Debian Codegraph @

rfc7383.html

raw
1 <pre>Internet Engineering Task Force (IETF) V. Smyslov
2 Request for Comments: 7383 ELVIS-PLUS
3 Category: Standards Track November 2014
4 ISSN: 2070-1721
5
6
7 <span class="h1">Internet Key Exchange Protocol Version 2 (IKEv2) Message Fragmentation</span>
8
9 Abstract
10
11 This document describes a way to avoid IP fragmentation of large
12 Internet Key Exchange Protocol version 2 (IKEv2) messages. This
13 allows IKEv2 messages to traverse network devices that do not allow
14 IP fragments to pass through.
15
16 Status of This Memo
17
18 This is an Internet Standards Track document.
19
20 This document is a product of the Internet Engineering Task Force
21 (IETF). It represents the consensus of the IETF community. It has
22 received public review and has been approved for publication by the
23 Internet Engineering Steering Group (IESG). Further information on
24 Internet Standards is available in <a href="./rfc5741#section-2">Section&nbsp;2 of RFC 5741</a>.
25
26 Information about the current status of this document, any errata,
27 and how to provide feedback on it may be obtained at
28 <a href="http://www.rfc-editor.org/info/rfc7383">http://www.rfc-editor.org/info/rfc7383</a>.
29
30 Copyright Notice
31
32 Copyright (c) 2014 IETF Trust and the persons identified as the
33 document authors. All rights reserved.
34
35 This document is subject to <a href="https://www.rfc-editor.org/bcp/bcp78">BCP 78</a> and the IETF Trust's Legal
36 Provisions Relating to IETF Documents
37 (<a href="http://trustee.ietf.org/license-info">http://trustee.ietf.org/license-info</a>) in effect on the date of
38 publication of this document. Please review these documents
39 carefully, as they describe your rights and restrictions with respect
40 to this document. Code Components extracted from this document must
41 include Simplified BSD License text as described in Section 4.e of
42 the Trust Legal Provisions and are provided without warranty as
43 described in the Simplified BSD License.
44
45
46
47
48
49
50
51
52 <span class="grey">Smyslov Standards Track [Page 1]</span></pre>
53 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-2" ></span>
54 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
55
56
57 Table of Contents
58
59 <a href="#section-1">1</a>. Introduction ....................................................<a href="#page-2">2</a>
60 <a href="#section-1.1">1.1</a>. Problem Description ........................................<a href="#page-2">2</a>
61 <a href="#section-1.2">1.2</a>. Proposed Solution ..........................................<a href="#page-3">3</a>
62 <a href="#section-1.3">1.3</a>. Conventions Used in This Document ..........................<a href="#page-4">4</a>
63 <a href="#section-2">2</a>. Protocol Details ................................................<a href="#page-4">4</a>
64 <a href="#section-2.1">2.1</a>. Overview ...................................................<a href="#page-4">4</a>
65 <a href="#section-2.2">2.2</a>. Limitations ................................................<a href="#page-4">4</a>
66 <a href="#section-2.3">2.3</a>. Negotiation ................................................<a href="#page-5">5</a>
67 <a href="#section-2.4">2.4</a>. Using IKE Fragmentation ....................................<a href="#page-5">5</a>
68 <a href="#section-2.5">2.5</a>. Fragmenting Message ........................................<a href="#page-6">6</a>
69 <a href="#section-2.5.1">2.5.1</a>. Selecting Fragment Size .............................<a href="#page-8">8</a>
70 <a href="#section-2.5.2">2.5.2</a>. PMTU Discovery ......................................<a href="#page-9">9</a>
71 2.5.3. Fragmenting Messages Containing Unprotected
72 Payloads ...........................................<a href="#page-11">11</a>
73 <a href="#section-2.6">2.6</a>. Receiving IKE Fragment Message ............................<a href="#page-11">11</a>
74 <a href="#section-2.6.1">2.6.1</a>. Replay Detection and Retransmissions ...............<a href="#page-13">13</a>
75 <a href="#section-3">3</a>. Interaction with Other IKE Extensions ..........................<a href="#page-14">14</a>
76 <a href="#section-4">4</a>. Transport Considerations .......................................<a href="#page-14">14</a>
77 <a href="#section-5">5</a>. Security Considerations ........................................<a href="#page-15">15</a>
78 <a href="#section-6">6</a>. IANA Considerations ............................................<a href="#page-16">16</a>
79 <a href="#section-7">7</a>. References .....................................................<a href="#page-16">16</a>
80 <a href="#section-7.1">7.1</a>. Normative References ......................................<a href="#page-16">16</a>
81 <a href="#section-7.2">7.2</a>. Informative References ....................................<a href="#page-16">16</a>
82 <a href="#appendix-A">Appendix A</a>. Design Rationale ......................................<a href="#page-19">19</a>
83 <a href="#appendix-B">Appendix B</a>. Correlation between IP Datagram Size and Encrypted
84 Payload Content Size ..................................<a href="#page-19">19</a>
85 Acknowledgements ..................................................<a href="#page-20">20</a>
86 Author's Address ..................................................<a href="#page-20">20</a>
87
88 <span class="h2"><a class="selflink" id="section-1" href="#section-1">1</a>. Introduction</span>
89
90 <span class="h3"><a class="selflink" id="section-1.1" href="#section-1.1">1.1</a>. Problem Description</span>
91
92 The Internet Key Exchange Protocol version 2 (IKEv2), specified in
93 [<a href="./rfc7296" title="&quot;Internet Key Exchange Protocol Version 2 (IKEv2)&quot;">RFC7296</a>], uses UDP as a transport for its messages. Most IKEv2
94 messages are relatively small, usually below several hundred bytes.
95 A notable exception is the IKE_AUTH exchange, which requires fairly
96 large messages, up to several KB, especially when certificates are
97 transferred. When the IKE message size exceeds the path MTU, it gets
98 fragmented at the IP level. The problem is that some network
99 devices, specifically some NAT boxes, do not allow IP fragments to
100 pass through. This apparently blocks IKE communication and,
101 therefore, prevents peers from establishing an IPsec Security
102 Association (SA). <a href="./rfc7296#section-2">Section&nbsp;2 of [RFC7296]</a> discusses the impact of IP
103 fragmentation on IKEv2 and acknowledges this problem.
104
105
106
107
108 <span class="grey">Smyslov Standards Track [Page 2]</span></pre>
109 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-3" ></span>
110 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
111
112
113 Widespread deployment of Carrier-Grade NATs (CGNs) introduces new
114 challenges. [<a href="./rfc6888" title="&quot;Common Requirements for Carrier-Grade NATs (CGNs)&quot;">RFC6888</a>] describes requirements for CGNs. It states
115 that CGNs must comply with <a href="./rfc4787#section-11">Section&nbsp;11 of [RFC4787]</a>, which requires
116 NATs to support receiving IP fragments (REQ-14). In real life,
117 fulfillment of this requirement creates an additional burden in terms
118 of memory, especially for high-capacity devices used in CGNs. It was
119 found by people deploying IKE that more and more ISPs use equipment
120 that drops IP fragments, thereby violating this requirement.
