Debian Codegraph @

rfc2453.html

raw
1 <pre>Network Working Group G. Malkin
2 Request for Comments: 2453 Bay Networks
3 Obsoletes: <a href="./rfc1723">1723</a>, <a href="./rfc1388">1388</a> November 1998
4 STD: 56
5 Category: Standards Track
6
7
8 <span class="h1">RIP Version 2</span>
9
10 Status of this Memo
11
12 This document specifies an Internet standards track protocol for the
13 Internet community, and requests discussion and suggestions for
14 improvements. Please refer to the current edition of the "Internet
15 Official Protocol Standards" (STD 1) for the standardization state
16 and status of this protocol. Distribution of this memo is unlimited.
17
18 Copyright Notice
19
20 Copyright (C) The Internet Society (1998). All Rights Reserved.
21
22 Abstract
23
24 This document specifies an extension of the Routing Information
25 Protocol (RIP), as defined in [<a href="#ref-1" title="&quot;Routing Information Protocol&quot;">1</a>], to expand the amount of useful
26 information carried in RIP messages and to add a measure of security.
27
28 A companion document will define the SNMP MIB objects for RIP-2 [<a href="#ref-2" title="&quot;RIP Version 2 MIB Extension&quot;">2</a>].
29 An additional document will define cryptographic security
30 improvements for RIP-2 [<a href="#ref-3" title="&quot;RIP-II MD5 Authentication&quot;">3</a>].
31
32 Acknowledgements
33
34 I would like to thank the IETF RIP Working Group for their help in
35 improving the RIP-2 protocol. Much of the text for the background
36 discussions about distance vector protocols and some of the
37 descriptions of the operation of RIP were taken from "Routing
38 Information Protocol" by C. Hedrick [<a href="#ref-1" title="&quot;Routing Information Protocol&quot;">1</a>]. Some of the final editing on
39 the document was done by Scott Bradner.
40
41
42
43
44
45
46
47
48
49
50
51
52 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
55
56
57 Table of Contents
58
59 <a href="#section-1">1</a>. Justification . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-3">3</a>
60 <a href="#section-2">2</a>. Current RIP . . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-3">3</a>
61 <a href="#section-3">3</a>. Basic Protocol . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-3">3</a>
62 <a href="#section-3.1">3.1</a> Introduction . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-3">3</a>
63 <a href="#section-3.2">3.2</a> Limitations of the Protocol . . . . . . . . . . . . . . . . <a href="#page-5">5</a>
64 <a href="#section-3.3">3.3</a>. Organization of this document . . . . . . . . . . . . . . . <a href="#page-6">6</a>
65 <a href="#section-3.4">3.4</a> Distance Vector Algorithms . . . . . . . . . . . . . . . . . <a href="#page-6">6</a>
66 <a href="#section-3.4.1">3.4.1</a> Dealing with changes in topology . . . . . . . . . . . . <a href="#page-12">12</a>
67 <a href="#section-3.4.2">3.4.2</a> Preventing instability . . . . . . . . . . . . . . . . . <a href="#page-13">13</a>
68 <a href="#section-3.4.3">3.4.3</a> Split horizon . . . . . . . . . . . . . . . . . . . . . . <a href="#page-15">15</a>
69 <a href="#section-3.4.4">3.4.4</a> Triggered updates . . . . . . . . . . . . . . . . . . . . <a href="#page-17">17</a>
70 <a href="#section-3.5">3.5</a> Protocol Specification . . . . . . . . . . . . . . . . . . <a href="#page-18">18</a>
71 <a href="#section-3.6">3.6</a> Message Format . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-20">20</a>
72 <a href="#section-3.7">3.7</a> Addressing Considerations . . . . . . . . . . . . . . . . . <a href="#page-22">22</a>
73 <a href="#section-3.8">3.8</a> Timers . . . . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-24">24</a>
74 <a href="#section-3.9">3.9</a> Input Processing . . . . . . . . . . . . . . . . . . . . . . <a href="#page-25">25</a>
75 <a href="#section-3.9.1">3.9.1</a> Request Messages . . . . . . . . . . . . . . . . . . . . <a href="#page-25">25</a>
76 <a href="#section-3.9.2">3.9.2</a> Response Messages . . . . . . . . . . . . . . . . . . . . <a href="#page-26">26</a>
77 <a href="#section-3.10">3.10</a> Output Processing . . . . . . . . . . . . . . . . . . . . . <a href="#page-28">28</a>
78 <a href="#section-3.10.1">3.10.1</a> Triggered Updates . . . . . . . . . . . . . . . . . . . . <a href="#page-29">29</a>
79 <a href="#section-3.10.2">3.10.2</a> Generating Response Messages. . . . . . . . . . . . . . . <a href="#page-30">30</a>
80 <a href="#section-4">4</a>. Protocol Extensions . . . . . . . . . . . . . . . . . . . . . <a href="#page-31">31</a>
81 <a href="#section-4.1">4.1</a> Authentication . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-31">31</a>
82 <a href="#section-4.2">4.2</a> Route Tag . . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-32">32</a>
83 <a href="#section-4.3">4.3</a> Subnet Mask . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-32">32</a>
84 <a href="#section-4.4">4.4</a> Next Hop . . . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-33">33</a>
85 <a href="#section-4.5">4.5</a> Multicasting . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-33">33</a>
86 <a href="#section-4.6">4.6</a> Queries . . . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-33">33</a>
87 <a href="#section-5">5</a>. Compatibility . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-34">34</a>
88 <a href="#section-5.1">5.1</a> Compatibility Switch . . . . . . . . . . . . . . . . . . . . <a href="#page-34">34</a>
89 <a href="#section-5.2">5.2</a> Authentication . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-34">34</a>
90 <a href="#section-5.3">5.3</a> Larger Infinity . . . . . . . . . . . . . . . . . . . . . . <a href="#page-35">35</a>
91 <a href="#section-5.4">5.4</a> Addressless Links . . . . . . . . . . . . . . . . . . . . . <a href="#page-35">35</a>
92 <a href="#section-6">6</a>. Interaction between version 1 and version 2 . . . . . . . . . <a href="#page-35">35</a>
93 <a href="#section-7">7</a>. Security Considerations . . . . . . . . . . . . . . . . . . . <a href="#page-36">36</a>
94 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-37">37</a>
95 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-37">37</a>
96 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . <a href="#page-38">38</a>
97 Full Copyright Statement . . . . . . . . . . . . . . . . . . . . . <a href="#page-39">39</a>
98
99
100
101
102
103
104
105
106
107
108 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
111
112
113 <span class="h2"><a class="selflink" id="section-1" href="#section-1">1</a>. Justification</span>
114
115 With the advent of OSPF and IS-IS, there are those who believe that
116 RIP is obsolete. While it is true that the newer IGP routing
117 protocols are far superior to RIP, RIP does have some advantages.
