mapping . . . . . . . . . . . . . . . 8 63 2.2.6. IP Time-To-Live (TTL) and Tunnel errors . . . . . . . 8 64 2.2.7. Effective MTU of the Tunnel . . . . . . . . . . . . . 8 65 2.3. Exclusive SA Use . . . . . . . . . . . . . . . . . . . . 8 66 2.4. Modes of Operation . . . . . . . . . . . . . . . . . . . 9 67 2.4.1. Non-Congestion Controlled Mode . . . . . . . . . . . 9 68 2.4.2. Congestion Controlled Mode . . . . . . . . . . . . . 9 69 3. Congestion Information . . . . . . . . . . . . . . . . . . . 10 70 3.1. ECN Support . . . . . . . . . . . . . . . . . . . . . . . 12 71 4. Configuration . . . . . . . . . . . . . . . . . . . . . . . . 12 72 4.1. Bandwidth . . . . . . . . . . . . . . . . . . . . . . . . 12 73 4.2. Fixed Packet Size . . . . . . . . . . . . . . . . . . . . 12 74 4.3. Congestion Control . . . . . . . . . . . . . . . . . . . 12 75 5. IKEv2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 76 5.1. USE_AGGFRAG Notification Message . . . . . . . . . . . . 13 77 6. Packet and Data Formats . . . . . . . . . . . . . . . . . . . 13 78 6.1. AGGFRAG_PAYLOAD Payload . . . . . . . . . . . . . . . . . 13 79 6.1.1. Non-Congestion Control AGGFRAG_PAYLOAD Payload Format 14 80 6.1.2. Congestion Control AGGFRAG_PAYLOAD Payload Format . . 15 81 6.1.3. Data Blocks . . . . . . . . . . . . . . . . . . . . . 16 82 6.1.4. IKEv2 USE_AGGFRAG Notification Message . . . . . . . 18 83 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 19 84 7.1. AGGFRAG_PAYLOAD Sub-Type Registry . . . . . . . . . . . . 19 85 7.2. USE_AGGFRAG Notify Message Status Type . . . . . . . . . 19 86 8. Security Considerations . . . . . . . . . . . . . . . . . . . 19 87 9. References . . . . . . . . . . . . . . . . . . . . . . . . . 20 88 9.1. Normative References . . . . . . . . . . . . . . . . . . 20 89 9.2. Informative References . . . . . . . . . . . . . . . . . 20 90 Appendix A. Example Of An Encapsulated IP Packet Flow . . . . . 22 91 Appendix B. A Send and Loss Event Rate Calculation . . . . . . . 23 92 Appendix C. Comparisons of IP-TFS . . . . . . . . . . . . . . . 23 93 C.1. Comparing Overhead . . . . . . . . . . . . . . . . . . . 23 94 C.1.1. IP-TFS Overhead . . . . . . . . . . . . . . . . . . . 23 95 C.1.2. ESP with Padding Overhead . . . . . . . . . . . . . . 24 97 C.2. Overhead Comparison . . . . . . . . . . . . . . . . . . . 25 98 C.3. Comparing Available Bandwidth . . . . . . . . . . . . . . 25 99 C.3.1. Ethernet . . . . . . . . . . . . . . . . . . . . . . 26 100 Appendix D. Acknowledgements . . . . . . . . . . . . . . . . . . 28 101 Appendix E. Contributors . . . . . . . . . . . . . . . . . . . . 28 102 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 28 104 1. Introduction 106 Traffic Analysis ([RFC4301], [AppCrypt]) is the act of extracting 107 information about data being sent through a network. While one may 108 directly obscure the data through the use of encryption [RFC4303], 109 the traffic pattern itself exposes information due to variations in 110 it's shape and timing ([I-D.iab-wire-image], [AppCrypt]). Hiding the 111 size and frequency of traffic is referred to as Traffic Flow 112 Confidentiality (TFC) per [RFC4303]. 114 [RFC4303] provides for TFC by allowing padding to be added to 115 encrypted IP packets and allowing for transmission of all-pad packets 116 (indicated using protocol 59). This method has the major limitation 117 that it can significantly under-utilize the available bandwidth. 119 The IP-TFS solution provides for full TFC without the aforementioned 120 bandwidth limitation. This is accomplished by using a constant-send- 121 rate IPsec [RFC4303] tunnel with fixed-sized encapsulating packets; 122 however, these fixed-sized packets can contain partial, whole or 123 multiple IP packets to maximize the bandwidth of the tunnel. A non- 124 constant send-rate is allowed, but the confidentiality properties of 125 its use are outside the scope of this document. 127 For a comparison of the overhead of IP-TFS with the RFC4303 128 prescribed TFC solution see Appendix C. 130 Additionally, IP-TFS provides for dealing with network congestion 131 [RFC2914]. This is important for when the IP-TFS user is not in full 132 control of the domain through which the IP-TFS tunnel path flows. 134 1.1. Terminology & Concepts 136 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 137 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 138 "OPTIONAL" in this document are to be interpreted as described in 139 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, 140 as shown here. 142 This document assumes familiarity with IP security concepts described 143 in [RFC4301]. 145 2. The IP-TFS Tunnel 147 As mentioned in Section 1 IP-TFS utilizes an IPsec [RFC4303] tunnel 148 (SA) as it's transport. To provide for full TFC, fixed-sized 149 encapsulating packets are sent at a constant rate on the tunnel. 151 The primary input to the tunnel algorithm is the requested bandwidth 152 used by the tunnel. Two values are then required to provide for this 153 bandwidth, the fixed size of the encapsulating packets, and rate at 154 which to send them. 156 The fixed packet size MAY either be specified manually or could be 157 determined through the other methods such as the Packetization Layer 158 MTU Discovery (PLMTUD) ([RFC4821], [RFC8899]) or Path MTU discovery 159 (PMTUD) ([RFC1191], [RFC8201]). PMTUD is known to have issues so 160 PLMTUD is considered the more robust option. 162 Given the encapsulating packet size and the requested tunnel used 163 bandwidth, the corresponding packet send rate can be calculated. The 164 packet send rate is the requested bandwidth divided by the size of 165 the encapsulating packet. 167 The egress of the IP-TFS tunnel MUST allow for and expect the ingress 168 (sending) side of the IP-TFS tunnel to vary the size and rate of sent 169 encapsulating packets, unless constrained by other policy. 171 2.1. Tunnel Content 173 As previously mentioned, one issue with the TFC padding solution in 174 [RFC4303] is the large amount of wasted bandwidth as only one IP 175 packet can be sent per encapsulating packet. In order to maximize 176 bandwidth IP-TFS breaks this one-to-one association. 178 IP-TFS aggregates as well as fragments the inner IP traffic flow into 179 fixed-sized encapsulating IPsec tunnel packets. Padding is only 180 added to the the tunnel packets if there is no data available to be 181 sent at the time of tunnel packet transmission, or if fragmentation 182 has been disabled by the receiver. 184 This is accomplished using a new Encapsulating Security Payload (ESP, 185 [RFC4303]) type which is identified by the number AGGFRAG_PAYLOAD 186 (Section 6.1). 