121
122 Security researchers have found, and continue to find, attack vectors
123 that rely on IP fragmentation. For these reasons, and also as
124 articulated in [<a href="#ref-FRAGDROP" title="&quot;Why Operators Filter Fragments and What It Implies&quot;">FRAGDROP</a>], many network operators filter all IPv6
125 fragments. Also, the default behavior of many currently deployed
126 firewalls is to discard IPv6 fragments.
127
128 In one recent study [<a href="#ref-BLACKHOLES">BLACKHOLES</a>], two researchers utilized a
129 measurement network to measure fragment filtering. They sent
130 packets, fragmented to the minimum MTU of 1280, to 502 IPv6-enabled
131 and reachable probes. They found that during any given trial period,
132 ten percent of the probes did not receive fragmented packets.
133
134 Thus, this problem is valid for both IPv4 and IPv6 and may be caused
135 by either deficiency of network devices or operational choice.
136
137 <span class="h3"><a class="selflink" id="section-1.2" href="#section-1.2">1.2</a>. Proposed Solution</span>
138
139 The solution to the problem described in this document is to perform
140 fragmentation of large messages by IKEv2 itself and replace them with
141 a series of smaller messages. In this case, the resulting IP
142 datagrams will be small enough so that no fragmentation at the IP
143 level will take place.
144
145 The primary goal of this solution is to allow IKEv2 to operate in
146 environments that might block IP fragments. This goal does not
147 assume that IP fragmentation should be avoided completely, but only
148 in those cases when it interferes with IKE operations. However, this
149 solution could be used to avoid IP fragmentation in all situations
150 where fragmentation within IKE is applicable, as recommended in
151 <a href="./rfc5405#section-3.2">Section&nbsp;3.2 of [RFC5405]</a>. Avoiding IP fragmentation would be
152 beneficial for IKEv2 in general. The Security Considerations section
153 of [<a href="./rfc7296" title="&quot;Internet Key Exchange Protocol Version 2 (IKEv2)&quot;">RFC7296</a>] mentions exhaustion of the IP reassembly buffers as one
154 of the possible attacks on the protocol. In [<a href="#ref-DOSUDPPROT">DOSUDPPROT</a>], several
155 aspects of attacks on IKE using IP fragmentation are discussed, and
156 one of the defenses it proposes is to perform fragmentation within
157 IKE, similar to the solution described in this document.
158
159
160
161
162
163
164 <span class="grey">Smyslov Standards Track [Page 3]</span></pre>
165 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-4" ></span>
166 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
167
168
169 <span class="h3"><a class="selflink" id="section-1.3" href="#section-1.3">1.3</a>. Conventions Used in This Document</span>
170
171 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
172 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
173 document are to be interpreted as described in [<a href="./rfc2119" title="&quot;Key words for use in RFCs to Indicate Requirement Levels&quot;">RFC2119</a>].
174
175 <span class="h2"><a class="selflink" id="section-2" href="#section-2">2</a>. Protocol Details</span>
176
177 <span class="h3"><a class="selflink" id="section-2.1" href="#section-2.1">2.1</a>. Overview</span>
178
179 The idea of the protocol described in this document is to split large
180 IKEv2 messages into a set of smaller ones, called IKE Fragment
181 messages. Fragmentation takes place before the original message is
182 encrypted and authenticated, so that each IKE Fragment message
183 receives individual protection. On the receiving side, IKE Fragment
184 messages are collected, verified, decrypted, and merged together to
185 get the original message before encryption. See <a href="#appendix-A">Appendix A</a> for
186 details on design rationale.
187
188 <span class="h3"><a class="selflink" id="section-2.2" href="#section-2.2">2.2</a>. Limitations</span>
189
190 Since IKE Fragment messages are cryptographically protected, SK_a and
191 SK_e must already be calculated. In general, it means that the
192 original message can be fragmented if and only if it contains an
193 Encrypted payload.
194
195 This implies that messages of the IKE_SA_INIT exchange cannot be
196 fragmented. In most cases, this is not a problem because IKE_SA_INIT
197 messages are usually small enough to avoid IP fragmentation. But in
198 some cases (advertising a badly structured long list of algorithms,
199 using large Modular Exponentiation (MODP) groups, etc.), these
200 messages may become fairly large and get fragmented at the IP level.
201 In this case, the solution described in this document will not help.
202
203 Among existing IKEv2 extensions, messages of an IKE_SESSION_RESUME
204 exchange, as defined in [<a href="./rfc5723" title="&quot;Internet Key Exchange Protocol Version 2 (IKEv2) Session Resumption&quot;">RFC5723</a>], cannot be fragmented either. See
205 <a href="#section-3">Section 3</a> for details.
206
207 Another limitation is that the minimum size of an IP datagram bearing
208 an IKE Fragment message is about 100 bytes, depending on the
209 algorithms employed. According to [<a href="./rfc0791" title="&quot;Internet Protocol&quot;">RFC0791</a>], the minimum IPv4
210 datagram size that is guaranteed not to be further fragmented is
211 68 bytes. So, even the smallest IKE Fragment messages could be
212 fragmented at the IP level in some circumstances. But such extremely
213 small Path MTU (PMTU) sizes are very rare in real life.
214
215
216
217
218
219
220 <span class="grey">Smyslov Standards Track [Page 4]</span></pre>
221 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-5" ></span>
222 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
223
224
225 <span class="h3"><a class="selflink" id="section-2.3" href="#section-2.3">2.3</a>. Negotiation</span>
226
227 The initiator indicates its support for IKE fragmentation and
228 willingness to use it by including a Notification payload of type
229 IKEV2_FRAGMENTATION_SUPPORTED in the IKE_SA_INIT request message. If
230 the responder also supports this extension and is willing to use it,
231 it includes this notification in the response message.
232
233 Initiator Responder
234 ----------- -----------
235 HDR, SAi1, KEi, Ni,
236 [N(IKEV2_FRAGMENTATION_SUPPORTED)] --&gt;
237
238 &lt;-- HDR, SAr1, KEr, Nr, [CERTREQ],
239 [N(IKEV2_FRAGMENTATION_SUPPORTED)]
240
241 The Notify payload is formatted as follows:
242
243 1 2 3
244 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
245 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
246 | Next Payload |C| RESERVED | Payload Length |
247 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
248 |Protocol ID(=0)| SPI Size (=0) | Notify Message Type |
249 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
250
251 o Protocol ID (1 octet) - MUST be 0.