118 Primarily, in a small network, RIP has very little overhead in terms
119 of bandwidth used and configuration and management time. RIP is also
120 very easy to implement, especially in relation to the newer IGPs.
121
122 Additionally, there are many, many more RIP implementations in the
123 field than OSPF and IS-IS combined. It is likely to remain that way
124 for some years yet.
125
126 Given that RIP will be useful in many environments for some period of
127 time, it is reasonable to increase RIP's usefulness. This is
128 especially true since the gain is far greater than the expense of the
129 change.
130
131 <span class="h2"><a class="selflink" id="section-2" href="#section-2">2</a>. Current RIP</span>
132
133 The current RIP-1 message contains the minimal amount of information
134 necessary for routers to route messages through a network. It also
135 contains a large amount of unused space, owing to its origins.
136
137 The current RIP-1 protocol does not consider autonomous systems and
138 IGP/EGP interactions, subnetting [<a href="#ref-11" title="&quot;Requirements for IP Version 4 Routers.&quot;">11</a>], and authentication since
139 implementations of these postdate RIP-1. The lack of subnet masks is
140 a particularly serious problem for routers since they need a subnet
141 mask to know how to determine a route. If a RIP-1 route is a network
142 route (all non-network bits 0), the subnet mask equals the network
143 mask. However, if some of the non-network bits are set, the router
144 cannot determine the subnet mask. Worse still, the router cannot
145 determine if the RIP-1 route is a subnet route or a host route.
146 Currently, some routers simply choose the subnet mask of the
147 interface over which the route was learned and determine the route
148 type from that.
149
150 <span class="h2"><a class="selflink" id="section-3" href="#section-3">3</a>. Basic Protocol</span>
151
152 <span class="h3"><a class="selflink" id="section-3.1" href="#section-3.1">3.1</a> Introduction</span>
153
154 RIP is a routing protocol based on the Bellman-Ford (or distance
155 vector) algorithm. This algorithm has been used for routing
156 computations in computer networks since the early days of the
157 ARPANET. The particular packet formats and protocol described here
158 are based on the program "routed," which is included with the
159 Berkeley distribution of Unix.
160
161
162
163
164 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
167
168
169 In an international network, such as the Internet, it is very
170 unlikely that a single routing protocol will used for the entire
171 network. Rather, the network will be organized as a collection of
172 Autonomous Systems (AS), each of which will, in general, be
173 administered by a single entity. Each AS will have its own routing
174 technology, which may differ among AS's. The routing protocol used
175 within an AS is referred to as an Interior Gateway Protocol (IGP). A
176 separate protocol, called an Exterior Gateway Protocol (EGP), is used
177 to transfer routing information among the AS's. RIP was designed to
178 work as an IGP in moderate-size AS's. It is not intended for use in
179 more complex environments. For information on the context into which
180 RIP-1 is expected to fit, see Braden and Postel [<a href="#ref-6" title="&quot;Requirements for Internet Gateways&quot;">6</a>].
181
182 RIP uses one of a class of routing algorithms known as Distance
183 Vector algorithms. The earliest description of this class of
184 algorithms known to the author is in Ford and Fulkerson [<a href="#ref-8" title="&quot;Flows in Networks&quot;">8</a>]. Because
185 of this, they are sometimes known as Ford-Fulkerson algorithms. The
186 term Bellman-Ford is also used, and derives from the fact that the
187 formulation is based on Bellman's equation [<a href="#ref-4" title="&quot;Dynamic Programming&quot;">4</a>]. The presentation in
188 this document is closely based on [<a href="#ref-5" title="&quot;Data Networks&quot;">5</a>]. This document contains a
189 protocol specification. For an introduction to the mathematics of
190 routing algorithms, see [<a href="#ref-1" title="&quot;Routing Information Protocol&quot;">1</a>]. The basic algorithms used by this
191 protocol were used in computer routing as early as 1969 in the
192 ARPANET. However, the specific ancestry of this protocol is within
193 the Xerox network protocols. The PUP protocols [<a href="#ref-7" title="&quot;Pup: An Internetwork Architecture&quot;">7</a>] used the Gateway
194 Information Protocol to exchange routing information. A somewhat
195 updated version of this protocol was adopted for the Xerox Network
196 Systems (XNS) architecture, with the name Routing Information
197 Protocol [<a href="#ref-9" title="&quot;Internet Transport Protocols&quot;">9</a>]. Berkeley's routed is largely the same as the Routing
198 Information Protocol, with XNS addresses replaced by a more general
199 address format capable of handling IPv4 and other types of address,
200 and with routing updates limited to one every 30 seconds. Because of
201 this similarity, the term Routing Information Protocol (or just RIP)
202 is used to refer to both the XNS protocol and the protocol used by
203 routed.
204
205 RIP is intended for use within the IP-based Internet. The Internet
206 is organized into a number of networks connected by special purpose
207 gateways known as routers. The networks may be either point-to-point
208 links or more complex networks such as Ethernet or token ring. Hosts
209 and routers are presented with IP datagrams addressed to some host.
210 Routing is the method by which the host or router decides where to
211 send the datagram. It may be able to send the datagram directly to
212 the destination, if that destination is on one of the networks that
213 are directly connected to the host or router. However, the
214 interesting case is when the destination is not directly reachable.
215
216
217
218
219
220 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
223
224
225 In this case, the host or router attempts to send the datagram to a
226 router that is nearer the destination. The goal of a routing
227 protocol is very simple: It is to supply the information that is
228 needed to do routing.
