Debian Codegraph @

rfc4951.html

raw
1 <pre>Network Working Group V. Jain, Ed.
2 Request for Comments: 4951 Riverstone Networks
3 Category: Standards Track August 2007
4
5
6 <span class="h1">Fail Over Extensions for Layer 2 Tunneling Protocol (L2TP) "failover"</span>
7
8 Status of This Memo
9
10 This document specifies an Internet standards track protocol for the
11 Internet community, and requests discussion and suggestions for
12 improvements. Please refer to the current edition of the "Internet
13 Official Protocol Standards" (STD 1) for the standardization state
14 and status of this protocol. Distribution of this memo is unlimited.
15
16 Copyright Notice
17
18 Copyright (C) The IETF Trust (2007).
19
20 Abstract
21
22 Layer 2 Tunneling Protocol (L2TP) is a connection-oriented protocol
23 that has a shared state between active endpoints. Some of this
24 shared state is vital for operation, but may be volatile in nature,
25 such as packet sequence numbers used on the L2TP Control Connection.
26 When failure of one side of a control connection occurs, a new
27 control connection is created and associated with the old connection
28 by exchanging information about the old connection. Such a mechanism
29 is not intended as a replacement for an active fail over with some
30 mirrored connection states, but as an aid for those parameters that
31 are particularly difficult to have immediately available. Protocol
32 extensions to L2TP defined in this document are intended to
33 facilitate state recovery, providing additional resiliency in an L2TP
34 network, and improving a remote system's layer 2 connectivity.
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
55
56
57 Table of Contents
58
59 <a href="#section-1">1</a>. Introduction ....................................................<a href="#page-3">3</a>
60 <a href="#section-1.1">1.1</a>. Terminology ................................................<a href="#page-4">4</a>
61 <a href="#section-1.2">1.2</a>. Abbreviations ..............................................<a href="#page-5">5</a>
62 <a href="#section-1.3">1.3</a>. Specification of Requirements ..............................<a href="#page-5">5</a>
63 <a href="#section-2">2</a>. Overview ........................................................<a href="#page-5">5</a>
64 <a href="#section-3">3</a>. Failover Protocol ...............................................<a href="#page-7">7</a>
65 <a href="#section-3.1">3.1</a>. Failover Capability Negotiation ............................<a href="#page-7">7</a>
66 <a href="#section-3.2">3.2</a>. Failover Recovery Procedure ................................<a href="#page-7">7</a>
67 <a href="#section-3.2.1">3.2.1</a>. Recovery Tunnel Establishment .......................<a href="#page-8">8</a>
68 <a href="#section-3.2.2">3.2.2</a>. Control Channel Reset ..............................<a href="#page-10">10</a>
69 <a href="#section-3.2.3">3.2.3</a>. Data Channel Reset .................................<a href="#page-10">10</a>
70 <a href="#section-3.3">3.3</a>. Session State Synchronization .............................<a href="#page-11">11</a>
71 <a href="#section-4">4</a>. New Control Messages ...........................................<a href="#page-12">12</a>
72 <a href="#section-4.1">4.1</a>. Failover Session Query ....................................<a href="#page-13">13</a>
73 <a href="#section-4.2">4.2</a>. Failover Session Response .................................<a href="#page-13">13</a>
74 <a href="#section-5">5</a>. New Attribute Value Pairs ......................................<a href="#page-14">14</a>
75 <a href="#section-5.1">5.1</a>. Failover Capability AVP ...................................<a href="#page-14">14</a>
76 <a href="#section-5.2">5.2</a>. Tunnel Recovery AVP .......................................<a href="#page-15">15</a>
77 <a href="#section-5.3">5.3</a>. Suggested Control Sequence AVP ............................<a href="#page-16">16</a>
78 <a href="#section-5.4">5.4</a>. Failover Session State AVP ................................<a href="#page-17">17</a>
79 <a href="#section-6">6</a>. Configuration Parameters .......................................<a href="#page-18">18</a>
80 <a href="#section-7">7</a>. IANA Considerations ............................................<a href="#page-19">19</a>
81 <a href="#section-8">8</a>. Security Considerations ........................................<a href="#page-19">19</a>
82 <a href="#section-9">9</a>. Acknowledgements ...............................................<a href="#page-19">19</a>
83 <a href="#section-10">10</a>. Contributors ..................................................<a href="#page-20">20</a>
84 <a href="#section-11">11</a>. References ....................................................<a href="#page-20">20</a>
85 <a href="#section-11.1">11.1</a>. Normative References .....................................<a href="#page-20">20</a>
86 <a href="#section-11.2">11.2</a>. Informative References ...................................<a href="#page-20">20</a>
87 <a href="#appendix-A">Appendix A</a> ........................................................<a href="#page-21">21</a>
88 <a href="#appendix-B">Appendix B</a> ........................................................<a href="#page-23">23</a>
89 <a href="#appendix-C">Appendix C</a> ........................................................<a href="#page-24">24</a>
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
111
112
113 <span class="h2"><a class="selflink" id="section-1" href="#section-1">1</a>. Introduction</span>
114
115 The goal of this document is to aid the overall resiliency of an L2TP
116 endpoint by introducing extensions to <a href="./rfc2661">RFC 2661</a> [<a href="#ref-L2TPv2" title="&quot;Layer Two Tunneling Protocol &quot;">L2TPv2</a>] and <a href="./rfc3931">RFC 3931</a>
117 [<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>] that will minimize the recovery time of the L2TP layer after
118 a failover, while minimizing the impact on its performance.
119 Therefore, it is assumed that the endpoint's overall architecture is
120 also supportive in the resiliency effort.
121
122 To ensure proper operation of an L2TP endpoint after a failover, the
123 associated information of the control connection and sessions between
124 them must be correct and consistent. This includes both the
125 configured and dynamic information. The configured information is
126 assumed to be correct and consistent after a failover, otherwise the
127 tunnels and sessions would not have been setup in the first place.
128
129 The dynamic information, which is also referred to as stateful
130 information, changes with the processing of the tunnel's control and
131 data packets. Currently, the only such information that is essential
132 to the tunnel's operation is its sequence numbers. For the tunnel
133 control channel, the inconsistencies in its sequence numbers can
134 result in the termination of the entire tunnel. For tunnel sessions,
135 the inconsistency in its sequence numbers, when used, can cause
136 significant data loss, which gives the perception of a "service loss"
137 to the end user.
138
139 Thus, an optimal resilient architecture that aims to minimize
140 "service loss" after a failover, must make provisions for the
141 tunnel's essential stateful information, i.e., its sequence numbers.