188 Other non-IP-TFS uses of this aggregation and fragmentation 189 encapsulation have been identified, such as increased performance 190 through packet aggregation, as well as handling MTU issues using 191 fragmentation. These uses are not defined here, but are also not 192 restricted by this document. 194 2.2. Payload Content 196 The AGGFRAG_PAYLOAD payload content defined in this document is 197 comprised of a 4 or 24 octet header followed by either a partial, a 198 full or multiple partial or full data blocks. The following diagram 199 illustrates this payload within the ESP packet. See Section 6.1 for 200 the exact formats of the AGGFRAG_PAYLOAD payload. 202 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 . Outer Encapsulating Header ... . 204 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 . ESP Header... . 206 +---------------------------------------------------------------+ 207 | [AGGFRAG subtype/flags] : BlockOffset | 208 +---------------------------------------------------------------+ 209 : [Optional Congestion Info] : 210 +---------------------------------------------------------------+ 211 | DataBlocks ... ~ 212 ~ ~ 213 ~ | 214 +---------------------------------------------------------------| 215 . ESP Trailer... . 216 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218 Figure 1: Layout of an IP-TFS IPsec Packet 220 The "BlockOffset" value is either zero or some offset into or past 221 the end of the "DataBlocks" data. 223 If the "BlockOffset" value is zero it means that the "DataBlocks" 224 data begins with a new data block. 226 Conversely, if the "BlockOffset" value is non-zero it points to the 227 start of the new data block, and the initial "DataBlocks" data 228 belongs to a previous data block that is still being re-assembled. 230 The "BlockOffset" can point past the end of the "DataBlocks" data 231 which indicates that the next data block occurs in a subsequent 232 encapsulating packet. 234 Having the "BlockOffset" always point at the next available data 235 block allows for recovering the next inner packet in the presence of 236 outer encapsulating packet loss. 238 An example IP-TFS packet flow can be found in Appendix A. 240 2.2.1. Data Blocks 242 +---------------------------------------------------------------+ 243 | Type | rest of IPv4, IPv6 or pad. 244 +-------- 246 Figure 2: Layout of IP-TFS data block 248 A data block is defined by a 4-bit type code followed by the data 249 block data. The type values have been carefully chosen to coincide 250 with the IPv4/IPv6 version field values so that no per-data block 251 type overhead is required to encapsulate an IP packet. Likewise, the 252 length of the data block is extracted from the encapsulated IPv4 or 253 IPv6 packet's length field. 255 2.2.2. No Implicit End Padding Required 257 It's worth noting that since a data block type is identified by its 258 first octet there is never a need for an implicit pad at the end of 259 an encapsulating packet. Even when the start of a data block occurs 260 near the end of a encapsulating packet such that there is no room for 261 the length field of the encapsulated header to be included in the 262 current encapsulating packet, the fact that the length comes at a 263 known location and is guaranteed to be present is enough to fetch the 264 length field from the subsequent encapsulating packet payload. Only 265 when there is no data to encapsulated is end padding required, and 266 then an explicit "Pad Data Block" would be used to identify the 267 padding. 269 2.2.3. Fragmentation, Sequence Numbers and All-Pad Payloads 271 In order for a receiver to be able to reassemble fragmented inner- 272 packets, the sender MUST send the inner-packet fragments back-to-back 273 in the logical outer packet stream (i.e., using consecutive ESP 274 sequence numbers). However, the sender is allowed to insert "all- 275 pad" payloads (i.e., payloads with a "BlockOffset" of zero and a 276 single pad "DataBlock") in between the packets carrying the inner- 277 packet fragment payloads. This possible interleaving of all-pad 278 payloads allows the sender to always be able to send a tunnel packet, 279 regardless of the encapsulation computational requirements. 281 When a receiver is reassembling an inner-packet, and it receives an 282 "all-pad" payload, it increments the expected sequence number that 283 the next inner-packet fragment is expected to arrive in. 285 Given the above, the receiver will need to handle out-of-order 286 arrival of outer ESP packets prior to reassembly processing. ESP 287 already provides for detecting replay attacks (normally) utilizing a 288 window. A similar sequence number based sliding window can be used 289 to correct re-ordering of the outer packet stream. Receiving a 290 larger (newer) sequence number packet advances the window, and 291 received older ESP packets whose sequence numbers the window has 292 passed by are dropped. A good choice for the size of this window 293 depends on the amount of re-ordering the user may normally 294 experience. As the amount of reordering that may be present is hard 295 to predict the window size SHOULD be configurable by the user. 296 Implementations MAY also dynamically adjust the reordering window 297 based on actual reordering seen in arriving packets. Finally, we 298 note that as IP-TFS is sending a continuous stream of packets there 299 is no requirement for timers (although there's no prohibition either) 300 as newly arrived packets will cause the window to advance and older 301 packets will then be processed as they leave the window. 303 2.2.3.1. Optional Extra Padding 305 When the tunnel bandwidth is not being fully utilized, an 306 implementation MAY pad-out the current encapsulating packet in order 307 to deliver an inner packet un-fragmented in the following outer 308 packet. The benefit would be to avoid inner-packet fragmentation in 309 the presence of a bursty offered load (non-bursty traffic will 310 naturally not fragment). The cost is complexity and added delay of 311 inner traffic. The main advantage to avoiding fragmentation is to 312 minimize inner packet loss in the presence of outer packet loss. 313 When this is worthwhile (e.g., how much loss and what type of loss is 314 required, given different inner traffic shapes and utilization, for 315 this to make sense), and what values to use for the allowable/added 316 delay may be worth researching, but is outside the scope of this 317 document. 319 While use of padding to avoid fragmentation does not impact 320 interoperability, used inappropriately it can reduce the effective 321 throughput of a tunnel. Implementations implementing the above 322 approach will need to take care to not reduce the effective capacity, 323 and overall utility, of the tunnel through the overuse of padding. 