229
230 <span class="h3"><a class="selflink" id="section-3.2" href="#section-3.2">3.2</a> Limitations of the Protocol</span>
231
232 This protocol does not solve every possible routing problem. As
233 mentioned above, it is primary intended for use as an IGP in networks
234 of moderate size. In addition, the following specific limitations
235 are be mentioned:
236
237 - The protocol is limited to networks whose longest path (the
238 network's diameter) is 15 hops. The designers believe that the
239 basic protocol design is inappropriate for larger networks. Note
240 that this statement of the limit assumes that a cost of 1 is used
241 for each network. This is the way RIP is normally configured. If
242 the system administrator chooses to use larger costs, the upper
243 bound of 15 can easily become a problem.
244
245 - The protocol depends upon "counting to infinity" to resolve certain
246 unusual situations. (This will be explained in the next section.)
247 If the system of networks has several hundred networks, and a
248 routing loop was formed involving all of them, the resolution of
249 the loop would require either much time (if the frequency of
250 routing updates were limited) or bandwidth (if updates were sent
251 whenever changes were detected). Such a loop would consume a large
252 amount of network bandwidth before the loop was corrected. We
253 believe that in realistic cases, this will not be a problem except
254 on slow lines. Even then, the problem will be fairly unusual,
255 since various precautions are taken that should prevent these
256 problems in most cases.
257
258 - This protocol uses fixed "metrics" to compare alternative routes.
259 It is not appropriate for situations where routes need to be chosen
260 based on real-time parameters such a measured delay, reliability,
261 or load. The obvious extensions to allow metrics of this type are
262 likely to introduce instabilities of a sort that the protocol is
263 not designed to handle.
264
265
266
267
268
269
270
271
272
273
274
275
276 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
279
280
281 <span class="h3"><a class="selflink" id="section-3.3" href="#section-3.3">3.3</a>. Organization of this document</span>
282
283 The main body of this document is organized into two parts, which
284 occupy the next two sections:
285
286 A conceptual development and justification of distance vector
287 algorithms in general.
288
289 The actual protocol description.
290
291 Each of these two sections can largely stand on its own. <a href="#section-3.4">Section 3.4</a>
292 attempts to give an informal presentation of the mathematical
293 underpinnings of the algorithm. Note that the presentation follows a
294 "spiral" method. An initial, fairly simple algorithm is described.
295 Then refinements are added to it in successive sections. <a href="#section-3.5">Section 3.5</a>
296 is the actual protocol description. Except where specific references
297 are made to <a href="#section-3.4">section 3.4</a>, it should be possible to implement RIP
298 entirely from the specifications given in <a href="#section-3.5">section 3.5</a>.
299
300 <span class="h3"><a class="selflink" id="section-3.4" href="#section-3.4">3.4</a> Distance Vector Algorithms</span>
301
302 Routing is the task of finding a path from a sender to a desired
303 destination. In the IP "Internet model" this reduces primarily to a
304 matter of finding a series of routers between the source and
305 destination networks. As long as a message or datagram remains on a
306 single network or subnet, any forwarding problems are the
307 responsibility of technology that is specific to the network. For
308 example, Ethernet and the ARPANET each define a way in which any
309 sender can talk to any specified destination within that one network.
310 IP routing comes in primarily when messages must go from a sender on
311 one network to a destination on a different one. In that case, the
312 message must pass through one or more routers connecting the
313 networks. If the networks are not adjacent, the message may pass
314 through several intervening networks, and the routers connecting
315 them. Once the message gets to a router that is on the same network
316 as the destination, that network's own technology is used to get to
317 the destination.
318
319 Throughout this section, the term "network" is used generically to
320 cover a single broadcast network (e.g., an Ethernet), a point to
321 point line, or the ARPANET. The critical point is that a network is
322 treated as a single entity by IP. Either no forwarding decision is
323 necessary (as with a point to point line), or that forwarding is done
324 in a manner that is transparent to IP, allowing IP to treat the
325 entire network as a single fully-connected system (as with an
326 Ethernet or the ARPANET). Note that the term "network" is used in a
327 somewhat different way in discussions of IP addressing. We are using
328 the term "network" here to refer to subnets in cases where subnet
329
330
331
332 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
335
336
337 addressing is in use.
338
339 A number of different approaches for finding routes between networks
340 are possible. One useful way of categorizing these approaches is on
341 the basis of the type of information the routers need to exchange in
342 order to be able to find routes. Distance vector algorithms are
343 based on the exchange of only a small amount of information. Each
344 entity (router or host) that participates in the routing protocol is
345 assumed to keep information about all of the destinations within the
346 system. Generally, information about all entities connected to one
347 network is summarized by a single entry, which describes the route to
348 all destinations on that network. This summarization is possible
349 because as far as IP is concerned, routing within a network is
350 invisible. Each entry in this routing database includes the next
351 router to which datagrams destined for the entity should be sent. In
352 addition, it includes a "metric" measuring the total distance to the
353 entity. Distance is a somewhat generalized concept, which may cover
354 the time delay in getting messages to the entity, the dollar cost of
355 sending messages to it, etc. Distance vector algorithms get their
356 name from the fact that it is possible to compute optimal routes when
357 the only information exchanged is the list of these distances.
358 Furthermore, information is only exchanged among entities that are
359 adjacent, that is, entities that share a common network.
360
361 Although routing is most commonly based on information about
362 networks, it is sometimes necessary to keep track of the routes to
363 individual hosts. The RIP protocol makes no formal distinction
364 between networks and hosts. It simply describes exchange of
365 information about destinations, which may be either networks or
366 hosts. (Note however, that it is possible for an implementor to
367 choose not to support host routes. See <a href="#section-3.2">section 3.2</a>.) In fact, the
368 mathematical developments are most conveniently thought of in terms
369 of routes from one host or router to another. When discussing the
370 algorithm in abstract terms, it is best to think of a routing entry
371 for a network as an abbreviation for routing entries for all of the
372 entities connected to that network. This sort of abbreviation makes
373 sense only because we think of networks as having no internal
374 structure that is visible at the IP level. Thus, we will generally
375 assign the same distance to every entity in a given network.
376
377 We said above that each entity keeps a routing database with one
378 entry for every possible destination in the system. An actual
379 implementation is likely to need to keep the following information
380 about each destination:
381
382
383
384
385
386
387
388 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
391
392
393 - address: in IP implementations of these algorithms, this will be
394 the IP address of the host or network.
395
396 - router: the first router along the route to the destination.
397
398 - interface: the physical network which must be used to reach the
399 first router.
400
401 - metric: a number, indicating the distance to the destination.