142 Currently, there are two options available: the first option is to
143 ensure that the backup endpoint is completely synchronized with the
144 active endpoint, with respect to the control and data sessions
145 sequence numbers. The other option is to reestablish all the tunnels
146 and their sessions after a failover. The drawback of the first
147 option is that it adds significant performance and complexity impact
148 to the endpoint's architecture, especially as tunnel and session
149 aggregation increases. The drawback of the second option is that it
150 increases the "service loss" time, especially as the architecture
151 scales.
152
153 To alleviate the above-mentioned drawbacks of the current options,
154 this document introduces a mechanism to bring the dynamic stateful
155 information of a tunnel to a correct and consistent state after a
156 failure. The proposed mechanism defines the recovery of tunnels and
157 sessions that were in an established state prior to the failure.
158
159
160
161
162
163
164 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
167
168
169 <span class="h3"><a class="selflink" id="section-1.1" href="#section-1.1">1.1</a>. Terminology</span>
170
171 Endpoint: L2TP control connection endpoint, i.e., either LAC or LNS
172 (also known as LCCE in [<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>]).
173
174 Active Endpoint: An endpoint that is currently providing service.
175
176 Backup Endpoint: A redundant endpoint standing by for the active
177 endpoint that has its database of active tunnels and sessions in sync
178 with its active endpoint.
179
180 Failed Endpoint: The endpoint that was the active endpoint at the
181 time of the failure.
182
183 Recovery Endpoint: The endpoint that initiates the failover protocol
184 to recover from the failure of an active endpoint.
185
186 Remote Endpoint: The endpoint that peers with active endpoint before
187 failure and with recovery endpoint after failure.
188
189 Failover: The action of a backup endpoint taking over the service of
190 an active endpoint. This could be due to administrative action or
191 failure of the active endpoint.
192
193 Old Tunnel: A control connection that existed before failure and is
194 subjected to recovery upon failover.
195
196 Recovery Tunnel: A new control connection established only to recover
197 an old tunnel.
198
199 Recovered Tunnel: After an old tunnel's control connection and
200 sessions are restored using the mechanism described in this document,
201 it is referred to as a Recovered Tunnel.
202
203 Control Channel Failure: Failure of the component responsible for
204 establishing/maintaining tunnels and sessions at an endpoint.
205
206 Data Channel Failure: Failure of the component responsible for
207 forwarding the L2TP encapsulated data.
208
209
210
211
212
213
214
215
216
217
218
219
220 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
223
224
225 <span class="h3"><a class="selflink" id="section-1.2" href="#section-1.2">1.2</a>. Abbreviations</span>
226
227 LAC L2TP Access Concentrator
228 LNS L2TP Network Server
229 LCCE L2TP Control Connection Endpoint
230 AVP Attribute Value Pair
231 SCCRQ Start-Control-Connection-Request
232 SCCRP Start-Control-Connection-Reply
233 ZLB Zero Length Body Message
234
235 <span class="h3"><a class="selflink" id="section-1.3" href="#section-1.3">1.3</a>. Specification of Requirements</span>
236
237 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
238 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
239 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>].
240
241 <span class="h2"><a class="selflink" id="section-2" href="#section-2">2</a>. Overview</span>
242
243 The following diagram depicts the redundancy architecture and
244 pertaining entities used to describe the failover protocol:
245
246 +--------------+
247 | L2TP active |
248 +----------+ ----| endpoint (A) |
249 | L2TP | / +--------------+
250 | endpoint |----------------------/
251 | (R) | \ +--------------+
252 +----------+ \ | L2TP backup |
253 ----| endpoint (B) |
254 +--------------+
255
256 Active and backup endpoints may reside on the same device, however,
257 they are not required to be that way. On other hand, some devices
258 may not have a standby module altogether, in which case the failed
259 endpoint, after reset, can become the recovery endpoint to recover
260 from its prior failure.
261
262 Therefore, in the above diagram, upon A's (active endpoint's)
263 failure:
264
265 - Endpoint A would be called the failed endpoint.
266
267 - If B is present, then it would become the recovery endpoint and
268 also an active endpoint.
269
270 - If B is not present, then A could become the recovery endpoint
271 after it restarts, provided it saved the information about
272 active tunnels/sessions in some persistent storage.
273
274
275
276 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
279
280
281 - R does not initiate the failover protocol; rather it waits for a
282 failure indication from recovery endpoint.
283
284 This document assumes that the actual detection of a failure in the
285 backup endpoint is done in an implementation-specific way. It also
286 assumes that failure detection by the backup endpoint is faster than
287 the L2TP control channel timeout between the active and remote
288 endpoints. Typically, active and backup endpoints reside on the same
289 physical device, allowing fast and reliable failure detection without
290 the need to communicate between these endpoints over the network.
291
292 Similarly, an active endpoint that acts also as a backup endpoint can
293 easily start the recovery once it has rebooted. However, when the
294 active and backup endpoints reside in separate devices, there is a
295 need for communication between them in order to detect failures. As
296 a solution for such situations, additional failure detection
297 protocols, e.g., [<a href="#ref-BFD-MULTIHOP" title="&quot;BFD for Multihop Paths&quot;">BFD-MULTIHOP</a>], may be needed.
298
299 A device could have three kinds of failures:
300
301 i) Control Channel Failure
302
303 ii) Data Channel Failure
304
305 iii) Control and Data Channel Failure
306
307 The protocol described in this document specifies the recovery in
308 conditions i) and iii). It is perceived that not much (stateful
309 information) could be recovered via a control protocol exchange in
310 case of ii).
311
312 The failover protocol consists of three phases:
313
314 1) Failover Capability Negotiation: An active endpoint and a remote
315 endpoint exchange failover capabilities and attributes to be used
316 during the recovery process.
317
318 2) Failover Recovery: A recovery endpoint establishes a new L2TP
319 control connection (called recovery tunnel) for every old tunnel
320 that it wishes to recover. The recovery tunnel serves three
321 purposes:
322
323 - It identifies the old tunnel that is being recovered.
324
325 - It provides a means of authentication and a three-way handshake
326 to ensure both ends agree on the failover for the specified old
327 tunnel.
328
329
330
331
332 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
335
336
337 - It could exchange the Ns and Nr values to be used in the
338 recovered tunnel.
339
340 Upon establishing the recovery tunnel, two endpoints reset the
341 control and data channel(s) on the recovered tunnel using the
342 procedures described in <a href="#section-3.2.2">Section 3.2.2</a> and <a href="#section-3.2.3">Section 3.2.3</a>,
343 respectively. The recovery tunnel could be torn down after that,
344 and sessions that were established resume traffic.