325 2.2.4. Empty Payload 327 In order to support reporting of congestion control information 328 (described later) on a non-AGGFRAG_PAYLOAD enabled SA, IP-TFS allows 329 for the sending of an AGGFRAG_PAYLOAD payload with no data blocks 330 (i.e., the ESP payload length is equal to the AGGFRAG_PAYLOAD header 331 length). This special payload is called an empty payload. 333 2.2.5. IP Header Value Mapping 335 [RFC4301] provides some direction on when and how to map various 336 values from an inner IP header to the outer encapsulating header, 337 namely the Don't-Fragment (DF) bit ([RFC0791] and [RFC8200]), the 338 Differentiated Services (DS) field [RFC2474] and the Explicit 339 Congestion Notification (ECN) field [RFC3168]. Unlike [RFC4301], IP- 340 TFS may and often will be encapsulating more than one IP packet per 341 ESP packet. To deal with this, these mappings are restricted 342 further. In particular IP-TFS never maps the inner DF bit as it is 343 unrelated to the IP-TFS tunnel functionality; IP-TFS never IP 344 fragments the inner packets and the inner packets will not affect the 345 fragmentation of the outer encapsulation packets. Likewise, the ECN 346 value need not be mapped as any congestion related to the constant- 347 send-rate IP-TFS tunnel is unrelated (by design!) to the inner 348 traffic flow. Finally, by default the DS field SHOULD NOT be copied 349 although an implementation MAY choose to allow for configuration to 350 override this behavior. An implementation SHOULD also allow the DS 351 value to be set by configuration. 353 2.2.6. IP Time-To-Live (TTL) and Tunnel errors 355 [RFC4301] specifies how to modify the inner packet TTL ([RFC0791]). 357 Any errors (e.g., ICMP errors arriving back at the tunnel ingress due 358 to tunnel traffic) should be handled the same as with non IP-TFS 359 IPsec tunnels. 361 2.2.7. Effective MTU of the Tunnel 363 Unlike [RFC4301], there is normally no effective MTU (EMTU) on an IP- 364 TFS tunnel as all IP packet sizes are properly transmitted without 365 requiring IP fragmentation prior to tunnel ingress. 367 If IP-TFS fragmentation has been disabled, then the tunnel's EMTU and 368 behaviors are the same as normal IPsec tunnels ([RFC4301]). 370 2.3. Exclusive SA Use 372 It is not the intention of this specification to allow for mixed use 373 of an AGGFRAG_PAYLOAD enabled SA. In other words, an SA that has 374 AGGFRAG_PAYLOAD enabled MUST NOT have non-AGGFRAG_PAYLOAD payloads 375 such as IP (IP protocol 4), TCP transport (IP protocol 6), or ESP pad 376 packets (protocol 59) intermixed with non-empty AGGFRAG_PAYLOAD 377 payloads. While it's possible to envision making the algorithm work 378 in the presence of sequence number skips in the AGGFRAG_PAYLOAD 379 payload stream, the added complexity is not deemed worthwhile. Other 380 IPsec uses can configure and use their own SAs. 382 2.4. Modes of Operation 384 Just as with normal IPsec/ESP tunnels, IP-TFS tunnels are 385 unidirectional. Bidirectional IP-TFS functionality is achieved by 386 setting up 2 IP-TFS tunnels, one in either direction. 388 An IP-TFS tunnel can operate in 2 modes, a non-congestion controlled 389 mode and congestion controlled mode. 391 2.4.1. Non-Congestion Controlled Mode 393 In the non-congestion controlled mode IP-TFS sends fixed-sized 394 packets at a constant rate. The packet send rate is constant and is 395 not automatically adjusted regardless of any network congestion 396 (e.g., packet loss). 398 For similar reasons as given in [RFC7510] the non-congestion 399 controlled mode should only be used where the user has full 400 administrative control over the path the tunnel will take. This is 401 required so the user can guarantee the bandwidth and also be sure as 402 to not be negatively affecting network congestion [RFC2914]. In this 403 case packet loss should be reported to the administrator (e.g., via 404 syslog, YANG notification, SNMP traps, etc) so that any failures due 405 to a lack of bandwidth can be corrected. 407 2.4.2. Congestion Controlled Mode 409 With the congestion controlled mode, IP-TFS adapts to network 410 congestion by lowering the packet send rate to accommodate the 411 congestion, as well as raising the rate when congestion subsides. 412 Since overhead is per packet, by allowing for maximal fixed-size 413 packets and varying the send rate transport overhead is minimized. 415 The output of the congestion control algorithm will adjust the rate 416 at which the ingress sends packets. While this document does not 417 require a specific congestion control algorithm, best current 418 practice RECOMMENDS that the algorithm conform to [RFC5348]. 419 Congestion control principles are documented in [RFC2914] as well. 420 An example of an implementation of the [RFC5348] algorithm which 421 matches the requirements of IP-TFS (i.e., designed for fixed-size 422 packet and send rate varied based on congestion) is documented in 423 [RFC4342]. 425 The required inputs for the TCP friendly rate control algorithm 426 described in [RFC5348] are the receiver's loss event rate and the 427 sender's estimated round-trip time (RTT). These values are provided 428 by IP-TFS using the congestion information header fields described in 429 Section 3. In particular these values are sufficient to implement 430 the algorithm described in [RFC5348]. 432 At a minimum, the congestion information must be sent, from the 433 receiver and from the sender, at least once per RTT. Prior to 434 establishing an RTT the information SHOULD be sent constantly from 435 the sender and the receiver so that an RTT estimate can be 436 established. The lack of receiving this information over multiple 437 consecutive RTT intervals should be considered a congestion event 438 that causes the sender to adjust it's sending rate lower. For 439 example, [RFC4342] calls this the "no feedback timeout" and it is 440 equal to 4 RTT intervals. When a "no feedback timeout" has occurred 441 [RFC4342] halves the sending rate. 443 An implementation MAY choose to always include the congestion 444 information in it's IP-TFS payload header if sending on an IP-TFS 445 enabled SA. Since IP-TFS normally will operate with a large packet 446 size, the congestion information should represent a small portion of 447 the available tunnel bandwidth. An implementation choosing to always 448 send the data MAY also choose to only update the "LossEventRate" and 449 "RTT" header field values it sends every "RTT" though. 