252
253 o SPI Size (1 octet) - MUST be 0, meaning no Security Parameter
254 Index (SPI) is present.
255
256 o Notify Message Type (2 octets) - MUST be 16430, the value assigned
257 for the IKEV2_FRAGMENTATION_SUPPORTED notification.
258
259 This notification contains no data.
260
261 <span class="h3"><a class="selflink" id="section-2.4" href="#section-2.4">2.4</a>. Using IKE Fragmentation</span>
262
263 IKE fragmentation MUST NOT be used unless both peers have indicated
264 their support for it. After that, it is up to the initiator of each
265 exchange to decide whether or not to use it. The responder usually
266 replies in the same form as the request message, but other
267 considerations might override this.
268
269 The initiator can employ various policies regarding the use of IKE
270 fragmentation. It might first try to send an unfragmented message
271 and resend it as fragmented only if no complete response is received
272 even after several retransmissions. Alternatively, it might choose
273
274
275
276 <span class="grey">Smyslov Standards Track [Page 5]</span></pre>
277 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-6" ></span>
278 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
279
280
281 to always send fragmented messages (however, see <a href="#section-3">Section 3</a>), or it
282 might fragment only large messages and messages that are expected to
283 result in large responses.
284
285 The following general guidelines apply:
286
287 o If either peer has information that a part of the transaction is
288 likely to be fragmented at the IP layer, causing interference with
289 the IKE exchange, that peer SHOULD use IKE fragmentation. This
290 information might be passed from a lower layer, provided by
291 configuration, or derived through heuristics. Examples of
292 heuristics are the lack of a complete response after several
293 retransmissions for the initiator, and receiving repeated
294 retransmissions of the request for the responder.
295
296 o If either peer knows that IKE fragmentation has been used in a
297 previous exchange in the context of the current IKE SA, that peer
298 SHOULD continue to use IKE fragmentation for the messages that are
299 larger than the current fragmentation threshold (see
300 <a href="#section-2.5.1">Section 2.5.1</a>).
301
302 o IKE fragmentation SHOULD NOT be used in cases where IP-layer
303 fragmentation of both the request and response messages is
304 unlikely. For example, there is no point in fragmenting liveness
305 check messages.
306
307 o If none of the above apply, the responder SHOULD respond in the
308 same form (fragmented or not) as the request message to which it
309 is responding. Note that the other guidelines might override this
310 because of information or heuristics available to the responder.
311
312 In most cases, IKE fragmentation will be used in the IKE_AUTH
313 exchange, especially if certificates are employed.
314
315 <span class="h3"><a class="selflink" id="section-2.5" href="#section-2.5">2.5</a>. Fragmenting Message</span>
316
317 Only messages that contain an Encrypted payload are subject to IKE
318 fragmentation. For the purpose of construction of IKE Fragment
319 messages, the original (unencrypted) content of the Encrypted payload
320 is split into chunks. The content is treated as a binary blob and is
321 split regardless of the boundaries of inner payloads. Each of the
322 resulting chunks is treated as an original content of the Encrypted
323 Fragment payload and is then encrypted and authenticated. Thus, the
324 Encrypted Fragment payload contains a chunk of the original content
325 of the Encrypted payload in encrypted form. The cryptographic
326 processing of the Encrypted Fragment payload is identical to that
327
328
329
330
331
332 <span class="grey">Smyslov Standards Track [Page 6]</span></pre>
333 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-7" ></span>
334 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
335
336
337 described in <a href="./rfc7296#section-3.14">Section&nbsp;3.14 of [RFC7296]</a>, as well as documents updating
338 such processing for particular algorithms or modes, such as
339 [<a href="./rfc5282" title="&quot;Using Authenticated Encryption Algorithms with the Encrypted Payload of the Internet Key Exchange version 2 (IKEv2) Protocol&quot;">RFC5282</a>].
340
341 As is the case for the Encrypted payload, the Encrypted Fragment
342 payload, if present in a message, MUST be the last payload in the
343 message.
344
345 The Encrypted Fragment payload is denoted SKF{...}, and its payload
346 type is 53. This payload is also called the "Encrypted and
347 Authenticated Fragment" payload.
348
349 1 2 3
350 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
351 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
352 | Next Payload |C| RESERVED | Payload Length |
353 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
354 | Fragment Number | Total Fragments |
355 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
356 | Initialization Vector |
357 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
358 ~ Encrypted content ~
359 + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
360 | | Padding (0-255 octets) |
361 +-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+
362 | | Pad Length |
363 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
364 ~ Integrity Checksum Data ~
365 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
366
367 Encrypted Fragment Payload
368
369 o Next Payload (1 octet) - in the very first fragment (with Fragment
370 Number equal to 1), this field MUST be set to the payload type of
371 the first inner payload (the same as for the Encrypted payload).
372 In the rest of the Fragment messages (with Fragment Number greater
373 than 1), this field MUST be set to zero.
374
375 o Fragment Number (2 octets, unsigned integer) - current Fragment
376 message number, starting from 1. This field MUST be less than or
377 equal to the next field (Total Fragments). This field MUST NOT be
378 zero.
379
380 o Total Fragments (2 octets, unsigned integer) - number of Fragment
381 messages into which the original message was divided. This field
382 MUST NOT be zero. With PMTU discovery, this field plays an
383 additional role. See <a href="#section-2.5.2">Section 2.5.2</a> for details.
384
385
386
387
388 <span class="grey">Smyslov Standards Track [Page 7]</span></pre>
389 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-8" ></span>
390 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
391
392
393 The other fields are identical to those specified in <a href="./rfc7296#section-3.14">Section&nbsp;3.14 of
394 [RFC7296]</a>.
395
396 When prepending the IKE header to the IKE Fragment messages, it MUST
397 be taken intact from the original message, except for the Length and
398 Next Payload fields. The Length field is adjusted to reflect the
399 length of the IKE Fragment message being constructed, and the Next
400 Payload field is set to the payload type of the first payload in that
401 message (in most cases, it will be the Encrypted Fragment payload).
402 After prepending the IKE header and all payloads that possibly
403 precede the Encrypted payload in the original message (if any; see
404 <a href="#section-2.5.3">Section 2.5.3</a>), the resulting messages are sent to the peer.
405
406 Below is an example of fragmenting a message.
407
408 HDR(MID=n), SK(NextPld=PLD1) {PLD1 ... PLDN}
409
410 Original Message
411
412
413 HDR(MID=n), SKF(NextPld=PLD1, Frag#=1, TotalFrags=m) {...},
414 HDR(MID=n), SKF(NextPld=0, Frag#=2, TotalFrags=m) {...},
415 ...