345
346 3) Session State Synchronization: The session state synchronization
347 process occurs on the recovered or the old tunnel and allows the
348 two endpoints to agree on the state of the various sessions in the
349 tunnel after failover. The inconsistency, which could arise due
350 to the failure, is handled in the following manner: first, the two
351 endpoints silently clear the sessions that were not in the
352 established state. Then, they utilize Failover Session Query
353 (FSQ) and Failover Session Response (FSR) on the recovered tunnel
354 to obtain the state of sessions as known to the peer endpoint and
355 clear the sessions accordingly.
356
357 <span class="h2"><a class="selflink" id="section-3" href="#section-3">3</a>. Failover Protocol</span>
358
359 The protocol consists of three steps describing specifications during
360 the life of a control connection - before and after failover.
361
362 <span class="h3"><a class="selflink" id="section-3.1" href="#section-3.1">3.1</a>. Failover Capability Negotiation</span>
363
364 The active and remote endpoints exchange the Failover Capability
365 attribute-value pair (AVP) in Start-Control-Connection-Request
366 (SCCRQ) and Start-Control-Connection-Reply (SCCRP) during control
367 connection establishment as a part of the normal (before failover)
368 operation. The Failover Capability AVP, defined in <a href="#section-5.1">Section 5.1</a>,
369 allows an endpoint to specify if it is control and/or data channel
370 failover capable and the time allowed for the recovery for the
371 tunnel.
372
373 <span class="h3"><a class="selflink" id="section-3.2" href="#section-3.2">3.2</a>. Failover Recovery Procedure</span>
374
375 The Failover Recovery Procedure described in this section is
376 performed only if there was a control channel failure. The selection
377 of the tunnels to be recovered is implementation specific.
378
379 The Failover Recovery Procedure consists of following three steps,
380 which are described in detail in the subsections below:
381
382 - Recovery tunnel establishment
383
384 - Control channel reset
385
386
387
388 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
391
392
393 - Data channel reset
394
395 <span class="h4"><a class="selflink" id="section-3.2.1" href="#section-3.2.1">3.2.1</a>. Recovery Tunnel Establishment</span>
396
397 The recovery endpoint establishes a new control connection, called
398 recovery tunnel, for every old tunnel it wishes to recover. The
399 purpose of the recovery tunnel is solely to recover the corresponding
400 old tunnel. There is a one to one relationship between recovery
401 tunnel and recovered/old tunnel
402
403 Recovery tunnel establishment considerations:
404
405 - An LCCE MUST follow the procedures described in [<a href="#ref-L2TPv2" title="&quot;Layer Two Tunneling Protocol &quot;">L2TPv2</a>] or
406 [<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>] to establish the recovery tunnel.
407
408 - The recovery tunnel MUST use the same L2TP version (and
409 establishment procedures) that was used for the old tunnel.
410
411 - The SCCRQ for Recovery tunnel MUST include the Tunnel Recovery
412 AVP, defined in <a href="#section-5.2">Section 5.2</a>, to identify the old tunnel that is
413 being recovered.
414
415 - The recovery tunnel MUST NOT include the Failover Capability AVP
416 in its SCCRQ or SCCRP messages.
417
418 - An endpoint SHOULD NOT send any message other than the following
419 on the recovery tunnel: SCCRQ, SCCRP, SCCCN, StopCCN, HELLO,
420 ZLB, and ACK ([<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>] only).
421
422 - An endpoint MUST NOT use any old Tunnel ID for the recovery
423 tunnel. The old tunnels MUST be valid until the recovery
424 process concludes.
425
426 - An endpoint MUST use the Tie Breaker AVP (<a href="#section-4.4.3">Section 4.4.3</a>
427 [<a href="#ref-L2TPv2" title="&quot;Layer Two Tunneling Protocol &quot;">L2TPv2</a>]) or Control Connection Tie Breaker AVP (<a href="#section-5.4.3">Section 5.4.3</a>
428 [<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>]) in the setup of the recovery tunnel to ensure that
429 only a single recovery tunnel (when both endpoints have
430 simultaneous failover) is established to recover an old tunnel.
431 The tunnel that wins the tie is used to decide the suggested Ns
432 and Nr values on the recovered tunnel. Therefore, the endpoint
433 that loses the tie, should reset the Ns and Nr values (<a href="#section-3.2.2">Section</a>
434 <a href="#section-3.2.2">3.2.2</a>) as if it were a remote endpoint. <a href="#appendix-B">Appendix B</a> illustrates
435 the double failover scenario.
436
437 - Tie Breaker AVP processing: The scope of a tie breaker AVP's
438 action for recovery and non recovery tunnels must be
439 independent, and is defined as follows:
440
441
442
443
444 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
447
448
449 o When Tie Breaker AVP is used in a non recovery tunnel, the
450 scope of the tie breaker AVP's action MUST only be within non
451 recovery tunnels. Therefore, losing a tie against a non
452 recovery tunnel MUST NOT result in termination of any
453 recovery tunnel.
454
455 o When a Tie Breaker AVP is used in a recovery tunnel, the
456 scope of tie breaker AVP's action is further restricted to
457 the recovery tunnel(s) for a single tunnel to be recovered.
458 Thus, an implementation MUST apply the tie breaker received
459 in a recovery tunnel only to those tunnels that are a)
460 recovery tunnels, and b) associated with the same tunnel to
461 be recovered. It MUST NOT impact the operation of non-
462 recovery tunnels and recovery tunnels associated with other
463 old tunnels to be recovered.
464
465 Upon getting an SCCRQ with a Tunnel Recovery AVP, an endpoint
466 validates the Recover Tunnel ID and the Recover Remote Tunnel ID and
467 responds with an SCCRP. It MUST terminate the recovery tunnel if:
468
469 - The Recover Tunnel ID or the Recover Remote Tunnel ID is
470 unknown.
471
472 - The active or remote endpoint (prior to failover) had not
473 indicated that it was failover capable.
474
475 - The L2TP version of recovery tunnel is different from the
476 version used in the old tunnel.
477
478 If the remote endpoint accepts the SCCRQ, it SHOULD include the
479 Suggested Control Sequence AVP, defined in <a href="#section-5.3">Section 5.3</a>, in the SCCRP
480 message.
481
482 Authentication considerations:
483
484 - To authenticate a peer endpoint during recovery tunnel
485 establishment, an endpoint MUST follow the procedure described
486 in either [<a href="#ref-L2TPv2" title="&quot;Layer Two Tunneling Protocol &quot;">L2TPv2</a>] <a href="#section-5.1.1">Section 5.1.1</a> or [<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>] <a href="#section-4.3">Section 4.3</a>. It
487 MUST use the same secret that was used to authenticate the old
488 tunnel.
489
490 - Not being able to authenticate could be a reason to terminate
491 the recovery tunnel.