402
403 - timer: the amount of time since the entry was last updated.
404
405 In addition, various flags and other internal information will
406 probably be included. This database is initialized with a
407 description of the entities that are directly connected to the
408 system. It is updated according to information received in messages
409 from neighboring routers.
410
411 The most important information exchanged by the hosts and routers is
412 carried in update messages. Each entity that participates in the
413 routing scheme sends update messages that describe the routing
414 database as it currently exists in that entity. It is possible to
415 maintain optimal routes for the entire system by using only
416 information obtained from neighboring entities. The algorithm used
417 for that will be described in the next section.
418
419 As we mentioned above, the purpose of routing is to find a way to get
420 datagrams to their ultimate destinations. Distance vector algorithms
421 are based on a table in each router listing the best route to every
422 destination in the system. Of course, in order to define which route
423 is best, we have to have some way of measuring goodness. This is
424 referred to as the "metric".
425
426 In simple networks, it is common to use a metric that simply counts
427 how many routers a message must go through. In more complex
428 networks, a metric is chosen to represent the total amount of delay
429 that the message suffers, the cost of sending it, or some other
430 quantity which may be minimized. The main requirement is that it
431 must be possible to represent the metric as a sum of "costs" for
432 individual hops.
433
434 Formally, if it is possible to get from entity i to entity j directly
435 (i.e., without passing through another router between), then a cost,
436 d(i,j), is associated with the hop between i and j. In the normal
437 case where all entities on a given network are considered to be the
438 same, d(i,j) is the same for all destinations on a given network, and
439 represents the cost of using that network. To get the metric of a
440 complete route, one just adds up the costs of the individual hops
441
442
443
444 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
447
448
449 that make up the route. For the purposes of this memo, we assume
450 that the costs are positive integers.
451
452 Let D(i,j) represent the metric of the best route from entity i to
453 entity j. It should be defined for every pair of entities. d(i,j)
454 represents the costs of the individual steps. Formally, let d(i,j)
455 represent the cost of going directly from entity i to entity j. It
456 is infinite if i and j are not immediate neighbors. (Note that d(i,i)
457 is infinite. That is, we don't consider there to be a direct
458 connection from a node to itself.) Since costs are additive, it is
459 easy to show that the best metric must be described by
460
461 D(i,i) = 0, all i
462 D(i,j) = min [d(i,k) + D(k,j)], otherwise
463 k
464 and that the best routes start by going from i to those neighbors k
465 for which d(i,k) + D(k,j) has the minimum value. (These things can
466 be shown by induction on the number of steps in the routes.) Note
467 that we can limit the second equation to k's that are immediate
468 neighbors of i. For the others, d(i,k) is infinite, so the term
469 involving them can never be the minimum.
470
471 It turns out that one can compute the metric by a simple algorithm
472 based on this. Entity i gets its neighbors k to send it their
473 estimates of their distances to the destination j. When i gets the
474 estimates from k, it adds d(i,k) to each of the numbers. This is
475 simply the cost of traversing the network between i and k. Now and
476 then i compares the values from all of its neighbors and picks the
477 smallest.
478
479 A proof is given in [<a href="#ref-2" title="&quot;RIP Version 2 MIB Extension&quot;">2</a>] that this algorithm will converge to the
480 correct estimates of D(i,j) in finite time in the absence of topology
481 changes. The authors make very few assumptions about the order in
482 which the entities send each other their information, or when the min
483 is recomputed. Basically, entities just can't stop sending updates
484 or recomputing metrics, and the networks can't delay messages
485 forever. (Crash of a routing entity is a topology change.) Also,
486 their proof does not make any assumptions about the initial estimates
487 of D(i,j), except that they must be non-negative. The fact that
488 these fairly weak assumptions are good enough is important. Because
489 we don't have to make assumptions about when updates are sent, it is
490 safe to run the algorithm asynchronously. That is, each entity can
491 send updates according to its own clock. Updates can be dropped by
492 the network, as long as they don't all get dropped. Because we don't
493 have to make assumptions about the starting condition, the algorithm
494 can handle changes. When the system changes, the routing algorithm
495 starts moving to a new equilibrium, using the old one as its starting
496 point. It is important that the algorithm will converge in finite
497
498
499
500 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
503
504
505 time no matter what the starting point. Otherwise certain kinds of
506 changes might lead to non-convergent behavior.
507
508 The statement of the algorithm given above (and the proof) assumes
509 that each entity keeps copies of the estimates that come from each of
510 its neighbors, and now and then does a min over all of the neighbors.
511 In fact real implementations don't necessarily do that. They simply
512 remember the best metric seen so far, and the identity of the
513 neighbor that sent it. They replace this information whenever they
514 see a better (smaller) metric. This allows them to compute the
515 minimum incrementally, without having to store data from all of the
516 neighbors.
517
518 There is one other difference between the algorithm as described in
519 texts and those used in real protocols such as RIP: the description
520 above would have each entity include an entry for itself, showing a
521 distance of zero. In fact this is not generally done. Recall that
522 all entities on a network are normally summarized by a single entry
523 for the network. Consider the situation of a host or router G that
524 is connected to network A. C represents the cost of using network A
525 (usually a metric of one). (Recall that we are assuming that the
526 internal structure of a network is not visible to IP, and thus the
527 cost of going between any two entities on it is the same.) In
528 principle, G should get a message from every other entity H on
529 network A, showing a cost of 0 to get from that entity to itself. G
530 would then compute C + 0 as the distance to H. Rather than having G
531 look at all of these identical messages, it simply starts out by
532 making an entry for network A in its table, and assigning it a metric
533 of C. This entry for network A should be thought of as summarizing
534 the entries for all other entities on network A. The only entity on
535 A that can't be summarized by that common entry is G itself, since
536 the cost of going from G to G is 0, not C. But since we never need
537 those 0 entries, we can safely get along with just the single entry
538 for network A. Note one other implication of this strategy: because
539 we don't need to use the 0 entries for anything, hosts that do not
540 function as routers don't need to send any update messages. Clearly
541 hosts that don't function as routers (i.e., hosts that are connected
542 to only one network) can have no useful information to contribute
543 other than their own entry D(i,i) = 0. As they have only the one
544 interface, it is easy to see that a route to any other network
545 through them will simply go in that interface and then come right
546 back out it. Thus the cost of such a route will be greater than the
547 best cost by at least C. Since we don't need the 0 entries, non-
548 routers need not participate in the routing protocol at all.