416 HDR(MID=n), SKF(NextPld=0, Frag#=m, TotalFrags=m) {...}
417
418 IKE Fragment Messages
419
420 <span class="h4"><a class="selflink" id="section-2.5.1" href="#section-2.5.1">2.5.1</a>. Selecting Fragment Size</span>
421
422 When splitting the content of an Encrypted payload into chunks, the
423 sender SHOULD choose their size so that the resulting IP datagrams
424 will be smaller than some fragmentation threshold. Implementations
425 may calculate the fragmentation threshold using various sources of
426 information.
427
428 If the sender has information about the PMTU size, it SHOULD use it.
429 The responder in the exchange may use the maximum size of the
430 received IKE Fragment message IP datagrams as a threshold when
431 constructing a fragmented response. Successful completion of
432 previous exchanges (including those exchanges that cannot employ IKE
433 fragmentation, e.g., IKE_SA_INIT) may be an indication that the
434 fragmentation threshold can be set to the size of the largest message
435 of those messages already sent.
436
437 Otherwise, for messages to be sent over IPv6, it is RECOMMENDED that
438 a value of 1280 bytes as a maximum IP datagram size be used
439 ([<a href="./rfc2460" title="&quot;Internet Protocol, Version 6 (IPv6) Specification&quot;">RFC2460</a>]). For messages to be sent over IPv4, it is RECOMMENDED
440 that a value of 576 bytes as a maximum IP datagram size be used. The
441
442
443
444 <span class="grey">Smyslov Standards Track [Page 8]</span></pre>
445 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-9" ></span>
446 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
447
448
449 presence of tunnels on the path may reduce these values.
450 Implementations may use other values if they are appropriate in the
451 current environment.
452
453 According to [<a href="./rfc0791" title="&quot;Internet Protocol&quot;">RFC0791</a>], the minimum IPv4 datagram size that is
454 guaranteed not to be further fragmented is 68 bytes, but it is
455 generally impossible to use such a small value for the solution
456 described in this document. Using 576 bytes is a compromise -- the
457 value is large enough for the presented solution and small enough to
458 avoid IP fragmentation in most situations. Several other UDP-based
459 protocols (Syslog, DNS, etc.) use 576 bytes as a safe low limit for
460 IP datagram size.
461
462 See <a href="#appendix-B">Appendix B</a> for correlation between IP datagram size and Encrypted
463 payload content size.
464
465 <span class="h4"><a class="selflink" id="section-2.5.2" href="#section-2.5.2">2.5.2</a>. PMTU Discovery</span>
466
467 The amount of traffic that the IKE endpoint produces during the
468 lifetime of an IKE SA is fairly modest -- it is usually below 100 KB
469 within a period of several hours. Most of this traffic consists of
470 relatively short messages -- usually below several hundred bytes. In
471 most cases, the only time when IKE endpoints exchange messages of
472 several KB in size is IKE SA establishment, and often each endpoint
473 sends exactly one such message.
474
475 For the reasons articulated above, implementing PMTU discovery in IKE
476 is OPTIONAL. It is believed that using the values recommended in
477 <a href="#section-2.5.1">Section 2.5.1</a> as a fragmentation threshold will be sufficient in most
478 cases. Using these values could lead to suboptimal fragmentation,
479 but it is acceptable given the amount of traffic IKE produces.
480 Implementations may support PMTU discovery if there are good reasons
481 to do it (for example, if they are intended to be used in
482 environments where the MTU size might be less than the values listed
483 in <a href="#section-2.5.1">Section 2.5.1</a>).
484
485 PMTU discovery in IKE follows recommendations given in <a href="./rfc4821#section-10.4">Section&nbsp;10.4
486 of [RFC4821]</a> with some modifications, induced by the distinctive
487 features of IKE listed above. The difference is that the PMTU search
488 is performed downward, while in [<a href="./rfc4821" title="&quot;Packetization Layer Path MTU Discovery&quot;">RFC4821</a>] it is performed upward.
489 The reason for this change is that IKE usually sends large messages
490 only when the IKE SA is being established, and in many cases there is
491 only one such message. If the probing were performed upward, this
492 message would be fragmented using the smallest allowable threshold,
493 and usually all other messages are small enough to avoid IP
494 fragmentation, so continued probing would be of little value.
495
496
497
498
499
500 <span class="grey">Smyslov Standards Track [Page 9]</span></pre>
501 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-10" ></span>
502 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
503
504
505 It is the initiator of the exchange who performs PMTU discovery.
506 This is done by probing several values of fragmentation threshold.
507 Implementations MUST be prepared to probe in every exchange that
508 utilizes IKE fragmentation to deal with possible changes in path MTU
509 over time. While doing probes, it MUST start from larger values and
510 refragment the original message, using the next smaller value of the
511 threshold if it did not receive a response in a reasonable time after
512 several retransmissions. The exact number of retransmissions and
513 length of timeouts are not covered in this specification because they
514 do not affect interoperability. However, the timeout interval is
515 supposed to be relatively short, so that unsuccessful probes would
516 not delay IKE operations too much. Performing a few retries within
517 several seconds for each probe seems appropriate, but different
518 environments may require different rules. When starting a new probe,
519 the node MUST reset its retransmission timers so that if it employs
520 exponential back-off the timers will start over. After reaching the
521 smallest allowed value for the fragmentation threshold, an
522 implementation MUST continue retransmitting until the exchange either
523 completes or times out using some timeout interval as discussed in
524 <a href="./rfc7296#section-2.4">Section&nbsp;2.4 of [RFC7296]</a>.
525
526 PMTU discovery in IKE is supposed to be coarse-grained, i.e., it is
527 expected that a node will try only a few fragmentation thresholds in
528 order to minimize delays caused by unsuccessful probes. If path MTU
529 information is not yet available, the endpoint may use the link MTU
530 size when it starts probing. In subsequent exchanges, the node
531 should start with the current value of the fragmentation threshold.
532
533 If an implementation is capable of receiving ICMP error messages, it
534 can additionally utilize classic PMTU discovery methods, as described
535 in [<a href="./rfc1191" title="&quot;Path MTU discovery&quot;">RFC1191</a>] and [<a href="./rfc1981" title="&quot;Path MTU Discovery for IP version 6&quot;">RFC1981</a>]. In particular, if the initiator receives
536 a Packet Too Big error in response to the probe, and it contains a
537 smaller value than the current fragmentation threshold, then the
538 initiator SHOULD stop retransmitting the probe and SHOULD select a
539 new value for the fragmentation threshold that is less than or equal
540 to the value from the ICMP message and meets the requirements listed
541 below.