492
493 - For L2TPv3 tunnels, a recovery tunnel MUST use the Control
494 Message authentication (i.e., exchange the nonce values), as
495 described in [<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>] <a href="#section-4.3">Section 4.3</a>, if the old tunnel was
496 configured to do control message authentication. An L2TPv3
497
498
499
500 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
503
504
505 recovered tunnel MUST reset its nonce values (both endpoints) to
506 the nonce values exchanged in the recovery tunnel.
507
508 For any reason, if the recovery endpoint could not establish the
509 recovery tunnel, then it MUST silently clear the old tunnel and
510 sessions within, concluding that the recovery process has failed.
511
512 Any control packet received on the recovered tunnel before control
513 channel reset (<a href="#section-3.2.2">Section 3.2.2</a>) MUST be silently discarded.
514
515 <span class="h4"><a class="selflink" id="section-3.2.2" href="#section-3.2.2">3.2.2</a>. Control Channel Reset</span>
516
517 Control channel reset allows new control messages to be sent and
518 received over the recovered tunnel.
519
520 Control channel reset procedure:
521
522 - An endpoint SHOULD flush the transmit/receive windows and reset
523 the control channel sequence numbers (i.e., Ns and Nr values) on
524 the recovered tunnel. The control channel on the recovery
525 endpoint is reset upon getting a valid SCCRP on the recovery
526 tunnel, whereas the control channel on the remote endpoint is
527 reset upon getting a valid SCCCN on the recovery tunnel. If the
528 recovery endpoint did not receive the Suggested Control Sequence
529 (SCS) AVP in the SCCRP then it MUST reset the Ns and Nr values
530 to zero. If the remote endpoint opted to not send the SCS AVP
531 then it MUST reset the Ns and Nr values to zero. Either
532 endpoint can tear down the recovery tunnel after the control
533 channel reset procedure is complete.
534
535 - An endpoint MUST prevent the establishment of new sessions until
536 it has cleared (or marked for clearance) the sessions that were
537 not in an established state, i.e., until after Step I, <a href="#section-3.3">Section</a>
538 <a href="#section-3.3">3.3</a> is complete.
539
540 <span class="h4"><a class="selflink" id="section-3.2.3" href="#section-3.2.3">3.2.3</a>. Data Channel Reset</span>
541
542 A data channel reset procedure is applicable only for the sessions
543 using sequence numbers. For L2TPv3 data channel, the terms Nr and Ns
544 in this document are used to mean "expected sequence number" and
545 "sequence number", respectively.
546
547 Data channel reset procedure:
548
549 - The recovery endpoint sets the Ns value to zero.
550
551 - The remote endpoint (recovery endpoint's peer) continues to use
552 the Ns values it was using previously.
553
554
555
556 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
559
560
561 - To reset Nr values during failover, if an endpoint receives 'n'
562 out of order but in sequence packets, then it MUST set the Nr
563 value based on the Ns value of the incoming packets, as
564 suggested in <a href="#appendix-C">Appendix C</a> of [<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>]. The value of 'n' SHOULD be
565 configurable.
566
567 - If one of the endpoints does not exhibit the capability
568 (indicated in 'D' bit in the Failover Capability AVP) to reset
569 the Nr value, then data channels using sequence numbers are
570 considered non recoverable. Those sessions SHOULD be torn down
571 by the recovery endpoint by sending a Call-Disconnect-Notify
572 (CDN).
573
574 - For data-channel-only failure, two endpoints MAY use the session
575 state query/response mechanism on the control channel to
576 synchronize the state of sessions as described in <a href="#section-3.3">Section 3.3</a>
577 below.
578
579 <span class="h3"><a class="selflink" id="section-3.3" href="#section-3.3">3.3</a>. Session State Synchronization</span>
580
581 If a control channel failure happens when a session was being
582 established or torn down, then it is possible for an endpoint to
583 consider a session in an established state while its peer considers
584 the same session non existent. Two such situations occur when
585 failure on an endpoint occurs immediately after sending:
586
587 - A CDN message that never made it to the peer.
588
589 - An ICCN message that never made it to the peer.
590
591 The following mechanism MUST be used to identify and clear the
592 sessions that exists on an endpoint, but not on its peer:
593
594 Step I: For control channel failure, after the recovery tunnel is
595 established, the sessions that were not in an established state MUST
596 be silently cleared (i.e., without sending a CDN message) by each
597 endpoint.
598
599 Step II: Both endpoints MAY identify the sessions that might have
600 been in inconsistent states, perhaps based on data channel
601 inactivity. FSQ and FSR messages have been introduced to synchronize
602 session state at any given point during the life of a session between
603 two endpoints. These messages are used when one endpoint determines
604 or suspects in an implementation specific manner that its session
605 state could be inconsistent with that of its peer's.
606
607 Step III: An endpoint sends a Failover Session Query (FSQ) message to
608 query the state of sessions as known to its peer. An FSQ message
609
610
611
612 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
615
616
617 contains one Failover Session State (FSS) AVP, defined in <a href="#section-5.4">Section</a>
618 <a href="#section-5.4">5.4</a>, for each session it wishes to query. Multiple FSS AVPs could be
619 included in one FSQ message. An FSQ message MUST include at least
620 one FSS AVP. An endpoint MAY send another FSQ message before getting
621 a response for its previous FSQs.
622
623 An inconsistency about a session's existence during failover could
624 result in an endpoint selecting the same Session ID for a new
625 session. In such a situation, it would send an ICRQ for an already
626 established session. Therefore, before all sessions are synchronized
627 using an FSQ/FSR mechanism, if endpoint receives an ICRQ for a
628 session in an established state, then it MUST respond to such an ICRQ
629 with a CDN. The CDN message must set Assigned/Local Session ID AVP
630 ([<a href="#ref-L2TPv2" title="&quot;Layer Two Tunneling Protocol &quot;">L2TPv2</a>] <a href="#section-4.4.4">Section 4.4.4</a>, [<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>] <a href="#section-5.4.4">Section 5.4.4</a>) to its local Session
631 ID and clear the session that it considered established. Use of a
632 least recently used Session ID for the new sessions could help reduce
633 this symptom during failover.
634
635 When an endpoint receives an FSQ message, it MUST ensure that for
636 each FSS AVP in the FSQ message, it includes an FSS AVP in the
637 Failover Session Response (FSR) message. An endpoint could respond
638 to multiple FSQs using one FSR message, or it could respond one FSQ
639 with multiple FSRs. FSSs are not required to be responded in the
640 same order in which they were received. For each FSS AVP received in
641 FSQ messages, an endpoint MUST validate the Remote Session ID and
642 determine if it is paired with the Session ID specified in the
643 message. If an FSS AVP is not valid (i.e., session is non-existing
644 or it is paired with different remote Session ID), then the Session
645 ID field in the FSS AVP in the FSR MUST be set to zero. When session
646 is discovered to be pairing with mismatching Session ID, the local
647 session MUST not be cleared, but rather marked stale, to be queried
648 later using an FSQ message. <a href="#appendix-C">Appendix C</a> presents an example dialogue
649 between two endpoints with mismatching Session IDs.