549
550 Let us summarize what a host or router G does. For each destination
551 in the system, G will keep a current estimate of the metric for that
552 destination (i.e., the total cost of getting to it) and the identity
553
554
555
556 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
559
560
561 of the neighboring router on whose data that metric is based. If the
562 destination is on a network that is directly connected to G, then G
563 simply uses an entry that shows the cost of using the network, and
564 the fact that no router is needed to get to the destination. It is
565 easy to show that once the computation has converged to the correct
566 metrics, the neighbor that is recorded by this technique is in fact
567 the first router on the path to the destination. (If there are
568 several equally good paths, it is the first router on one of them.)
569 This combination of destination, metric, and router is typically
570 referred to as a route to the destination with that metric, using
571 that router.
572
573 4.ne The method so far only has a way to lower the metric, as the
574 existing metric is kept until a smaller one shows up. It is possible
575 that the initial estimate might be too low. Thus, there must be a
576 way to increase the metric. It turns out to be sufficient to use the
577 following rule: suppose the current route to a destination has metric
578 D and uses router G. If a new set of information arrived from some
579 source other than G, only update the route if the new metric is
580 better than D. But if a new set of information arrives from G
581 itself, always update D to the new value. It is easy to show that
582 with this rule, the incremental update process produces the same
583 routes as a calculation that remembers the latest information from
584 all the neighbors and does an explicit minimum. (Note that the
585 discussion so far assumes that the network configuration is static.
586 It does not allow for the possibility that a system might fail.)
587
588 To summarize, here is the basic distance vector algorithm as it has
589 been developed so far. (Note that this is not a statement of the RIP
590 protocol. There are several refinements still to be added.) The
591 following procedure is carried out by every entity that participates
592 in the routing protocol. This must include all of the routers in the
593 system. Hosts that are not routers may participate as well.
594
595 - Keep a table with an entry for every possible destination in the
596 system. The entry contains the distance D to the destination, and
597 the first router G on the route to that network. Conceptually,
598 there should be an entry for the entity itself, with metric 0, but
599 this is not actually included.
600
601 - Periodically, send a routing update to every neighbor. The update
602 is a set of messages that contain all of the information from the
603 routing table. It contains an entry for each destination, with the
604 distance shown to that destination.
605
606 - When a routing update arrives from a neighbor G', add the cost
607 associated with the network that is shared with G'. (This should
608 be the network over which the update arrived.) Call the resulting
609
610
611
612 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
615
616
617 distance D'. Compare the resulting distances with the current
618 routing table entries. If the new distance D' for N is smaller
619 than the existing value D, adopt the new route. That is, change
620 the table entry for N to have metric D' and router G'. If G' is
621 the router from which the existing route came, i.e., G' = G, then
622 use the new metric even if it is larger than the old one.
623
624 <span class="h4"><a class="selflink" id="section-3.4.1" href="#section-3.4.1">3.4.1</a> Dealing with changes in topology</span>
625
626 The discussion above assumes that the topology of the network is
627 fixed. In practice, routers and lines often fail and come back up.
628 To handle this possibility, we need to modify the algorithm slightly.
629
630 The theoretical version of the algorithm involved a minimum over all
631 immediate neighbors. If the topology changes, the set of neighbors
632 changes. Therefore, the next time the calculation is done, the
633 change will be reflected. However, as mentioned above, actual
634 implementations use an incremental version of the minimization. Only
635 the best route to any given destination is remembered. If the router
636 involved in that route should crash, or the network connection to it
637 break, the calculation might never reflect the change. The algorithm
638 as shown so far depends upon a router notifying its neighbors if its
639 metrics change. If the router crashes, then it has no way of
640 notifying neighbors of a change.
641
642 In order to handle problems of this kind, distance vector protocols
643 must make some provision for timing out routes. The details depend
644 upon the specific protocol. As an example, in RIP every router that
645 participates in routing sends an update message to all its neighbors
646 once every 30 seconds. Suppose the current route for network N uses
647 router G. If we don't hear from G for 180 seconds, we can assume
648 that either the router has crashed or the network connecting us to it
649 has become unusable. Thus, we mark the route as invalid. When we
650 hear from another neighbor that has a valid route to N, the valid
651 route will replace the invalid one. Note that we wait for 180
652 seconds before timing out a route even though we expect to hear from
653 each neighbor every 30 seconds. Unfortunately, messages are
654 occasionally lost by networks. Thus, it is probably not a good idea
655 to invalidate a route based on a single missed message.
656
657 As we will see below, it is useful to have a way to notify neighbors
658 that there currently isn't a valid route to some network. RIP, along
659 with several other protocols of this class, does this through a
660 normal update message, by marking that network as unreachable. A
661 specific metric value is chosen to indicate an unreachable
662 destination; that metric value is larger than the largest valid
663 metric that we expect to see. In the existing implementation of RIP,
664 16 is used. This value is normally referred to as "infinity", since
665
666
667
668 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
671
672
673 it is larger than the largest valid metric. 16 may look like a
674 surprisingly small number. It is chosen to be this small for reasons
675 that we will see shortly. In most implementations, the same
676 convention is used internally to flag a route as invalid.
677
678 <span class="h4"><a class="selflink" id="section-3.4.2" href="#section-3.4.2">3.4.2</a> Preventing instability</span>
679
680 The algorithm as presented up to this point will always allow a host
681 or router to calculate a correct routing table. However, that is
682 still not quite enough to make it useful in practice. The proofs
683 referred to above only show that the routing tables will converge to
684 the correct values in finite time. They do not guarantee that this
685 time will be small enough to be useful, nor do they say what will
686 happen to the metrics for networks that become inaccessible.
687
688 It is easy enough to extend the mathematics to handle routes becoming
689 inaccessible. The convention suggested above will do that. We
690 choose a large metric value to represent "infinity". This value must
691 be large enough that no real metric would ever get that large. For
692 the purposes of this example, we will use the value 16. Suppose a
693 network becomes inaccessible. All of the immediately neighboring
694 routers time out and set the metric for that network to 16. For
695 purposes of analysis, we can assume that all the neighboring routers
696 have gotten a new piece of hardware that connects them directly to
697 the vanished network, with a cost of 16. Since that is the only
698 connection to the vanished network, all the other routers in the
699 system will converge to new routes that go through one of those
700 routers. It is easy to see that once convergence has happened, all
701 the routers will have metrics of at least 16 for the vanished
702 network. Routers one hop away from the original neighbors would end
703 up with metrics of at least 17; routers two hops away would end up
704 with at least 18, etc. As these metrics are larger than the maximum
705 metric value, they are all set to 16. It is obvious that the system
706 will now converge to a metric of 16 for the vanished network at all
707 routers.