542
543 In the case of PMTU discovery, the Total Fragments field is used to
544 distinguish between different sets of fragments, i.e., the sets that
545 were created by fragmenting the original message using different
546 fragmentation thresholds. Since the sender starts from larger
547 fragments and then makes them smaller, the value in the Total
548 Fragments field increases with each new probe. When selecting the
549 next smaller value for the fragmentation threshold, the sender MUST
550 ensure that the value in the Total Fragments field is really
551 increased. This requirement should not be a problem for the sender,
552 because PMTU discovery in IKE is supposed to be coarse-grained, so
553
554
555
556 <span class="grey">Smyslov Standards Track [Page 10]</span></pre>
557 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-11" ></span>
558 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
559
560
561 the difference between previous and next fragmentation thresholds
562 should be significant anyway. The need to distinguish between the
563 sets is vital for the receiver, since receiving a valid fragment from
564 a newer set means that it has to start the reassembly process over
565 and not mix fragments from different sets.
566
567 <span class="h4"><a class="selflink" id="section-2.5.3" href="#section-2.5.3">2.5.3</a>. Fragmenting Messages Containing Unprotected Payloads</span>
568
569 Currently, there are no IKEv2 exchanges that define messages,
570 containing both unprotected payloads and payloads, that are protected
571 by the Encrypted payload. However, IKEv2 does not prohibit such
572 construction. If some future IKEv2 extension defines such a message
573 and it needs to be fragmented, all unprotected payloads MUST be
574 placed in the first fragment (with the Fragment Number field equal to
575 1), along with the Encrypted Fragment payload, which MUST be present
576 in every IKE Fragment message and be the last payload in it.
577
578 Below is an example of a fragmenting message that contains both
579 protected and unprotected payloads.
580
581 HDR(MID=n), PLD0, SK(NextPld=PLD1) {PLD1 ... PLDN}
582
583 Original Message
584
585
586 HDR(MID=n), PLD0, SKF(NextPld=PLD1, Frag#=1, TotalFrags=m) {...},
587 HDR(MID=n), SKF(NextPld=0, Frag#=2, TotalFrags=m) {...},
588 ...
589 HDR(MID=n), SKF(NextPld=0, Frag#=m, TotalFrags=m) {...}
590
591 IKE Fragment Messages
592
593 Note that the size of each IP datagram bearing IKE Fragment messages
594 should not exceed the fragmentation threshold, including the first
595 one, that contains unprotected payloads. This will reduce the size
596 of the Encrypted Fragment payload content in the first IKE Fragment
597 message to accommodate all unprotected payloads. In an extreme case,
598 the Encrypted Fragment payload will contain no data, but it still
599 must be present in the message, because only its presence allows the
600 receiver to determine that the sender has used IKE fragmentation.
601
602 <span class="h3"><a class="selflink" id="section-2.6" href="#section-2.6">2.6</a>. Receiving IKE Fragment Message</span>
603
604 The receiver identifies the IKE Fragment message by the presence of
605 an Encrypted Fragment payload in it. In most cases, it will be the
606 first and only payload in the message; however, this may not be true
607 for some hypothetical IKE exchanges (see <a href="#section-2.5.3">Section 2.5.3</a>).
608
609
610
611
612 <span class="grey">Smyslov Standards Track [Page 11]</span></pre>
613 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-12" ></span>
614 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
615
616
617 Upon receiving the IKE Fragment message, the following actions are
618 performed:
619
620 o Check message validity - in particular, check whether the values
621 in the Fragment Number and the Total Fragments fields in the
622 Encrypted Fragment payload are valid. The following tests need to
623 be performed.
624
625 * check that the Fragment Number and the Total Fragments fields
626 contain non-zero values
627
628 * check that the value in the Fragment Number field is less than
629 or equal to the value in the Total Fragments field
630
631 * if reassembling has already started, check that the value in
632 the Total Fragments field is equal to or greater than the Total
633 Fragments field in the fragments that have already been stored
634 in the reassembling queue
635
636 If any of these tests fail, the message MUST be silently
637 discarded.
638
639 o Check that this IKE Fragment message is new for the receiver and
640 not a replay. If an IKE Fragment message with the same Message
641 ID, Fragment Number, and Total Fragments fields is already present
642 in the reassembling queue, this message is considered a replay and
643 MUST be silently discarded.
644
645 o Verify IKE Fragment message authenticity by checking the Integrity
646 Check Value (ICV) in the Encrypted Fragment payload. If the ICV
647 check fails, the message MUST be silently discarded.
648
649 o If reassembling is not finished yet and the Total Fragments field
650 in the received fragment is greater than the Total Fragments field
651 in those fragments that are in the reassembling queue, the
652 receiver MUST discard all received fragments and start the
653 reassembly process over with just the received IKE Fragment
654 message.
655
656 o Store the message in the reassembling queue waiting for the rest
657 of the fragments to arrive.
658
659 When all IKE Fragment messages (as indicated in the Total Fragments
660 field) are received, the decrypted content of all Encrypted Fragment
661 payloads is merged together to form the content of the original
662 Encrypted payload and, therefore, along with the IKE header and
663
664
665
666
667
668 <span class="grey">Smyslov Standards Track [Page 12]</span></pre>
669 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-13" ></span>
670 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
671
672
673 unprotected payloads (if any), the original message. Then, it is
674 processed as if it was received, verified, and decrypted as a regular
675 IKE message.
676
677 If the receiver does not get all IKE fragments needed to reassemble
678 the original message within a timeout interval, it MUST discard all
679 IKE Fragment messages received so far for the exchange. The next
680 actions depend on the role of the receiver in the exchange.
681
682 o The initiator acts as described in <a href="./rfc7296#section-2.1">Section&nbsp;2.1 of [RFC7296]</a>. It
683 either retransmits the fragmented request message or deems the IKE
684 SA to have failed and deletes it. The number of retransmits and
685 length of timeouts for the initiator are not covered in this
686 specification, since they are assumed to be the same as in a
687 regular IKEv2 exchange and are discussed in <a href="./rfc7296#section-2.4">Section&nbsp;2.4 of
688 [RFC7296]</a>.
689
690 o The responder in this case acts as if no request message was
691 received. It would delete any memory of the incomplete request
692 message and not treat it as an IKE SA failure. It is RECOMMENDED
693 that the reassembling timeout for the responder be equal to the
694 time interval that the implementation waits before completely
695 giving up when acting as the initiator of an exchange.
696 <a href="./rfc7296#section-2.4">Section&nbsp;2.4 of [RFC7296]</a> gives recommendations for selecting this
697 interval. Implementations can use a shorter timeout to conserve
698 memory.