650
651 When responding to an FSQ with an FSR message, the Remote Session ID
652 in the FSS AVP of the FSR message is always set to the received value
653 of the Session ID in the FSS AVP of the FSQ message.
654
655 When an endpoint receives an FSR message, for each FSS AVP it MUST
656 use the Remote Session ID field to identify the local session and
657 silently (without sending a CDN) clear the session if the Session ID
658 in the AVP was zero. Otherwise, it MUST consider the session to be
659 in an established state and recovered.
660
661 <span class="h2"><a class="selflink" id="section-4" href="#section-4">4</a>. New Control Messages</span>
662
663 This document introduces two new messages that could be sent over an
664 established/recovered control connection.
665
666
667
668 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
671
672
673 <span class="h3"><a class="selflink" id="section-4.1" href="#section-4.1">4.1</a>. Failover Session Query</span>
674
675 The Failover Session Query (FSQ) control message is used by an
676 endpoint during the recovery process to query the state of various
677 sessions. It triggers a response from the peer, which contains the
678 requested state of various sessions.
679
680 This control message is encoded as follows:
681
682 Vendor ID = 0 (IETF)
683 Attribute Type = 21
684
685 The following AVPs MUST be present in the FSQ control message:
686
687 Message Type
688 Failover Session State
689
690 The following AVPs MAY be present in the FSQ control message:
691
692 Random Vector
693 Message digest ([<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>] tunnels only)
694
695 Other AVPs MUST NOT be sent in this control message and SHOULD be
696 ignored on receipt.
697
698 The M-bit on the Message Type AVP for this control message MUST be
699 set to 0.
700
701 <span class="h3"><a class="selflink" id="section-4.2" href="#section-4.2">4.2</a>. Failover Session Response</span>
702
703 The Failover Session Response (FSR) control message is used by an
704 endpoint during the recovery process to respond with the local state
705 of various sessions. It is sent as a response to an FSQ message. An
706 endpoint MAY choose to respond to an FSQ message with multiple FSR
707 messages.
708
709 This control message is encoded as follows:
710
711 Vendor ID = 0 (IETF)
712 Attribute Type = 22
713
714 The following AVPs MUST be present in the FSR control message:
715
716 Message Type
717 Failover Session State
718
719
720
721
722
723
724 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
727
728
729 The following AVPs MAY be present in the FSR control message:
730
731 Random Vector
732 Message digest ([<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>] tunnels only)
733
734 Other AVPs MUST NOT be sent in this control message and SHOULD be
735 ignored on receipt.
736
737 The M-bit on the Message Type AVP for this control message MUST be
738 set to 0.
739
740 <span class="h2"><a class="selflink" id="section-5" href="#section-5">5</a>. New Attribute Value Pairs</span>
741
742 The following sections contain a list of new L2TP AVPs defined in
743 this document.
744
745 <span class="h3"><a class="selflink" id="section-5.1" href="#section-5.1">5.1</a>. Failover Capability AVP</span>
746
747 The Failover Capability AVP, Attribute Type 76, indicates the
748 capabilities of an endpoint required for the recovery process. The
749 AVP format is defined as follows:
750
751 Failover Capability AVP
752
753 0 1 2 3
754 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
755 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
756 |M|H| rsvd | Length | 0 |
757 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
758 | Attribute Type 76 | Reserved |D|C|
759 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
760 | Recovery Time (in milliseconds) |
761 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
762
763 The AVP MAY be hidden (the H-bit set to 0 or 1). The AVP is not
764 mandatory (the M-bit MUST be set to 0).
765
766 The C bit governs the failover capability for the control channel.
767 When the C bit is set, it indicates that the endpoint can recover
768 from a control channel failure using the procedure described in
769 <a href="#section-3.2.2">Section 3.2.2</a>.
770
771 When the C bit is not set, it indicates that the endpoint cannot
772 recover from a control channel failover. In this case, the D bit
773 MUST be set. Note that a control channel failover in this case would
774 be fatal for the tunnel and all associated data channels.
775
776
777
778
779
780 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
783
784
785 The D bit governs the failover capability for data channels that use
786 sequence numbers. Data channels that do not use sequence numbers do
787 not need help to recover from a data channel failure.
788
789 When the D bit is set, it indicates that the endpoint is capable of
790 resetting Nr value of data channels using the procedure described in
791 <a href="#section-3.2.3">Section 3.2.3</a> Data Channel reset procedure.
792
793 When the D bit is not set, it indicates that the endpoint cannot
794 recover data channels that use sequence numbers. In the case of a
795 failure, such data channels would be lost.
796
797 The Failover Capability AVP MUST NOT be sent with C bit and D bit
798 cleared.
799
800 The Recovery Time, applicable only when the C bit is set, is the time
801 in milliseconds an endpoint asks its peer to wait before assuming the
802 recovery process has failed. This timer starts when an endpoint's
803 control channel timeout ([<a href="#ref-L2TPv2" title="&quot;Layer Two Tunneling Protocol &quot;">L2TPv2</a>] <a href="#section-5.8">Section 5.8</a>, [<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>] <a href="#section-4.2">Section 4.2</a>)
804 is started, and is not stopped (before expiry) until an endpoint
805 successfully authenticates its peer during recovery. A value of zero
806 does not mean that failover will not occur, it means no additional
807 time is requested from the peer. The timer is also stopped if a
808 control channel message is acknowledged by the peer in the situation
809 when there was no failover, but the loss of the control channel
810 message was a temporary phenomenon.
811
812 This AVP MUST NOT be included in any control message other than SCCRQ
813 and SCCRP messages.
814
815 <span class="h3"><a class="selflink" id="section-5.2" href="#section-5.2">5.2</a>. Tunnel Recovery AVP</span>
816
817 The Tunnel Recovery AVP, Attribute Type 77, indicates that a sender
818 would like to recover the tunnel identified in this AVP due to a
819 failure. The AVP format is defined as follows:
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
839
840
841 Tunnel Recovery AVP for L2TPv3 tunnels:
842
843 0 1 2 3
844 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
845 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
846 |M|H| rsvd | Length | 0 |
847 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
848 | Attribute Type 77 | Reserved |
849 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
850 | Recover Tunnel ID |
851 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
852 | Recover Remote Tunnel ID |
853 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
854
855
856 Tunnel Recovery AVP for L2TPv2 tunnels:
857
858 0 1 2 3
859 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
860 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
861 |M|H| rsvd | Length | 0 |
862 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
863 | Attribute Type 77 | Reserved |
864 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
865 | Reserved | Recover Tunnel ID |
866 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
867 | Reserved | Recover Remote Tunnel ID |
868 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
869
870 This AVP MUST not be hidden (the H-bit is set to 0). The AVP is
871 mandatory (the M-bit is set to 1).