451 When an implementation is choosing a congestion control algorithm (or 452 a selection of algorithms) one should remember that IP-TFS is not 453 providing for reliable delivery of IP traffic, and so per packet ACKs 454 are not required and are not provided. 456 It's worth noting that the variable send-rate of a congestion 457 controlled IP-TFS tunnel, is not private; however, this send-rate is 458 being driven by network congestion, and as long as the encapsulated 459 (inner) traffic flow shape and timing are not directly affecting the 460 (outer) network congestion, the variations in the tunnel rate will 461 not weaken the provided inner traffic flow confidentiality. 463 2.4.2.1. Circuit Breakers 465 In additional to congestion control, implementations MAY choose to 466 define and implement circuit breakers [RFC8084] as a recovery method 467 of last resort. Enabling circuit breakers is also a reason a user 468 may wish to enable congestion information reports even when using the 469 non-congestion controlled mode of operation. The definition of 470 circuit breakers are outside the scope of this document. 472 3. Congestion Information 474 In order to support the congestion control mode, the sender needs to 475 know the loss event rate and also be able to approximate the RTT 476 ([RFC5348]). In order to obtain these values the receiver sends 477 congestion control information on it's SA back to the sender. Thus, 478 in order to support congestion control the receiver must have a 479 paired SA back to the sender (this is always the case when the tunnel 480 was created using IKEv2). If the SA back to the sender is a non- 481 AGGFRAG_PAYLOAD enabled SA then an AGGFRAG_PAYLOAD empty payload 482 (i.e., header only) is used to convey the information. 484 In order to calculate a loss event rate compatible with [RFC5348], 485 the receiver needs to have a round-trip time estimate. Thus the 486 sender communicates this estimate in the "RTT" header field. On 487 startup this value will be zero as no RTT estimate is yet known. 489 In order for the sender to estimate it's "RTT" value, the sender 490 places a timestamp value in the "TVal" header field. On first 491 receipt of this "TVal", the receiver records the new "TVal" value 492 along with the time it arrived locally, subsequent receipt of the 493 same "TVal" MUST not update the recorded time. When the receiver 494 sends it's CC header it places this latest recorded value in the 495 "TEcho" header field, along with 2 delay values, "Echo Delay" and 496 "Transmit Delay". The "Echo Delay" value is the time delta from the 497 recorded arrival time of "TVal" and the current clock in 498 microseconds. The second value, "Transmit Delay", is the receiver's 499 current transmission delay on the tunnel (i.e., the average time 500 between sending packets on it's half of the IP-TFS tunnel). When the 501 sender receives back it's "TVal" in the "TEcho" header field it 502 calculates 2 RTT estimates. The first is the actual delay found by 503 subtracting the "TEcho" value from it's current clock and then 504 subtracting "Echo Delay" as well. The second RTT estimate is found 505 by adding the received "Transmit Delay" header value to the senders 506 own transmission delay (i.e., the average time between sending 507 packets on it's half of the IP-TFS tunnel). The larger of these 2 508 RTT estimates SHOULD be used as the "RTT" value. The two estimates 509 are required to handle different combinations of faster or slower 510 tunnel packet paths with faster or slower fixed tunnel rates. 511 Choosing the larger of the two values guarantees that the "RTT" is 512 never considered faster than the aggregate transmission delay based 513 on the IP-TFS tunnel rate (the second estimate), as well as never 514 being considered faster than the actual RTT along the tunnel packet 515 path (the first estimate). 517 The receiver also calculates, and communicates in the "LossEventRate" 518 header field, the loss event rate for use by the sender. This is 519 slightly different from [RFC4342] which periodically sends all the 520 loss interval data back to the sender so that it can do the 521 calculation. See Appendix B for a suggested way to calculate the 522 loss event rate value. Initially this value will be zero (indicating 523 no loss) until enough data has been collected by the receiver to 524 update it. 526 3.1. ECN Support 528 In additional to normal packet loss information IP-TFS supports use 529 of the ECN bits in the encapsulating IP header [RFC3168] for 530 identifying congestion. If ECN use is enabled and a packet arrives 531 at the egress endpoint with the Congestion Experienced (CE) value 532 set, then the receiver considers that packet as being dropped, 533 although it does not drop it. The receiver MUST set the E bit in any 534 AGGFRAG_PAYLOAD payload header containing a "LossEventRate" value 535 derived from a CE value being considered. 537 As noted in [RFC3168] the ECN bits are not protected by IPsec and 538 thus may constitute a covert channel. For this reason ECN use SHOULD 539 NOT be enabled by default. 541 4. Configuration 543 IP-TFS is meant to be deployable with a minimal amount of 544 configuration. All IP-TFS specific configuration should be able to 545 be specified at the unidirectional tunnel ingress (sending) side. It 546 is intended that non-IKEv2 operation is supported, at least, with 547 local static configuration. 549 4.1. Bandwidth 551 Bandwidth is a local configuration option. For non-congestion 552 controlled mode the bandwidth SHOULD be configured. For congestion 553 controlled mode one can configure the bandwidth or have no 554 configuration and let congestion control discover the maximum 555 bandwidth available. No standardized configuration method is 556 required. 558 4.2. Fixed Packet Size 560 The fixed packet size to be used for the tunnel encapsulation packets 561 MAY be configured manually or can be automatically determined using 562 other methods such as PLMTUD ([RFC4821], [RFC8899]) or PMTUD 563 ([RFC1191], [RFC8201]). As PMTUD is known to have issues, PLMTUD is 564 considered the more robust option. No standardized configuration 565 method is required. 567 4.3. Congestion Control 569 Congestion control is a local configuration option. No standardized 570 configuration method is required. 572 5. IKEv2 574 5.1. USE_AGGFRAG Notification Message 576 As mentioned previously IP-TFS tunnels utilize ESP payloads of type 577 AGGFRAG_PAYLOAD. 579 When using IKEv2, a new "USE_AGGFRAG" Notification Message is used to 580 enable use of the AGGFRAG_PAYLOAD payload on a child SA pair. The 581 method used is similar to how USE_TRANSPORT_MODE is negotiated, as 582 described in [RFC7296]. 