699
700 <span class="h4"><a class="selflink" id="section-2.6.1" href="#section-2.6.1">2.6.1</a>. Replay Detection and Retransmissions</span>
701
702 According to <a href="./rfc7296#section-2.2">Section&nbsp;2.2 of [RFC7296]</a>, the Message ID is used, in
703 particular, to identify retransmissions of IKE messages. Each
704 request or response message, sent by either side, must have a unique
705 Message ID, or be considered a retransmission otherwise. This logic
706 has already been updated by [<a href="./rfc6311" title="&quot;Protocol Support for High Availability of IKEv2/ IPsec&quot;">RFC6311</a>], which deliberately allows any
707 number of messages with zero Message ID. This document also updates
708 this logic for those situations where IKE fragmentation is in use.
709
710 If an incoming message contains an Encrypted Fragment payload, the
711 values of the Fragment Number and Total Fragments fields MUST be used
712 along with the Message ID to detect retransmissions and replays.
713
714 If the responder receives a retransmitted fragment of a request when
715 it has already processed that request and has sent back a response,
716 that event MUST only trigger a retransmission of the response message
717 (fragmented or not) if the Fragment Number field in the received
718 fragment is set to 1; otherwise, it MUST be ignored.
719
720
721
722
723
724 <span class="grey">Smyslov Standards Track [Page 13]</span></pre>
725 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-14" ></span>
726 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
727
728
729 <span class="h2"><a class="selflink" id="section-3" href="#section-3">3</a>. Interaction with Other IKE Extensions</span>
730
731 IKE fragmentation is compatible with most IKE extensions, such as IKE
732 Session Resumption ([<a href="./rfc5723" title="&quot;Internet Key Exchange Protocol Version 2 (IKEv2) Session Resumption&quot;">RFC5723</a>]), the Quick Crash Detection Method
733 ([<a href="./rfc6290" title="&quot;A Quick Crash Detection Method for the Internet Key Exchange Protocol (IKE)&quot;">RFC6290</a>]), and so on. It neither affects their operation nor is
734 affected by them. It is believed that IKE fragmentation will also be
735 compatible with future IKE extensions, if they follow general
736 principles of formatting, sending, and receiving IKE messages, as
737 described in [<a href="./rfc7296" title="&quot;Internet Key Exchange Protocol Version 2 (IKEv2)&quot;">RFC7296</a>].
738
739 When IKE fragmentation is used with IKE Session Resumption
740 ([<a href="./rfc5723" title="&quot;Internet Key Exchange Protocol Version 2 (IKEv2) Session Resumption&quot;">RFC5723</a>]), messages of an IKE_SESSION_RESUME exchange cannot be
741 fragmented, since they do not contain an Encrypted payload. These
742 messages may be large due to the ticket size. To avoid IP
743 fragmentation in this situation, it is recommended that smaller
744 tickets be used, e.g., by utilizing a "ticket by reference" approach
745 instead of "ticket by value".
746
747 Protocol Support for High Availability of IKEv2/IPsec, described in
748 [<a href="./rfc6311" title="&quot;Protocol Support for High Availability of IKEv2/ IPsec&quot;">RFC6311</a>], requires special care when deciding whether to fragment an
749 IKE message or not. Since it deliberately allows any number of
750 synchronization exchanges to have the same Message ID, namely zero,
751 standard IKEv2 replay detection logic, based on checking the Message
752 ID, is not applicable for such messages, and the receiver has to
753 check message content to detect replays. When implementing IKE
754 fragmentation along with [<a href="./rfc6311" title="&quot;Protocol Support for High Availability of IKEv2/ IPsec&quot;">RFC6311</a>], IKE Message ID Synchronization
755 messages MUST NOT be sent fragmented, to simplify the receiver's task
756 of detecting replays. Fortunately, these messages are small, and
757 there is no point in fragmenting them anyway.
758
759 <span class="h2"><a class="selflink" id="section-4" href="#section-4">4</a>. Transport Considerations</span>
760
761 With IKE fragmentation, if any single IKE Fragment message gets lost,
762 the receiver becomes unable to reassemble the original message. So,
763 in general, using IKE fragmentation implies a higher probability that
764 the message will not be delivered to the peer. Although in most
765 network environments the difference will be insignificant, on some
766 lossy networks it may become noticeable. When using IKE
767 fragmentation, implementations MAY use longer timeouts and do more
768 retransmits than usual before considering the peer dead.
769
770 Note that Fragment messages are not individually acknowledged. The
771 response Fragment messages are all sent back together only when all
772 fragments of the request are received, and the original request
773 message is reassembled and successfully processed.
774
775
776
777
778
779
780 <span class="grey">Smyslov Standards Track [Page 14]</span></pre>
781 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-15" ></span>
782 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
783
784
785 <span class="h2"><a class="selflink" id="section-5" href="#section-5">5</a>. Security Considerations</span>
786
787 Most of the security considerations for IKE fragmentation are the
788 same as those for the base IKEv2 protocol described in [<a href="./rfc7296" title="&quot;Internet Key Exchange Protocol Version 2 (IKEv2)&quot;">RFC7296</a>].
789 This extension introduces the Encrypted Fragment payload to protect
790 the content of an IKE Message Fragment. This allows the receiver to
791 individually check the authenticity of fragments, thus protecting
792 peers from a DoS attack.
793
794 The Security Considerations section of [<a href="./rfc7296" title="&quot;Internet Key Exchange Protocol Version 2 (IKEv2)&quot;">RFC7296</a>] mentions a possible
795 attack on IKE where an attacker could prevent an exchange from
796 completing by exhausting the IP reassembly buffers. The mechanism
797 described in this document allows IKE to avoid IP fragmentation and
798 therefore increases its robustness to DoS attacks.
799
800 The following attack is possible with IKE fragmentation. An attacker
801 can initiate an IKE_SA_INIT exchange, complete it, compute SK_a and
802 SK_e, and then send a large but still incomplete set of IKE_AUTH
803 fragments. These fragments will pass the ICV check and will be
804 stored in reassembly buffers, but since the set is incomplete, the
805 reassembling will never succeed and eventually will time out. If the
806 set is large, this attack could potentially exhaust the receiver's
807 memory resources.
808
809 To mitigate the impact of this attack, it is RECOMMENDED that the
810 receiver limit the number of fragments it stores in the reassembling
811 queue so that the sum of the sizes of Encrypted Fragment payload
812 contents (after decryption) for fragments that are already placed
813 into the reassembling queue is less than some value that is
814 reasonable for the implementation. If the peer sends so many
815 fragments that the above condition is not met, the receiver can
816 consider this situation to be either an attack or a broken sender
817 implementation. In either case, the receiver SHOULD drop the
818 connection and discard all the received fragments.
819
820 This value can be predefined, can be a configurable option, or can be
821 calculated dynamically, depending on the receiver's memory load.
822 Some care should be taken when selecting this value because if it is
823 too small it might prevent a legitimate peer from establishing an IKE
824 SA if the size of messages it sends exceeds this value. It is NOT
825 RECOMMENDED for this value to exceed 64 KB because any IKE message
826 before fragmentation would likely be shorter than that.