872
873 The Recover Tunnel ID encodes the local Tunnel ID that an endpoint
874 wants recovered. The Recover Remote Tunnel ID encodes the remote
875 Tunnel ID corresponding to the old tunnel.
876
877 This AVP MUST NOT be included in any control message other than the
878 SCCRQ message when establishing a Recovery Tunnel.
879
880 <span class="h3"><a class="selflink" id="section-5.3" href="#section-5.3">5.3</a>. Suggested Control Sequence AVP</span>
881
882 The Suggested Control Sequence (SCS) AVP, Attribute Type 78,
883 specifies the Ns and Nr values to for the recovered tunnel. This AVP
884 is included in an SCCRP message of a recovery tunnel by remote
885 endpoint. The AVP format is defined as follows:
886
887
888
889
890
891
892 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
895
896
897 0 1 2 3
898 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
899 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
900 |M|H| rsvd | Length | 0 |
901 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
902 | Attribute Type 78 | Reserved |
903 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
904 | Suggested Ns | Suggested Nr |
905 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
906
907 This AVP MAY be hidden (the H-bit set to 0 or 1). The AVP is not
908 mandatory (the M-bit is set to 0).
909
910 This is an optional AVP, suggesting Ns and Nr values to be used by
911 the recovery endpoint. If this AVP is present in an SCCRP message
912 during recovery tunnel establishment, the recovery endpoint MUST set
913 the Ns and Nr values of the recovered tunnel to the respective
914 suggested values. When this AVP is not sent in an SCCRP or not
915 present in an incoming SCCRP, the Ns and Nr values for the recovered
916 tunnel are set to zero. Use of this AVP helps avoid the interference
917 in the recovered tunnel's control channel with old control packets.
918
919 This AVP MUST NOT be included in any control message other than the
920 SCCRP message when establishing a Recovery Tunnel.
921
922 <span class="h3"><a class="selflink" id="section-5.4" href="#section-5.4">5.4</a>. Failover Session State AVP</span>
923
924 The Failover Session State (FSS) AVP, Attribute Type 79, is used to
925 query the state of a session from the peer end to clear the sessions
926 that otherwise would remain in an undefined state after failover.
927 The AVP format is defined as follows:
928
929 FSS AVP format for L2TPv3 sessions:
930
931 0 1 2 3
932 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
933 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
934 |M|H| rsvd | Length | 0 |
935 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
936 | Attribute Type 79 | Reserved |
937 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
938 | Session ID |
939 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
940 | Remote Session ID |
941 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
942
943
944
945
946
947
948 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
951
952
953 FSS AVP format for L2TPv2 sessions:
954
955 0 1 2 3
956 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
957 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
958 |M|H| rsvd | Length | 0 |
959 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
960 | Attribute Type 79 | Reserved |
961 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
962 | Reserved | Session ID |
963 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
964 | Reserved | Remote Session ID |
965 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
966
967 This AVP MAY be hidden (the H-bit set to 0 or 1). The AVP is
968 mandatory (the M-bit is set to 1).
969
970 The Session ID identifies the local Session ID that the sender had
971 assigned, for which it would like to query the state on its peer. A
972 Remote Session Id is the remote Session ID for the same session.
973
974 An FSS AVP MUST NOT be used in any message other than FSQ and FSR
975 messages.
976
977 <span class="h2"><a class="selflink" id="section-6" href="#section-6">6</a>. Configuration Parameters</span>
978
979 An L2TP endpoint MAY expose the following configuration parameters to
980 be specified for control connections:
981
982 - Control Channel Failover Capability: Failover Capability AVP
983 (<a href="#section-5.1">Section 5.1</a>), C bit.
984
985 - Data Channel Failover Capability: Failover Capability AVP
986 (<a href="#section-5.1">Section 5.1</a>), D bit.
987
988 - Recovery Time: Failover Capability AVP (<a href="#section-5.1">Section 5.1</a>).
989
990 The L2TP MIB defined in [<a href="#ref-L2TPv2-MIB" title="&quot;Layer Two Tunneling Protocol &quot;">L2TPv2-MIB</a>] and [<a href="#ref-L2TPv3-MIB" title="&quot;Layer Two Tunneling Protocol (version 3) Management Information Base&quot;">L2TPv3-MIB</a>], defines a
991 number of objects that may be used for monitoring the status L2TP
992 nodes, but is seldom used for configuration purposes. It is expected
993 that the above mentioned parameters will be configured by using a
994 Command Line Interface (CLI) or other proprietary mechanism.
995
996 Asynchronous notifications for failover and recovery events may be
997 sent by L2TP nodes to network management applications, but the
998 specification of the protocol and format to be used for these
999 notifications is out of the scope of this document.
1000
1001
1002
1003
1004 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
1007
1008
1009 <span class="h2"><a class="selflink" id="section-7" href="#section-7">7</a>. IANA Considerations</span>
1010
1011 This document defines the following values assigned by IANA.
1012
1013 - Four Control Message Attribute Value Pairs (<a href="#section-10.1">Section 10.1</a> [<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>]):
1014
1015 Failover Capability : 76
1016 Tunnel Recovery : 77
1017 Suggested Control Sequence : 78
1018 Failover Session State : 79
1019
1020
1021 - Two Message Type (Attribute Type 0) Values (<a href="#section-10.2">Section 10.2</a> [<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>]):
1022
1023 Failover Session Query : 21
1024 Failover Session Response : 22
1025
1026 <span class="h2"><a class="selflink" id="section-8" href="#section-8">8</a>. Security Considerations</span>
1027
1028 A spoofed failover request (SCCRQ with Tunnel Recovery AVP) on behalf
1029 of an endpoint might cause a control channel termination if
1030 authentication measures mentioned in <a href="#section-3.2.1">Section 3.2.1</a> are not used.
1031
1032 Even if the authentication measures (as described in <a href="#section-3.2.1">Section 3.2.1</a>)
1033 were used, it is still possible to learn an identity of an
1034 operational tunnel from an endpoint by issuing it spoofed failover
1035 requests that fail the authentication procedure. The probability of
1036 succeeding with a spoofed failover request is 1 in (2^16 - 1) for
1037 [<a href="#ref-L2TPv2" title="&quot;Layer Two Tunneling Protocol &quot;">L2TPv2</a>] and 1 in (2^32 - 1) for [<a href="#ref-L2TPv3" title="&quot;Layer Two Tunneling Protocol - Version 3 (L2TPv3)&quot;">L2TPv3</a>]. The discovered identity
1038 of an operational tunnel could then be misused to send control
1039 messages for a possible hindrance to the control connection.