584 To request using the AGGFRAG_PAYLOAD payload on the Child SA pair, 585 the initiator includes the USE_AGGFRAG notification in an SA payload 586 requesting a new Child SA (either during the initial IKE_AUTH or 587 during non-rekeying CREATE_CHILD_SA exchanges). If the request is 588 accepted then response MUST also include a notification of type 589 USE_AGGFRAG. If the responder declines the request the child SA will 590 be established without AGGFRAG_PAYLOAD payload use enabled. If this 591 is unacceptable to the initiator, the initiator MUST delete the child 592 SA. 594 The USE_AGGFRAG notification MUST NOT be sent, and MUST be ignored, 595 during a CREATE_CHILD_SA rekeying exchange as it is not allowed to 596 change use of the AGGFRAG_PAYLOAD payload type during rekeying. 598 The USE_AGGFRAG notification contains a 1 octet payload of flags that 599 specify any requirements from the sender of the message. If any 600 requirement flags are not understood or cannot be supported by the 601 receiver then the receiver should not enable use of AGGFRAG_PAYLOAD 602 payload type (either by not responding with the USE_AGGFRAG 603 notification, or in the case of the initiator, by deleting the child 604 SA if the now established non-AGGFRAG_PAYLOAD using SA is 605 unacceptable). 607 The notification type and payload flag values are defined in 608 Section 6.1.4. 610 6. Packet and Data Formats 612 6.1. AGGFRAG_PAYLOAD Payload 614 ESP Payload Type: 0x5 616 An IP-TFS payload is identified by the ESP payload type 617 AGGFRAG_PAYLOAD which has the value 0x5. The first octet of this 618 payload indicates the format of the remaining payload data. 620 0 1 2 3 4 5 6 7 621 +-+-+-+-+-+-+-+-+-+-+- 622 | Sub-type | ... 623 +-+-+-+-+-+-+-+-+-+-+- 625 Sub-type: 626 An 8 bit value indicating the payload format. 628 This specification defines 2 payload sub-types. These payload 629 formats are defined in the following sections. 631 6.1.1. Non-Congestion Control AGGFRAG_PAYLOAD Payload Format 633 The non-congestion control AGGFRAG_PAYLOAD payload is comprised of a 634 4 octet header followed by a variable amount of "DataBlocks" data as 635 shown below. 637 1 2 3 638 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 1 639 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 640 | Sub-Type (0) | Reserved | BlockOffset | 641 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 642 | DataBlocks ... 643 +-+-+-+-+-+-+-+-+-+-+- 645 Sub-type: 646 An octet indicating the payload format. For this non-congestion 647 control format, the value is 0. 649 Reserved: 650 An octet set to 0 on generation, and ignored on receipt. 652 BlockOffset: 653 A 16 bit unsigned integer counting the number of octets of 654 "DataBlocks" data before the start of a new data block. 655 "BlockOffset" can count past the end of the "DataBlocks" data in 656 which case all the "DataBlocks" data belongs to the previous data 657 block being re-assembled. If the "BlockOffset" extends into 658 subsequent packets it continues to only count subsequent 659 "DataBlocks" data (i.e., it does not count subsequent packets 660 non-"DataBlocks" octets). 662 DataBlocks: 663 Variable number of octets that begins with the start of a data 664 block, or the continuation of a previous data block, followed by 665 zero or more additional data blocks. 667 6.1.2. Congestion Control AGGFRAG_PAYLOAD Payload Format 669 The congestion control AGGFRAG_PAYLOAD payload is comprised of a 24 670 octet header followed by a variable amount of "DataBlocks" data as 671 shown below. 673 1 2 3 674 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 1 675 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 676 | Sub-type (1) | Reserved |E| BlockOffset | 677 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 678 | LossEventRate | 679 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 680 | RTT | Echo Delay ... 681 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 682 ... Echo Delay | Transmit Delay | 683 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 684 | TVal | 685 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 686 | TEcho | 687 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 688 | DataBlocks ... 689 +-+-+-+-+-+-+-+-+-+-+- 691 Sub-type: 692 An octet indicating the payload format. For this congestion 693 control format, the value is 1. 695 Reserved: 696 A 7 bit field set to 0 on generation, and ignored on receipt. 698 E: 699 A 1 bit value if set indicates that Congestion Experienced (CE) 700 ECN bits were received and used in deriving the reported 701 "LossEventRate". 703 BlockOffset: 704 The same value as the non-congestion controlled payload format 705 value. 707 LossEventRate: 708 A 32 bit value specifying the inverse of the current loss event 709 rate as calculated by the receiver. A value of zero indicates no 710 loss. Otherwise the loss event rate is "1/LossEventRate". 712 RTT: 713 A 22 bit value specifying the sender's current round-trip time 714 estimate in microseconds. The value MAY be zero prior to the 715 sender having calculated a round-trip time estimate. The value 716 SHOULD be set to zero on non-AGGFRAG_PAYLOAD enabled SAs. If the 717 value is equal to or larger than "0x3FFFFF" it MUST be set to 718 "0x3FFFFF". 720 Echo Delay: 721 A 21 bit value specifying the delay in microseconds incurred 722 between the receiver first receiving the "TVal" value which it is 723 sending back in "TEcho". If the value is equal to or larger than 724 "0x1FFFFF" it MUST be set to "0x1FFFFF". 726 Transmit Delay: 727 A 21 bit value specifying the transmission delay in microseconds. 728 This is the fixed (or average) delay on the receiver between it 729 sending packets on the IPTFS tunnel. If the value is equal to or 730 larger than "0x1FFFFF" it MUST be set to "0x1FFFFF". 732 TVal: 733 An opaque 32 bit value that will be echoed back by the receiver in 734 later packets in the "TEcho" field, along with an "Echo Delay" 735 value of how long that echo took. 737 TEcho: 738 The opaque 32 bit value from a received packet's "TVal" field. 739 The received "TVal" is placed in "TEcho" along with an "Echo 740 Delay" value indicating how long it has been since receiving the 741 "TVal" value. 743 DataBlocks: 744 Variable number of octets that begins with the start of a data 745 block, or the continuation of a previous data block, followed by 746 zero or more additional data blocks. For the special case of 747 sending congestion control information on an non-IP-TFS enabled SA 748 this value MUST be empty (i.e., be zero octets long). 