708
709 Unfortunately, the question of how long convergence will take is not
710 amenable to quite so simple an answer. Before going any further, it
711 will be useful to look at an example (taken from [<a href="#ref-2" title="&quot;RIP Version 2 MIB Extension&quot;">2</a>]). Note that
712 what we are about to show will not happen with a correct
713 implementation of RIP. We are trying to show why certain features
714 are needed. In the following example the letters correspond to
715 routers, and the lines to networks.
716
717
718
719
720
721
722
723
724 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
727
728
729 A-----B
730 \ / \
731 \ / |
732 C / all networks have cost 1, except
733 | / for the direct link from C to D, which
734 |/ has cost 10
735 D
736 |&lt;=== target network
737
738 Each router will have a table showing a route to each network.
739
740 However, for purposes of this illustration, we show only the routes
741 from each router to the network marked at the bottom of the diagram.
742
743 D: directly connected, metric 1
744 B: route via D, metric 2
745 C: route via B, metric 3
746 A: route via B, metric 3
747
748 Now suppose that the link from B to D fails. The routes should now
749 adjust to use the link from C to D. Unfortunately, it will take a
750 while for this to this to happen. The routing changes start when B
751 notices that the route to D is no longer usable. For simplicity, the
752 chart below assumes that all routers send updates at the same time.
753 The chart shows the metric for the target network, as it appears in
754 the routing table at each router.
755
756 time ------&gt;
757
758 D: dir, 1 dir, 1 dir, 1 dir, 1 ... dir, 1 dir, 1
759 B: unreach C, 4 C, 5 C, 6 C, 11 C, 12
760 C: B, 3 A, 4 A, 5 A, 6 A, 11 D, 11
761 A: B, 3 C, 4 C, 5 C, 6 C, 11 C, 12
762
763 dir = directly connected
764 unreach = unreachable
765
766 Here's the problem: B is able to get rid of its failed route using a
767 timeout mechanism, but vestiges of that route persist in the system
768 for a long time. Initially, A and C still think they can get to D
769 via B. So, they keep sending updates listing metrics of 3. In the
770 next iteration, B will then claim that it can get to D via either A
771 or C. Of course, it can't. The routes being claimed by A and C are
772 now gone, but they have no way of knowing that yet. And even when
773 they discover that their routes via B have gone away, they each think
774 there is a route available via the other. Eventually the system
775 converges, as all the mathematics claims it must. But it can take
776 some time to do so. The worst case is when a network becomes
777
778
779
780 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
783
784
785 completely inaccessible from some part of the system. In that case,
786 the metrics may increase slowly in a pattern like the one above until
787 they finally reach infinity. For this reason, the problem is called
788 "counting to infinity".
789
790 You should now see why "infinity" is chosen to be as small as
791 possible. If a network becomes completely inaccessible, we want
792 counting to infinity to be stopped as soon as possible. Infinity
793 must be large enough that no real route is that big. But it
794 shouldn't be any bigger than required. Thus the choice of infinity
795 is a tradeoff between network size and speed of convergence in case
796 counting to infinity happens. The designers of RIP believed that the
797 protocol was unlikely to be practical for networks with a diameter
798 larger than 15.
799
800 There are several things that can be done to prevent problems like
801 this. The ones used by RIP are called "split horizon with poisoned
802 reverse", and "triggered updates".
803
804 <span class="h4"><a class="selflink" id="section-3.4.3" href="#section-3.4.3">3.4.3</a> Split horizon</span>
805
806 Note that some of the problem above is caused by the fact that A and
807 C are engaged in a pattern of mutual deception. Each claims to be
808 able to get to D via the other. This can be prevented by being a bit
809 more careful about where information is sent. In particular, it is
810 never useful to claim reachability for a destination network to the
811 neighbor(s) from which the route was learned. "Split horizon" is a
812 scheme for avoiding problems caused by including routes in updates
813 sent to the router from which they were learned. The "simple split
814 horizon" scheme omits routes learned from one neighbor in updates
815 sent to that neighbor. "Split horizon with poisoned reverse"
816 includes such routes in updates, but sets their metrics to infinity.
817
818 If A thinks it can get to D via C, its messages to C should indicate
819 that D is unreachable. If the route through C is real, then C either
820 has a direct connection to D, or a connection through some other
821 router. C's route can't possibly go back to A, since that forms a
822 loop. By telling C that D is unreachable, A simply guards against
823 the possibility that C might get confused and believe that there is a
824 route through A. This is obvious for a point to point line. But
825 consider the possibility that A and C are connected by a broadcast
826 network such as an Ethernet, and there are other routers on that
827 network. If A has a route through C, it should indicate that D is
828 unreachable when talking to any other router on that network. The
829 other routers on the network can get to C themselves. They would
830 never need to get to C via A. If A's best route is really through C,
831 no other router on that network needs to know that A can reach D.
832 This is fortunate, because it means that the same update message that
833
834
835
836 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
839
840
841 is used for C can be used for all other routers on the same network.
842 Thus, update messages can be sent by broadcast.
843
844 In general, split horizon with poisoned reverse is safer than simple
845 split horizon. If two routers have routes pointing at each other,
846 advertising reverse routes with a metric of 16 will break the loop
847 immediately. If the reverse routes are simply not advertised, the
848 erroneous routes will have to be eliminated by waiting for a timeout.
849 However, poisoned reverse does have a disadvantage: it increases the
850 size of the routing messages. Consider the case of a campus backbone
851 connecting a number of different buildings. In each building, there
852 is a router connecting the backbone to a local network. Consider
853 what routing updates those routers should broadcast on the backbone
854 network. All that the rest of the network really needs to know about
855 each router is what local networks it is connected to. Using simple
856 split horizon, only those routes would appear in update messages sent
857 by the router to the backbone network. If split horizon with
858 poisoned reverse is used, the router must mention all routes that it
859 learns from the backbone, with metrics of 16. If the system is
860 large, this can result in a large update message, almost all of whose
861 entries indicate unreachable networks.