1040 Typically, control messages that are outside the endpoint's receive
1041 window are discarded. However, if Suggested Control Sequence AVP
1042 (<a href="#section-5.3">Section 5.3</a>) is not used during the actual failover process, the
1043 sequence numbers might be reset to zero, thereby making the receive
1044 window predictable. To improve security under such circumstances, an
1045 endpoint may be configured with the possible set of recovery
1046 endpoints that could recover a tunnel, and use of Suggested Control
1047 Sequence AVP when recovering a tunnel.
1048
1049 <span class="h2"><a class="selflink" id="section-9" href="#section-9">9</a>. Acknowledgements</span>
1050
1051 Leo Huber provided suggestions to help define the failover concept.
1052 Mark Townsley, Carlos Pignataro, and Ignacio Goyret reviewed the
1053 document and provided valuable suggestions.
1054
1055
1056
1057
1058
1059
1060 <span class="grey">Jain, et al. 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="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
1063
1064
1065 <span class="h2"><a class="selflink" id="section-10" href="#section-10">10</a>. Contributors</span>
1066
1067 Paul Howard Juniper Networks
1068 Vipin Jain Riverstone Networks
1069 Sam Henderson Cisco Systems
1070 Keyur Parikh Harris Corporations
1071
1072 <span class="h2"><a class="selflink" id="section-11" href="#section-11">11</a>. References</span>
1073
1074 <span class="h3"><a class="selflink" id="section-11.1" href="#section-11.1">11.1</a>. Normative References</span>
1075
1076 [<a id="ref-RFC2119">RFC2119</a>] Bradner, S., "Key words for use in RFCs to Indicate
1077 Requirement Levels", <a href="https://www.rfc-editor.org/bcp/bcp14">BCP 14</a>, <a href="./rfc2119">RFC 2119</a>, March 1997.
1078
1079 [<a id="ref-L2TPv2">L2TPv2</a>] Townsley, W., Valencia, A., Rubens, A., Pall, G.,
1080 Zorn, G., and B. Palter, "Layer Two Tunneling Protocol
1081 "L2TP"", <a href="./rfc2661">RFC 2661</a>, August 1999.
1082
1083 [<a id="ref-L2TPv3">L2TPv3</a>] Lau, J., Townsley, M., and I. Goyret, "Layer Two
1084 Tunneling Protocol - Version 3 (L2TPv3)", <a href="./rfc3931">RFC 3931</a>,
1085 March 2005.
1086
1087 <span class="h3"><a class="selflink" id="section-11.2" href="#section-11.2">11.2</a>. Informative References</span>
1088
1089 [<a id="ref-L2TPv2-MIB">L2TPv2-MIB</a>] Caves, E., Calhoun, P., and R. Wheeler, "Layer Two
1090 Tunneling Protocol "L2TP" Management Information
1091 Base", <a href="./rfc3371">RFC 3371</a>, August 2002.
1092
1093
1094 [<a id="ref-L2TPv3-MIB">L2TPv3-MIB</a>] Nadeau, T. and K. Koushik, "Layer Two Tunneling
1095 Protocol (version 3) Management Information Base",
1096 Work in Progress, August 2006.
1097
1098 [<a id="ref-BFD-MULTIHOP">BFD-MULTIHOP</a>] Katz, D. and D. Ward, <a style="text-decoration: none" href='https://www.google.com/search?sitesearch=datatracker.ietf.org%2Fdoc%2Fhtml%2F&amp;q=inurl:draft-+%22BFD+for+Multihop+Paths%22'>"BFD for Multihop Paths"</a>, Work
1099 in Progress, March 2007.
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116 <span class="grey">Jain, et al. Standards Track [Page 20]</span></pre>
1117 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-21" ></span>
1118 <span class="grey"><a href="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
1119
1120
1121 Appendix A
1122
1123 Description below outlines the failover protocol operation for an
1124 example tunnel. The failover protocol does not preclude an endpoint
1125 from recovering multiple tunnels in parallel. It also allows an
1126 endpoint to send multiple FSQs, each including multiple FSS AVPs, to
1127 recover quickly.
1128
1129 Failover Capability Negotiation (<a href="#section-3.1">Section 3.1</a>):
1130
1131 Endpoint Peer
1132 (assigned tid = x, failover capable)
1133 SCCRQ --------------------------------------&gt; validate SCCRQ
1134
1135 (assigned tid = y, failover capable)
1136 validate &lt;-------------------------------------- send SCCRP
1137 SCCRP, etc.
1138
1139 .... &lt;after tunnel gets created, sessions are established&gt; ....
1140
1141
1142 &lt; This Node fails &gt;
1143
1144 The Recovery endpoint establishes the recovery tunnel (<a href="#section-3.2.1">Section 3.2.1</a>).
1145 Initiate recovery tunnel establishment for the old tunnel 'x':
1146
1147 Recovery Endpoint Peer
1148
1149 (assigned tid = z, Recovery AVP)
1150 SCCRQ -----------------------------------&gt; Detects failover
1151 (recover tid = x, recover remote tid = y) validate SCCRQ
1152
1153
1154 (Suggested Control Sequence AVP, Suggested Ns/Nr = 3/100)
1155 validate &lt;----------------------------------- send SCCRP
1156 SCCRP (recover tid = y, recover remote tid = x)
1157 reset Ns = 3, Nr = 100
1158 on the recovered tunnel
1159
1160 SCCCN -----------------------------------&gt; validate and reset
1161 Ns = 100, Nr = 3 on
1162 the recovered tunnel
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172 <span class="grey">Jain, et al. Standards Track [Page 21]</span></pre>
1173 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-22" ></span>
1174 <span class="grey"><a href="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
1175
1176
1177 Terminate the recovery tunnel
1178
1179 tid = 'z'
1180 StopCCN --------------------------------------&gt; Cleanup 'w'
1181
1182
1183 Session states are synchronized both endpoints may send FSQs and
1184 cleanup stale sessions (<a href="#section-3.3">Section 3.3</a>)
1185
1186 (FSS AVP for sessions s1, s2, s3..)
1187 send FSQ -------------------------------------&gt; compute the state
1188 of sessions in FSQ
1189
1190 (FSS AVP for sessions s1, s2, s3...)
1191 deletes &lt;-------------------------------------- send FSR
1192 stale sessions, if any
1193
1194
1195 (FSS AVP for sessions s7, s8, s9...)