827
828 If IKE fragments arrive in order, it is possible, but not advised,
829 for the receiver to parse the beginning of the message that is being
830 reassembled and extract the already-available payloads before the
831 reassembly is complete. It can be dangerous to take any action based
832 on the content of these payloads, because the fragments that have not
833
834
835
836 <span class="grey">Smyslov Standards Track [Page 15]</span></pre>
837 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-16" ></span>
838 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
839
840
841 yet been received might contain payloads that could change the
842 meaning of them (or could even make the whole message invalid), and
843 this can potentially be exploited by an attacker. It is important to
844 address this threat by ensuring that all the fragments are received
845 prior to parsing the reassembled message, as described in
846 <a href="#section-2.6">Section 2.6</a>.
847
848 <span class="h2"><a class="selflink" id="section-6" href="#section-6">6</a>. IANA Considerations</span>
849
850 This document defines a new payload in the "IKEv2 Payload Types"
851 registry:
852
853 53 Encrypted and Authenticated Fragment SKF
854
855 This document also defines a new Notify Message Type in the "IKEv2
856 Notify Message Types - Status Types" registry:
857
858 16430 IKEV2_FRAGMENTATION_SUPPORTED
859
860 <span class="h2"><a class="selflink" id="section-7" href="#section-7">7</a>. References</span>
861
862 <span class="h3"><a class="selflink" id="section-7.1" href="#section-7.1">7.1</a>. Normative References</span>
863
864 [<a id="ref-RFC2119">RFC2119</a>] Bradner, S., "Key words for use in RFCs to Indicate
865 Requirement Levels", <a href="https://www.rfc-editor.org/bcp/bcp14">BCP 14</a>, <a href="./rfc2119">RFC 2119</a>, March 1997,
866 &lt;<a href="http://www.rfc-editor.org/info/rfc2119">http://www.rfc-editor.org/info/rfc2119</a>&gt;.
867
868 [<a id="ref-RFC7296">RFC7296</a>] Kaufman, C., Hoffman, P., Nir, Y., Eronen, P., and T.
869 Kivinen, "Internet Key Exchange Protocol Version 2
870 (IKEv2)", STD 79, <a href="./rfc7296">RFC 7296</a>, October 2014,
871 &lt;<a href="http://www.rfc-editor.org/info/rfc7296">http://www.rfc-editor.org/info/rfc7296</a>&gt;.
872
873 [<a id="ref-RFC6311">RFC6311</a>] Singh, R., Kalyani, G., Nir, Y., Sheffer, Y., and D.
874 Zhang, "Protocol Support for High Availability of IKEv2/
875 IPsec", <a href="./rfc6311">RFC 6311</a>, July 2011,
876 &lt;<a href="http://www.rfc-editor.org/info/rfc6311">http://www.rfc-editor.org/info/rfc6311</a>&gt;.
877
878 <span class="h3"><a class="selflink" id="section-7.2" href="#section-7.2">7.2</a>. Informative References</span>
879
880 [<a id="ref-RFC0791">RFC0791</a>] Postel, J., "Internet Protocol", STD 5, <a href="./rfc791">RFC 791</a>,
881 September 1981, &lt;<a href="http://www.rfc-editor.org/info/rfc791">http://www.rfc-editor.org/info/rfc791</a>&gt;.
882
883 [<a id="ref-RFC1191">RFC1191</a>] Mogul, J. and S. Deering, "Path MTU discovery", <a href="./rfc1191">RFC 1191</a>,
884 November 1990, &lt;<a href="http://www.rfc-editor.org/info/rfc1191">http://www.rfc-editor.org/info/rfc1191</a>&gt;.
885
886 [<a id="ref-RFC1981">RFC1981</a>] McCann, J., Deering, S., and J. Mogul, "Path MTU Discovery
887 for IP version 6", <a href="./rfc1981">RFC 1981</a>, August 1996,
888 &lt;<a href="http://www.rfc-editor.org/info/rfc1981">http://www.rfc-editor.org/info/rfc1981</a>&gt;.
889
890
891
892 <span class="grey">Smyslov Standards Track [Page 16]</span></pre>
893 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-17" ></span>
894 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
895
896
897 [<a id="ref-RFC2460">RFC2460</a>] Deering, S. and R. Hinden, "Internet Protocol, Version 6
898 (IPv6) Specification", <a href="./rfc2460">RFC 2460</a>, December 1998,
899 &lt;<a href="http://www.rfc-editor.org/info/rfc2460">http://www.rfc-editor.org/info/rfc2460</a>&gt;.
900
901 [<a id="ref-RFC4787">RFC4787</a>] Audet, F. and C. Jennings, "Network Address Translation
902 (NAT) Behavioral Requirements for Unicast UDP", <a href="https://www.rfc-editor.org/bcp/bcp127">BCP 127</a>,
903 <a href="./rfc4787">RFC 4787</a>, January 2007,
904 &lt;<a href="http://www.rfc-editor.org/info/rfc4787">http://www.rfc-editor.org/info/rfc4787</a>&gt;.
905
906 [<a id="ref-RFC4821">RFC4821</a>] Mathis, M. and J. Heffner, "Packetization Layer Path MTU
907 Discovery", <a href="./rfc4821">RFC 4821</a>, March 2007,
908 &lt;<a href="http://www.rfc-editor.org/info/rfc4821">http://www.rfc-editor.org/info/rfc4821</a>&gt;.
909
910 [<a id="ref-RFC5282">RFC5282</a>] Black, D. and D. McGrew, "Using Authenticated Encryption
911 Algorithms with the Encrypted Payload of the Internet Key
912 Exchange version 2 (IKEv2) Protocol", <a href="./rfc5282">RFC 5282</a>,
913 August 2008, &lt;<a href="http://www.rfc-editor.org/info/rfc5282">http://www.rfc-editor.org/info/rfc5282</a>&gt;.
914
915 [<a id="ref-RFC5405">RFC5405</a>] Eggert, L. and G. Fairhurst, "Unicast UDP Usage Guidelines
916 for Application Designers", <a href="https://www.rfc-editor.org/bcp/bcp145">BCP 145</a>, <a href="./rfc5405">RFC 5405</a>,
917 November 2008, &lt;<a href="http://www.rfc-editor.org/info/rfc5405">http://www.rfc-editor.org/info/rfc5405</a>&gt;.
918
919 [<a id="ref-RFC5723">RFC5723</a>] Sheffer, Y. and H. Tschofenig, "Internet Key Exchange
920 Protocol Version 2 (IKEv2) Session Resumption", <a href="./rfc5723">RFC 5723</a>,
921 January 2010, &lt;<a href="http://www.rfc-editor.org/info/rfc5723">http://www.rfc-editor.org/info/rfc5723</a>&gt;.
922
923 [<a id="ref-RFC6290">RFC6290</a>] Nir, Y., Wierbowski, D., Detienne, F., and P. Sethi, "A
924 Quick Crash Detection Method for the Internet Key Exchange
925 Protocol (IKE)", <a href="./rfc6290">RFC 6290</a>, June 2011,
926 &lt;<a href="http://www.rfc-editor.org/info/rfc6290">http://www.rfc-editor.org/info/rfc6290</a>&gt;.