862
863 In a static sense, advertising reverse routes with a metric of 16
864 provides no additional information. If there are many routers on one
865 broadcast network, these extra entries can use significant bandwidth.
866 The reason they are there is to improve dynamic behavior. When
867 topology changes, mentioning routes that should not go through the
868 router as well as those that should can speed up convergence.
869 However, in some situations, network managers may prefer to accept
870 somewhat slower convergence in order to minimize routing overhead.
871 Thus implementors may at their option implement simple split horizon
872 rather than split horizon with poisoned reverse, or they may provide
873 a configuration option that allows the network manager to choose
874 which behavior to use. It is also permissible to implement hybrid
875 schemes that advertise some reverse routes with a metric of 16 and
876 omit others. An example of such a scheme would be to use a metric of
877 16 for reverse routes for a certain period of time after routing
878 changes involving them, and thereafter omitting them from updates.
879
880 The router requirements RFC [<a href="#ref-11" title="&quot;Requirements for IP Version 4 Routers.&quot;">11</a>] specifies that all implementation of
881 RIP must use split horizon and should also use split horizon with
882 poisoned reverse, although there may be a knob to disable poisoned
883 reverse.
884
885
886
887
888
889
890
891
892 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
895
896
897 <span class="h4"><a class="selflink" id="section-3.4.4" href="#section-3.4.4">3.4.4</a> Triggered updates</span>
898
899 Split horizon with poisoned reverse will prevent any routing loops
900 that involve only two routers. However, it is still possible to end
901 up with patterns in which three routers are engaged in mutual
902 deception. For example, A may believe it has a route through B, B
903 through C, and C through A. Split horizon cannot stop such a loop.
904 This loop will only be resolved when the metric reaches infinity and
905 the network involved is then declared unreachable. Triggered updates
906 are an attempt to speed up this convergence. To get triggered
907 updates, we simply add a rule that whenever a router changes the
908 metric for a route, it is required to send update messages almost
909 immediately, even if it is not yet time for one of the regular update
910 message. (The timing details will differ from protocol to protocol.
911 Some distance vector protocols, including RIP, specify a small time
912 delay, in order to avoid having triggered updates generate excessive
913 network traffic.) Note how this combines with the rules for
914 computing new metrics. Suppose a router's route to destination N
915 goes through router G. If an update arrives from G itself, the
916 receiving router is required to believe the new information, whether
917 the new metric is higher or lower than the old one. If the result is
918 a change in metric, then the receiving router will send triggered
919 updates to all the hosts and routers directly connected to it. They
920 in turn may each send updates to their neighbors. The result is a
921 cascade of triggered updates. It is easy to show which routers and
922 hosts are involved in the cascade. Suppose a router G times out a
923 route to destination N. G will send triggered updates to all of its
924 neighbors. However, the only neighbors who will believe the new
925 information are those whose routes for N go through G. The other
926 routers and hosts will see this as information about a new route that
927 is worse than the one they are already using, and ignore it. The
928 neighbors whose routes go through G will update their metrics and
929 send triggered updates to all of their neighbors. Again, only those
930 neighbors whose routes go through them will pay attention. Thus, the
931 triggered updates will propagate backwards along all paths leading to
932 router G, updating the metrics to infinity. This propagation will
933 stop as soon as it reaches a portion of the network whose route to
934 destination N takes some other path.
935
936 If the system could be made to sit still while the cascade of
937 triggered updates happens, it would be possible to prove that
938 counting to infinity will never happen. Bad routes would always be
939 removed immediately, and so no routing loops could form.
940
941 Unfortunately, things are not so nice. While the triggered updates
942 are being sent, regular updates may be happening at the same time.
943 Routers that haven't received the triggered update yet will still be
944 sending out information based on the route that no longer exists. It
945
946
947
948 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
951
952
953 is possible that after the triggered update has gone through a
954 router, it might receive a normal update from one of these routers
955 that hasn't yet gotten the word. This could reestablish an orphaned
956 remnant of the faulty route. If triggered updates happen quickly
957 enough, this is very unlikely. However, counting to infinity is
958 still possible.
959
960 The router requirements RFC [<a href="#ref-11" title="&quot;Requirements for IP Version 4 Routers.&quot;">11</a>] specifies that all implementation of
961 RIP must implement triggered update for deleted routes and may
962 implement triggered updates for new routes or change of routes. RIP
963 implementations must also limit the rate which of triggered updates
964 may be trandmitted. (see <a href="#section-3.10.1">section 3.10.1</a>)
965
966 <span class="h3"><a class="selflink" id="section-3.5" href="#section-3.5">3.5</a> Protocol Specification</span>
967
968 RIP is intended to allow routers to exchange information for
969 computing routes through an IPv4-based network. Any router that uses
970 RIP is assumed to have interfaces to one or more networks, otherwise
971 it isn't really a router. These are referred to as its directly-
972 connected networks. The protocol relies on access to certain
973 information about each of these networks, the most important of which
974 is its metric. The RIP metric of a network is an integer between 1
975 and 15, inclusive. It is set in some manner not specified in this
976 protocol; however, given the maximum path limit of 15, a value of 1
977 is usually used. Implementations should allow the system
978 administrator to set the metric of each network. In addition to the
979 metric, each network will have an IPv4 destination address and subnet
980 mask associated with it. These are to be set by the system
981 administrator in a manner not specified in this protocol.
982
983 Any host that uses RIP is assumed to have interfaces to one or more
984 networks. These are referred to as its "directly-connected
985 networks". The protocol relies on access to certain information
986 about each of these networks. The most important is its metric or
987 "cost". The metric of a network is an integer between 1 and 15
988 inclusive. It is set in some manner not specified in this protocol.
989 Most existing implementations always use a metric of 1. New
990 implementations should allow the system administrator to set the cost
991 of each network. In addition to the cost, each network will have an
992 IPv4 network number and a subnet mask associated with it. These are
993 to be set by the system administrator in a manner not specified in
994 this protocol.