1196 compute &lt;-------------------------------------- send FSQ
1197 the sate of
1198 sessions in FSQ
1199
1200
1201 (FSS AVP for sessions s7, s8, s9...)
1202 send FSR --------------------------------------&gt; delete stale
1203 sessions, if any
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228 <span class="grey">Jain, et al. Standards Track [Page 22]</span></pre>
1229 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-23" ></span>
1230 <span class="grey"><a href="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
1231
1232
1233 Appendix B
1234
1235 This section shows an example dialogue to illustrate double failure
1236 recovery. The notable difference, as described in <a href="#section-3.2.1">Section 3.2.1</a>, in
1237 the procedure from single failover scenario is the use of a tie
1238 breaker by one of the recovery endpoints to use the recovery tunnel
1239 established by its peer (also a recovery endpoint) as a recovery
1240 tunnel.
1241
1242 Recovery endpoint Recovery endpoint
1243
1244 (assume old tid = A) (assume old tid = B)
1245
1246 Recovery AVP = (A, B)
1247 SCCRQ -----------------------+
1248 (with tie (recovery tunnel 'C') |
1249 breaker |
1250 AVP) |
1251 Recovery AVP = (B, A) |
1252 +- valid &lt;--------------------------- Send SCCRQ
1253 | SCCRQ (recovery tunnel 'D') | (with tie breaker AVP)
1254 | This endpoint |
1255 | loses tie; |
1256 | Discards tunnel 'C' +--&gt; Valid SCCRQ
1257 | This endpoint wins tie;
1258 | Discards SCCRQ
1259 |
1260 | (may include SCS AVP)
1261 +-&gt;Send SCCRP -------------------------&gt; Validate SCCRP
1262 Reset 'B';
1263 Set Ns, Nr values --+
1264 |
1265 |
1266 |
1267 Validate SCCN &lt;---------------------- Send SCCN -------+
1268 Reset 'A';
1269 Set Ns, Nr values
1270
1271 FSQs and FSRs for the old tunnel (A, B) are exchanged on the
1272 recovered tunnel by both endpoints.
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284 <span class="grey">Jain, et al. Standards Track [Page 23]</span></pre>
1285 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-24" ></span>
1286 <span class="grey"><a href="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
1287
1288
1289 Appendix C
1290
1291 Session ID mismatch could not be a result of failure on one of the
1292 endpoints. However, failover session recovery procedure could
1293 exacerbate the situation, resulting into a permanent mismatch in
1294 Session IDs between two endpoints. The dialogue below outlines the
1295 behavior described in <a href="#section-3.3">Section 3.3</a>, Step III to handle such situations
1296 gracefully.
1297
1298 Recovery endpoint Remote endpoint
1299
1300 (assume a mismatch) (assume a mismatch)
1301 Sid = A, Remote Sid = B Sid = B, Remote Sid = C
1302 Sid = C, Remote Sid = D
1303
1304
1305 FSS AVP (A, B)
1306 send FSQ -------------------------&gt; No (B, A) pair exist;
1307 rather (B, C) exist.
1308 If it clears B then peer doesn't
1309 know if C is stale on other end.
1310
1311 Instead if it marks B stale
1312 and queries the session state
1313 via FSQ, C would be cleared on
1314 the other end.
1315
1316 FSS AVP (0, A)
1317 Clears A &lt;-------------------------- send FSR
1318
1319 ... some time later ...
1320
1321 FSS AVP (B, C)
1322 No (C,B) &lt;-------------------------- send FSQ
1323 Mark C Stale
1324
1325 FSS AVP (0, B)
1326 Send FSR --------------------------&gt; Clears B
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340 <span class="grey">Jain, et al. Standards Track [Page 24]</span></pre>
1341 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-25" ></span>
1342 <span class="grey"><a href="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
1343
1344
1345 Author Information
1346
1347 Vipin Jain
1348 Riverstone Networks
1349 5200 Great America Parkway
1350 Santa Clara, CA 95054
1351 EMail: [email protected]
1352
1353 Paul W. Howard
1354 Juniper Networks
1355 10 Technology Park Drive
1356 Westford, MA 01886
1357 EMail: [email protected]
1358
1359 Sam Henderson
1360 Cisco Systems
1361 7025 Kit Creek Rd.
1362 PO Box 14987
1363 Research Triangle Park, NC 27709
1364 EMail: [email protected]
1365
1366 Keyur Parikh
1367 Harris Corporation
1368 4393 Digitalway
1369 Mason, OH 45040
1370 EMail: [email protected]
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396 <span class="grey">Jain, et al. Standards Track [Page 25]</span></pre>
1397 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-26" ></span>
1398 <span class="grey"><a href="./rfc4951">RFC 4951</a> FAILOVER August 2007</span>
1399
1400
1401 Full Copyright Statement
1402
1403 Copyright (C) The IETF Trust (2007).
1404
1405 This document is subject to the rights, licenses and restrictions
1406 contained in <a href="https://www.rfc-editor.org/bcp/bcp78">BCP 78</a>, and except as set forth therein, the authors
1407 retain all their rights.
1408
1409 This document and the information contained herein are provided on an
1410 "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
1411 OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND
1412 THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS
1413 OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF
1414 THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
1415 WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
1416
1417 Intellectual Property
1418
1419 The IETF takes no position regarding the validity or scope of any
1420 Intellectual Property Rights or other rights that might be claimed to
1421 pertain to the implementation or use of the technology described in
1422 this document or the extent to which any license under such rights
1423 might or might not be available; nor does it represent that it has
1424 made any independent effort to identify any such rights. Information
1425 on the procedures with respect to rights in RFC documents can be
1426 found in <a href="https://www.rfc-editor.org/bcp/bcp78">BCP 78</a> and <a href="https://www.rfc-editor.org/bcp/bcp79">BCP 79</a>.
1427
1428 Copies of IPR disclosures made to the IETF Secretariat and any
1429 assurances of licenses to be made available, or the result of an
1430 attempt made to obtain a general license or permission for the use of
1431 such proprietary rights by implementers or users of this
1432 specification can be obtained from the IETF on-line IPR repository at
1433 <a href="http://www.ietf.org/ipr">http://www.ietf.org/ipr</a>.
1434
1435 The IETF invites any interested party to bring to its attention any
1436 copyrights, patents or patent applications, or other proprietary
1437 rights that may cover technology that may be required to implement
1438 this standard. Please address the information to the IETF at
1439 [email protected].
1440
1441 Acknowledgement
1442
1443 Funding for the RFC Editor function is currently provided by the
1444 Internet Society.
1445
1446
1447
1448
1449
1450
1451
1452 Jain, et al. Standards Track [Page 26]
1453 </pre>