750 6.1.3. Data Blocks 752 1 2 3 753 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 1 754 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 755 | Type | IPv4, IPv6 or pad... 756 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+- 758 Type: 759 A 4 bit field where 0x0 identifies a pad data block, 0x4 indicates 760 an IPv4 data block, and 0x6 indicates an IPv6 data block. 762 6.1.3.1. IPv4 Data Block 764 1 2 3 765 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 1 766 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 767 | 0x4 | IHL | TypeOfService | TotalLength | 768 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 769 | Rest of the inner packet ... 770 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+- 772 These values are the actual values within the encapsulated IPv4 773 header. In other words, the start of this data block is the start of 774 the encapsulated IP packet. 776 Type: 777 A 4 bit value of 0x4 indicating IPv4 (i.e., first nibble of the 778 IPv4 packet). 780 TotalLength: 781 The 16 bit unsigned integer "Total Length" field of the IPv4 inner 782 packet. 784 6.1.3.2. IPv6 Data Block 786 1 2 3 787 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 1 788 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 789 | 0x6 | TrafficClass | FlowLabel | 790 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 791 | PayloadLength | Rest of the inner packet ... 792 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+- 794 These values are the actual values within the encapsulated IPv6 795 header. In other words, the start of this data block is the start of 796 the encapsulated IP packet. 798 Type: 799 A 4 bit value of 0x6 indicating IPv6 (i.e., first nibble of the 800 IPv6 packet). 802 PayloadLength: 803 The 16 bit unsigned integer "Payload Length" field of the inner 804 IPv6 inner packet. 806 6.1.3.3. Pad Data Block 808 1 2 3 809 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 1 810 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 811 | 0x0 | Padding ... 812 +-+-+-+-+-+-+-+-+-+-+- 814 Type: 815 A 4 bit value of 0x0 indicating a padding data block. 817 Padding: 818 extends to end of the encapsulating packet. 820 6.1.4. IKEv2 USE_AGGFRAG Notification Message 822 As discussed in Section 5.1 a notification message USE_AGGFRAG is 823 used to negotiate use of the ESP AGGFRAG_PAYLOAD payload type. 825 The USE_AGGFRAG Notification Message State Type is (TBD2). 827 The notification payload contains 1 octet of requirement flags. 828 There are currently 2 requirement flags defined. This may be revised 829 by later specifications. 831 +-+-+-+-+-+-+-+-+ 832 |0|0|0|0|0|0|C|D| 833 +-+-+-+-+-+-+-+-+ 835 0: 836 6 bits - reserved, MUST be zero on send, unless defined by later 837 specifications. 839 C: 840 Congestion Control bit. If set, then the sender is requiring that 841 congestion control information MUST be returned to it periodically 842 as defined in Section 3. 844 D: 845 Don't Fragment bit, if set indicates the sender of the notify 846 message does not support receiving packet fragments (i.e., inner 847 packets MUST be sent using a single "Data Block"). This value 848 only applies to what the sender is capable of receiving; the 849 sender MAY still send packet fragments unless similarly restricted 850 by the receiver in it's USE_AGGFRAG notification. 852 7. IANA Considerations 854 7.1. AGGFRAG_PAYLOAD Sub-Type Registry 856 This document requests IANA create a registry called "AGGFRAG_PAYLOAD 857 Sub-Type Registry" under a new category named "ESP AGGFRAG_PAYLOAD 858 Parameters". The registration policy for this registry is "Standards 859 Action" ([RFC8126] and [RFC7120]). 861 Name: 862 AGGFRAG_PAYLOAD Sub-Type Registry 864 Description: 865 AGGFRAG_PAYLOAD Payload Formats. 867 Reference: 868 This document 870 This initial content for this registry is as follows: 872 Sub-Type Name Reference 873 -------------------------------------------------------- 874 0 Non-Congestion Control Format This document 875 1 Congestion Control Format This document 876 3-255 Reserved 878 7.2. USE_AGGFRAG Notify Message Status Type 880 This document requests a status type USE_AGGFRAG be allocated from 881 the "IKEv2 Notify Message Types - Status Types" registry. 883 Value: 884 TBD2 886 Name: 887 USE_AGGFRAG 889 Reference: 890 This document 892 8. Security Considerations 894 This document describes a mechanism to add Traffic Flow 895 Confidentiality to IP traffic. Use of this mechanism is expected to 896 increase the security of the traffic being transported. Other than 897 the additional security afforded by using this mechanism, IP-TFS 898 utilizes the security protocols [RFC4303] and [RFC7296] and so their 899 security considerations apply to IP-TFS as well. 901 As noted previously in Section 2.4.2, for TFC to be fully maintained 902 the encapsulated traffic flow should not be affecting network 903 congestion in a predictable way, and if it would be then non- 904 congestion controlled mode use should be considered instead. 906 9. References 908 9.1. Normative References 910 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 911 Requirement Levels", BCP 14, RFC 2119, 912 DOI 10.17487/RFC2119, March 1997, 913 . 915 [RFC4303] Kent, S., "IP Encapsulating Security Payload (ESP)", 916 RFC 4303, DOI 10.17487/RFC4303, December 2005, 917 . 919 [RFC7296] Kaufman, C., Hoffman, P., Nir, Y., Eronen, P., and T. 920 Kivinen, "Internet Key Exchange Protocol Version 2 921 (IKEv2)", STD 79, RFC 7296, DOI 10.17487/RFC7296, October 922 2014, . 924 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 925 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 926 May 2017, . 928 9.2. Informative References 930 [AppCrypt] 931 Schneier, B., "Applied Cryptography: Protocols, 932 Algorithms, and Source Code in C", 11 2017. 934 [I-D.iab-wire-image] 935 Trammell, B. and M. Kuehlewind, "The Wire Image of a 936 Network Protocol", draft-iab-wire-image-01 (work in 937 progress), November 2018. 939 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 940 DOI 10.17487/RFC0791, September 1981, 941 . 943 [RFC1191] Mogul, J. and S. Deering, "Path MTU discovery", RFC 1191, 944 DOI 10.17487/RFC1191, November 1990, 945 . 947 [RFC2474] Nichols, K., Blake, S., Baker, F., and D. Black, 948 "Definition of the Differentiated Services Field (DS 949 Field) in the IPv4 and IPv6 Headers", RFC 2474, 950 DOI 10.17487/RFC2474, December 1998, 951 . 953 [RFC2914] Floyd, S., "Congestion Control Principles", BCP 41, 954 RFC 2914, DOI 10.17487/RFC2914, September 2000, 955 . 957 [RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 958 of Explicit Congestion Notification (ECN) to IP", 959 RFC 3168, DOI 10.17487/RFC3168, September 2001, 960 . 962 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 963 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 964 December 2005, . 