927
928 [<a id="ref-RFC6888">RFC6888</a>] Perreault, S., Yamagata, I., Miyakawa, S., Nakagawa, A.,
929 and H. Ashida, "Common Requirements for Carrier-Grade NATs
930 (CGNs)", <a href="https://www.rfc-editor.org/bcp/bcp127">BCP 127</a>, <a href="./rfc6888">RFC 6888</a>, April 2013,
931 &lt;<a href="http://www.rfc-editor.org/info/rfc6888">http://www.rfc-editor.org/info/rfc6888</a>&gt;.
932
933 [<a id="ref-FRAGDROP">FRAGDROP</a>] Jaeggli, J., Colitti, L., Kumari, W., Vyncke, E., Kaeo,
934 M., and T. Taylor, "Why Operators Filter Fragments and
935 What It Implies", Work in Progress, <a href="./draft-taylor-v6ops-fragdrop-02">draft-taylor-v6ops-</a>
936 <a href="./draft-taylor-v6ops-fragdrop-02">fragdrop-02</a>, December 2013.
937
938
939
940
941
942
943
944
945
946
947
948 <span class="grey">Smyslov Standards Track [Page 17]</span></pre>
949 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-18" ></span>
950 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
951
952
953 [<a id="ref-BLACKHOLES">BLACKHOLES</a>]
954 De Boer, M. and J. Bosma, "Discovering Path MTU black
955 holes on the Internet using RIPE Atlas", July 2012,
956 &lt;<a href="http://www.nlnetlabs.nl/downloads/publications/pmtu-black-holes-msc-thesis.pdf">http://www.nlnetlabs.nl/downloads/publications/</a>
957 <a href="http://www.nlnetlabs.nl/downloads/publications/pmtu-black-holes-msc-thesis.pdf">pmtu-black-holes-msc-thesis.pdf</a>&gt;.
958
959 [<a id="ref-DOSUDPPROT">DOSUDPPROT</a>]
960 Kaufman, C., Perlman, R., and B. Sommerfeld, "DoS
961 protection for UDP-based protocols", ACM Conference on
962 Computer and Communications Security, October 2003.
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004 <span class="grey">Smyslov Standards Track [Page 18]</span></pre>
1005 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-19" ></span>
1006 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
1007
1008
1009 <span class="h2"><a class="selflink" id="appendix-A" href="#appendix-A">Appendix A</a>. Design Rationale</span>
1010
1011 The simplest approach to IKE fragmentation would have been to
1012 fragment a message that is fully formed and ready to be sent.
1013 However, if a message got fragmented after being encrypted and
1014 authenticated, this could make a simple DoS attack possible. The
1015 attacker could infrequently emit forged but valid-looking fragments
1016 into the network, and some of these fragments would be fetched by the
1017 receiver into the reassembling queue. The receiver would not be able
1018 to distinguish forged fragments from valid ones and would only be
1019 able to determine that some of the received fragments were forged
1020 after the whole message was reassembled and its authenticity check
1021 failed.
1022
1023 To prevent this kind of attack and also reduce vulnerability to some
1024 other kinds of DoS attacks, it was decided to perform fragmentation
1025 before applying cryptographic protection to the message. In this
1026 case, each Fragment message becomes individually encrypted and
1027 authenticated; this allows the receiver to determine forged fragments
1028 and not store them in the reassembling queue.
1029
1030 <span class="h2"><a class="selflink" id="appendix-B" href="#appendix-B">Appendix B</a>. Correlation between IP Datagram Size and Encrypted Payload</span>
1031 Content Size
1032
1033 In the case of IPv4, the content size of the Encrypted Payload is
1034 less than the IP datagram size by the sum of the following values:
1035
1036 o IPv4 header size (typically 20 bytes, up to 60 if IP options are
1037 present)
1038
1039 o UDP header size (8 bytes)
1040
1041 o non-ESP (Encapsulating Security Payload) marker size (4 bytes if
1042 present)
1043
1044 o IKE header size (28 bytes)
1045
1046 o Encrypted payload header size (4 bytes)
1047
1048 o initialization vector (IV) size (variable)
1049
1050 o padding and its size (at least 1 byte)
1051
1052 o ICV size (variable)
1053
1054 The sum may be estimated as 61..105 bytes + IV + ICV + padding.
1055
1056
1057
1058
1059
1060 <span class="grey">Smyslov Standards Track [Page 19]</span></pre>
1061 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-20" ></span>
1062 <span class="grey"><a href="./rfc7383">RFC 7383</a> IKEv2 Fragmentation November 2014</span>
1063
1064
1065 In the case of IPv6, the content size of the Encrypted Payload is
1066 less than the IP datagram size by the sum of the following values:
1067
1068 o IPv6 header size (40 bytes)
1069
1070 o IPv6 extension headers (optional; size varies)
1071
1072 o UDP header size (8 bytes)
1073
1074 o non-ESP marker size (4 bytes if present)
1075
1076 o IKE header size (28 bytes)
1077
1078 o Encrypted payload header size (4 bytes)
1079
1080 o IV size (variable)
1081
1082 o padding and its size (at least 1 byte)
1083
1084 o ICV size (variable)
1085
1086 If no extension header is present, the sum may be estimated as
1087 81..85 bytes + IV + ICV + padding. If extension headers are present,
1088 the payload content size is further reduced by the sum of the size of
1089 the extension headers. The length of each extension header can be
1090 calculated as 8 * (Hdr Ext Len) bytes, except for the fragment
1091 header, which is always 8 bytes in length.
1092
1093 Acknowledgements
1094
1095 The author would like to thank Tero Kivinen, Yoav Nir, Paul Wouters,
1096 Yaron Sheffer, Joe Touch, Derek Atkins, Ole Troan, and others for
1097 their reviews and valuable comments. Thanks to Ron Bonica for
1098 contributing text to the Introduction section. Thanks to Paul
1099 Hoffman and Barry Leiba for improving text clarity.
1100
1101 Author's Address
1102
1103 Valery Smyslov
1104 ELVIS-PLUS
1105 PO Box 81
1106 Moscow (Zelenograd) 124460
1107 Russian Federation
1108
1109 Phone: +7 495 276 0211
1110 EMail: [email protected]
1111
1112
1113
1114
1115
1116 Smyslov Standards Track [Page 20]
1117 </pre>