966 [RFC4342] Floyd, S., Kohler, E., and J. Padhye, "Profile for 967 Datagram Congestion Control Protocol (DCCP) Congestion 968 Control ID 3: TCP-Friendly Rate Control (TFRC)", RFC 4342, 969 DOI 10.17487/RFC4342, March 2006, 970 . 972 [RFC4821] Mathis, M. and J. Heffner, "Packetization Layer Path MTU 973 Discovery", RFC 4821, DOI 10.17487/RFC4821, March 2007, 974 . 976 [RFC5348] Floyd, S., Handley, M., Padhye, J., and J. Widmer, "TCP 977 Friendly Rate Control (TFRC): Protocol Specification", 978 RFC 5348, DOI 10.17487/RFC5348, September 2008, 979 . 981 [RFC7120] Cotton, M., "Early IANA Allocation of Standards Track Code 982 Points", BCP 100, RFC 7120, DOI 10.17487/RFC7120, January 983 2014, . 985 [RFC7510] Xu, X., Sheth, N., Yong, L., Callon, R., and D. Black, 986 "Encapsulating MPLS in UDP", RFC 7510, 987 DOI 10.17487/RFC7510, April 2015, 988 . 990 [RFC8084] Fairhurst, G., "Network Transport Circuit Breakers", 991 BCP 208, RFC 8084, DOI 10.17487/RFC8084, March 2017, 992 . 994 [RFC8126] Cotton, M., Leiba, B., and T. Narten, "Guidelines for 995 Writing an IANA Considerations Section in RFCs", BCP 26, 996 RFC 8126, DOI 10.17487/RFC8126, June 2017, 997 . 999 [RFC8200] Deering, S. and R. Hinden, "Internet Protocol, Version 6 1000 (IPv6) Specification", STD 86, RFC 8200, 1001 DOI 10.17487/RFC8200, July 2017, 1002 . 1004 [RFC8201] McCann, J., Deering, S., Mogul, J., and R. Hinden, Ed., 1005 "Path MTU Discovery for IP version 6", STD 87, RFC 8201, 1006 DOI 10.17487/RFC8201, July 2017, 1007 . 1009 [RFC8899] Fairhurst, G., Jones, T., Tuexen, M., Ruengeler, I., and 1010 T. Voelker, "Packetization Layer Path MTU Discovery for 1011 Datagram Transports", RFC 8899, DOI 10.17487/RFC8899, 1012 September 2020, . 1014 Appendix A. Example Of An Encapsulated IP Packet Flow 1016 Below an example inner IP packet flow within the encapsulating tunnel 1017 packet stream is shown. Notice how encapsulated IP packets can start 1018 and end anywhere, and more than one or less than 1 may occur in a 1019 single encapsulating packet. 1021 Offset: 0 Offset: 100 Offset: 2900 Offset: 1400 1022 [ ESP1 (1500) ][ ESP2 (1500) ][ ESP3 (1500) ][ ESP4 (1500) ] 1023 [--800--][--800--][60][-240-][--4000----------------------][pad] 1025 Figure 3: Inner and Outer Packet Flow 1027 The encapsulated IP packet flow (lengths include IP header and 1028 payload) is as follows: an 800 octet packet, an 800 octet packet, a 1029 60 octet packet, a 240 octet packet, a 4000 octet packet. 1031 The "BlockOffset" values in the 4 IP-TFS payload headers for this 1032 packet flow would thus be: 0, 100, 2900, 1400 respectively. The 1033 first encapsulating packet ESP1 has a zero "BlockOffset" which points 1034 at the IP data block immediately following the IP-TFS header. The 1035 following packet ESP2s "BlockOffset" points inward 100 octets to the 1036 start of the 60 octet data block. The third encapsulating packet 1037 ESP3 contains the middle portion of the 4000 octet data block so the 1038 offset points past its end and into the forth encapsulating packet. 1039 The fourth packet ESP4s offset is 1400 pointing at the padding which 1040 follows the completion of the continued 4000 octet packet. 1042 Appendix B. A Send and Loss Event Rate Calculation 1044 The current best practice indicates that congestion control SHOULD be 1045 done in a TCP friendly way. A TCP friendly congestion control 1046 algorithm is described in [RFC5348]. For this IP-TFS use case (as 1047 with [RFC4342]) the (fixed) packet size is used as the segment size 1048 for the algorithm. The main formula in the algorithm for the send 1049 rate is then as follows: 1051 1 1052 X = ----------------------------------------------- 1053 R * (sqrt(2*p/3) + 12*sqrt(3*p/8)*p*(1+32*p^2)) 1055 Where "X" is the send rate in packets per second, "R" is the round 1056 trip time estimate and "p" is the loss event rate (the inverse of 1057 which is provided by the receiver). 1059 In addition the algorithm in [RFC5348] also uses an "X_recv" value 1060 (the receiver's receive rate). For IP-TFS one MAY set this value 1061 according to the sender's current tunnel send-rate ("X"). 1063 The IP-TFS receiver, having the RTT estimate from the sender can use 1064 the same method as described in [RFC5348] and [RFC4342] to collect 1065 the loss intervals and calculate the loss event rate value using the 1066 weighted average as indicated. The receiver communicates the inverse 1067 of this value back to the sender in the AGGFRAG_PAYLOAD payload 1068 header field "LossEventRate". 1070 The IP-TFS sender now has both the "R" and "p" values and can 1071 calculate the correct sending rate. If following [RFC5348] the 1072 sender SHOULD also use the slow start mechanism described therein 1073 when the IP-TFS SA is first established. 1075 Appendix C. Comparisons of IP-TFS 1077 C.1. Comparing Overhead 1079 C.1.1. IP-TFS Overhead 1081 The overhead of IP-TFS is 40 bytes per outer packet. Therefore the 1082 octet overhead per inner packet is 40 divided by the number of outer 1083 packets required (fractional allowed). The overhead as a percentage 1084 of inner packet size is a constant based on the Outer MTU size. 1086 OH = 40 / Outer Payload Size / Inner Packet Size 1087 OH % of Inner Packet Size = 100 * OH / Inner Packet Size 1088 OH % of Inner Packet Size = 4000 / Outer Payload Size 1089 Type IP-TFS IP-TFS IP-TFS 1090 MTU 576 1500 9000 1091 PSize 536 1460 8960 1092 ------------------------------- 1093 40 7.46% 2.74% 0.45% 1094 576 7.46% 2.74% 0.45% 1095 1500 7.46% 2.74% 0.45% 1096 9000 7.46% 2.74% 0.45% 1098 Figure 4: IP-TFS Overhead as Percentage of Inner Packet Size 1100 C.1.2. ESP with Padding Overhead 1102 The overhead per inner packet for constant-send-rate padded ESP 1103 (i.e., traditional IPsec TFC) is 36 octets plus any padding, unless 1104 fragmentation is required. 1106 When fragmentation of the inner packet is required to fit in the 1107 outer IPsec packet, overhead is the number of outer packets required 1108 to carry the fragmented inner packet times both the inner IP overhead 1109 (20) and the outer packet overhead (36) minus the initial inner IP 1110 overhead plus any required tail padding in the last encapsulation 1111 packet. The required tail padding is the number of required packets 1112 times the difference of the Outer Payload Size and the IP Overhead 1113 minus the Inner Payload Size. So: 1115 Inner Paylaod Size = IP Packet Size - IP Overhead 1116 Outer Payload Size = MTU - IPsec Overhead 1118 Inner Payload Size 1119 NF0 = ---------------------------------- 1120 Outer Payload Size - IP Overhead 1122 NF = CEILING(NF0) 1124 OH = NF * (IP Overhead + IPsec Overhead) 1125 - IP Overhead 1126 + NF * (Outer Payload Size - IP Overhead) 1127 - Inner Payload Size 1129 OH = NF * (IPsec Overhead + Outer Payload Size) 1130 - (IP Overhead + Inner Payload Size) 1132 OH = NF * (IPsec Overhead + Outer Payload Size) 1133 - Inner Packet Size 1135 C.2. Overhead Comparison 1137 The following tables collect the overhead values for some common L3 1138 MTU sizes in order to compare them. The first table is the number of 1139 octets of overhead for a given L3 MTU sized packet. The second table 1140 is the percentage of overhead in the same MTU sized packet. 1142 Type ESP+Pad ESP+Pad ESP+Pad IP-TFS IP-TFS IP-TFS 1143 L3 MTU 576 1500 9000 576 1500 9000 1144 PSize 540 1464 8964 536 1460 8960 1145 ----------------------------------------------------------- 1146 40 500 1424 8924 3.0 1.1 0.2 1147 128 412 1336 8836 9.6 3.5 0.6 1148 256 284 1208 8708 19.1 7.0 1.1 1149 536 4 928 8428 40.0 14.7 2.4 1150 576 576 888 8388 43.0 15.8 2.6 1151 1460 268 4 7504 109.0 40.0 6.5 1152 1500 228 1500 7464 111.9 41.1 6.7 1153 8960 1408 1540 4 668.7 245.5 40.0 1154 9000 1368 1500 9000 671.6 246.6 40.2 1156 Figure 5: Overhead comparison in octets 1158 Type ESP+Pad ESP+Pad ESP+Pad IP-TFS IP-TFS IP-TFS 1159 MTU 576 1500 9000 576 1500 9000 1160 PSize 540 1464 8964 536 1460 8960 1161 ----------------------------------------------------------- 1162 40 1250.0% 3560.0% 22310.0% 7.46% 2.74% 0.45% 1163 128 321.9% 1043.8% 6903.1% 7.46% 2.74% 0.45% 1164 256 110.9% 471.9% 3401.6% 7.46% 2.74% 0.45% 1165 536 0.7% 173.1% 1572.4% 7.46% 2.74% 0.45% 1166 576 100.0% 154.2% 1456.2% 7.46% 2.74% 0.45% 1167 1460 18.4% 0.3% 514.0% 7.46% 2.74% 0.45% 1168 1500 15.2% 100.0% 497.6% 7.46% 2.74% 0.45% 1169 8960 15.7% 17.2% 0.0% 7.46% 2.74% 0.45% 1170 9000 15.2% 16.7% 100.0% 7.46% 2.74% 0.45% 1172 Figure 6: Overhead as Percentage of Inner Packet Size 1174 C.3. Comparing Available Bandwidth 1176 Another way to compare the two solutions is to look at the amount of 1177 available bandwidth each solution provides. The following sections 1178 consider and compare the percentage of available bandwidth. For the 1179 sake of providing a well understood baseline normal (unencrypted) 1180 Ethernet as well as normal ESP values are included. 1182 C.3.1. Ethernet 1184 In order to calculate the available bandwidth the per packet overhead 1185 is calculated first. The total overhead of Ethernet is 14+4 octets 1186 of header and CRC plus and additional 20 octets of framing (preamble, 1187 start, and inter-packet gap) for a total of 38 octets. Additionally 1188 the minimum payload is 46 octets. 1190 Size E + P E + P E + P IPTFS IPTFS IPTFS Enet ESP 1191 MTU 590 1514 9014 590 1514 9014 any any 1192 OH 74 74 74 78 78 78 38 74 1193 ------------------------------------------------------------ 1194 40 614 1538 9038 45 42 40 84 114 1195 128 614 1538 9038 146 134 129 166 202 1196 256 614 1538 9038 293 269 258 294 330 1197 536 614 1538 9038 614 564 540 574 610 1198 576 1228 1538 9038 659 606 581 614 650 1199 1460 1842 1538 9038 1672 1538 1472 1498 1534 1200 1500 1842 3076 9038 1718 1580 1513 1538 1574 1201 8960 11052 10766 9038 10263 9438 9038 8998 9034 1202 9000 11052 10766 18076 10309 9480 9078 9038 9074 1204 Figure 7: L2 Octets Per Packet 1206 Size E + P E + P E + P IPTFS IPTFS IPTFS Enet ESP 1207 MTU 590 1514 9014 590 1514 9014 any any 1208 OH 74 74 74 78 78 78 38 74 1209 -------------------------------------------------------------- 1210 40 2.0M 0.8M 0.1M 27.3M 29.7M 31.0M 14.9M 11.0M 1211 128 2.0M 0.8M 0.1M 8.5M 9.3M 9.7M 7.5M 6.2M 1212 256 2.0M 0.8M 0.1M 4.3M 4.6M 4.8M 4.3M 3.8M 1213 536 2.0M 0.8M 0.1M 2.0M 2.2M 2.3M 2.2M 2.0M 1214 576 1.0M 0.8M 0.1M 1.9M 2.1M 2.2M 2.0M 1.9M 1215 1460 678K 812K 138K 747K 812K 848K 834K 814K 1216 1500 678K 406K 138K 727K 791K 826K 812K 794K 1217 8960 113K 116K 138K 121K 132K 138K 138K 138K 1218 9000 113K 116K 69K 121K 131K 137K 138K 137K 1220 Figure 8: Packets Per Second on 10G Ethernet 1222 Size E + P E + P E + P IPTFS IPTFS IPTFS Enet ESP 1223 590 1514 9014 590 1514 9014 any any 1224 74 74 74 78 78 78 38 74 1225 ---------------------------------------------------------------------- 1226 40 6.51% 2.60% 0.44% 87.30% 94.93% 99.14% 47.62% 35.09% 1227 128 20.85% 8.32% 1.42% 87.30% 94.93% 99.14% 77.11% 63.37% 1228 256 41.69% 16.64% 2.83% 87.30% 94.93% 99.14% 87.07% 77.58% 1229 536 87.30% 34.85% 5.93% 87.30% 94.93% 99.14% 93.38% 87.87% 1230 576 46.91% 37.45% 6.37% 87.30% 94.93% 99.14% 93.81% 88.62% 1231 1460 79.26% 94.93% 16.15% 87.30% 94.93% 99.14% 97.46% 95.18% 1232 1500 81.43% 48.76% 16.60% 87.30% 94.93% 99.14% 97.53% 95.30% 1233 8960 81.07% 83.22% 99.14% 87.30% 94.93% 99.14% 99.58% 99.18% 1234 9000 81.43% 83.60% 49.79% 87.30% 94.93% 99.14% 99.58% 99.18% 1236 Figure 9: Percentage of Bandwidth on 10G Ethernet 1238 A sometimes unexpected result of using IP-TFS (or any packet 1239 aggregating tunnel) is that, for small to medium sized packets, the 1240 available bandwidth is actually greater than native Ethernet. This 1241 is due to the reduction in Ethernet framing overhead. This increased 1242 bandwidth is paid for with an increase in latency. This latency is 1243 the time to send the unrelated octets in the outer tunnel frame. The 1244 following table illustrates the latency for some common values on a 1245 10G Ethernet link. The table also includes latency introduced by 1246 padding if using ESP with padding. 1248 ESP+Pad ESP+Pad IP-TFS IP-TFS 1249 1500 9000 1500 9000 1251 ------------------------------------------ 1252 40 1.14 us 7.14 us 1.17 us 7.17 us 1253 128 1.07 us 7.07 us 1.10 us 7.10 us 1254 256 0.97 us 6.97 us 1.00 us 7.00 us 1255 536 0.74 us 6.74 us 0.77 us 6.77 us 1256 576 0.71 us 6.71 us 0.74 us 6.74 us 1257 1460 0.00 us 6.00 us 0.04 us 6.04 us 1258 1500 1.20 us 5.97 us 0.00 us 6.00 us 1260 Figure 10: Added Latency 1262 Notice that the latency values are very similar between the two 1263 solutions; however, whereas IP-TFS provides for constant high 1264 bandwidth, in some cases even exceeding native Ethernet, ESP with 1265 padding often greatly reduces available bandwidth. 1267 Appendix D. Acknowledgements 1269 We would like to thank Don Fedyk for help in reviewing and editing 1270 this work. We would also like to thank Valery Smyslov for reviews 1271 and suggestions for improvements as well as Joseph Touch for the 1272 transport area review and suggested improvements. 1274 Appendix E. Contributors 1276 The following people made significant contributions to this document. 1278 Lou Berger 1279 LabN Consulting, L.L.C. 1281 Email: lberger@labn.net 1283 Author's Address 1285 Christian Hopps 1286 LabN Consulting, L.L.C. 1288 Email: chopps@chopps.org