995
996 Note that the rules specified in <a href="#section-3.7">section 3.7</a> assume that there is a
997 single subnet mask applying to each IPv4 network, and that only the
998 subnet masks for directly-connected networks are known. There may be
999 systems that use different subnet masks for different subnets within
1000 a single network. There may also be instances where it is desirable
1001
1002
1003
1004 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1007
1008
1009 for a system to know the subnets masks of distant networks. Network-
1010 wide distribution of routing information which contains different
1011 subnet masks is permitted if all routers in the network are running
1012 the extensions presented in this document. However, if all routers in
1013 the network are not running these extensions distribution of routing
1014 information containing different subnet masks must be limited to
1015 avoid interoperability problems. See sections <a href="#section-3.7">3.7</a> and <a href="#section-4.3">4.3</a> for the
1016 rules governing subnet distribution.
1017
1018 Each router that implements RIP is assumed to have a routing table.
1019 This table has one entry for every destination that is reachable
1020 throughout the system operating RIP. Each entry contains at least
1021 the following information:
1022
1023 - The IPv4 address of the destination.
1024
1025 - A metric, which represents the total cost of getting a datagram
1026 from the router to that destination. This metric is the sum of the
1027 costs associated with the networks that would be traversed to get
1028 to the destination.
1029
1030 - The IPv4 address of the next router along the path to the
1031 destination (i.e., the next hop). If the destination is on one of
1032 the directly-connected networks, this item is not needed.
1033
1034 - A flag to indicate that information about the route has changed
1035 recently. This will be referred to as the "route change flag."
1036
1037 - Various timers associated with the route. See <a href="#section-3.6">section 3.6</a> for more
1038 details on timers.
1039
1040 The entries for the directly-connected networks are set up by the
1041 router using information gathered by means not specified in this
1042 protocol. The metric for a directly-connected network is set to the
1043 cost of that network. As mentioned, 1 is the usual cost. In that
1044 case, the RIP metric reduces to a simple hop-count. More complex
1045 metrics may be used when it is desirable to show preference for some
1046 networks over others (e.g., to indicate of differences in bandwidth
1047 or reliability).
1048
1049 To support the extensions detailed in this document, each entry must
1050 additionally contain a subnet mask. The subnet mask allows the router
1051 (along with the IPv4 address of the destination) to identify the
1052 different subnets within a single network as well as the subnets
1053 masks of distant networks.
1054
1055
1056
1057
1058
1059
1060 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1063
1064
1065 Implementors may also choose to allow the system administrator to
1066 enter additional routes. These would most likely be routes to hosts
1067 or networks outside the scope of the routing system. They are
1068 referred to as "static routes." Entries for destinations other than
1069 these initial ones are added and updated by the algorithms described
1070 in the following sections.
1071
1072 In order for the protocol to provide complete information on routing,
1073 every router in the AS must participate in the protocol. In cases
1074 where multiple IGPs are in use, there must be at least one router
1075 which can leak routing information between the protocols.
1076
1077 <span class="h3"><a class="selflink" id="section-3.6" href="#section-3.6">3.6</a> Message Format</span>
1078
1079 RIP is a UDP-based protocol. Each router that uses RIP has a routing
1080 process that sends and receives datagrams on UDP port number 520, the
1081 RIP-1/RIP-2 port. All communications intended for another routers's
1082 RIP process are sent to the RIP port. All routing update messages
1083 are sent from the RIP port. Unsolicited routing update messages have
1084 both the source and destination port equal to the RIP port. Update
1085 messages sent in response to a request are sent to the port from
1086 which the request came. Specific queries may be sent from ports
1087 other than the RIP port, but they must be directed to the RIP port on
1088 the target machine.
1089
1090 The RIP packet format is:
1091
1092 0 1 2 3
1093 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
1094 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1095 | command (1) | version (1) | must be zero (2) |
1096 +---------------+---------------+-------------------------------+
1097 | |
1098 ~ RIP Entry (20) ~
1099 | |
1100 +---------------+---------------+---------------+---------------+
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1119
1120
1121 There may be between 1 and 25 (inclusive) RIP entries. A RIP-1 entry
1122 has the following format:
1123
1124 0 1 2 3
1125 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
1126 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1127 | address family identifier (2) | must be zero (2) |
1128 +-------------------------------+-------------------------------+
1129 | IPv4 address (4) |
1130 +---------------------------------------------------------------+
1131 | must be zero (4) |
1132 +---------------------------------------------------------------+
1133 | must be zero (4) |
1134 +---------------------------------------------------------------+
1135 | metric (4) |
1136 +---------------------------------------------------------------+
1137
1138 Field sizes are given in octets. Unless otherwise specified, fields
1139 contain binary integers, in network byte order, with the most-
1140 significant octet first (big-endian). Each tick mark represents one
1141 bit.
1142
1143 Every message contains a RIP header which consists of a command and a
1144 version number. This section of the document describes version 1 of
1145 the protocol; <a href="#section-4">section 4</a> describes the version 2 extensions. The
1146 command field is used to specify the purpose of this message. The
1147 commands implemented in version 1 and 2 are:
1148
1149 1 - request A request for the responding system to send all or
1150 part of its routing table.
1151
1152 2 - response A message containing all or part of the sender's
1153 routing table. This message may be sent in response
1154 to a request, or it may be an unsolicited routing
1155 update generated by the sender.
1156
1157 For each of these message types, in version 1, the remainder of the
1158 datagram contains a list of Route Entries (RTEs). Each RTE in this
1159 list contains an Address Family Identifier (AFI), destination IPv4
1160 address, and the cost to reach that destination (metric).
1161
1162 The AFI is the type of address. For RIP-1, only AF_INET (2) is
1163 generally supported.
1164
1165 The metric field contains a value between 1 and 15 (inclusive) which
1166 specifies the current metric for the destination; or the value 16
1167 (infinity), which indicates that the destination is not reachable.
1168
1169
1170
1171
1172 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1175
1176
1177 <span class="h3"><a class="selflink" id="section-3.7" href="#section-3.7">3.7</a> Addressing Considerations</span>
1178
1179 Distance vector routing can be used to describe routes to individual
1180 hosts or to networks. The RIP protocol allows either of these
1181 possibilities. The destinations appearing in request and response
1182 messages can be networks, hosts, or a special code used to indicate a
1183 default address. In general, the kinds of routes actually used will
1184 depend upon the routing strategy used for the particular network.
1185 Many networks are set up so that routing information for individual
1186 hosts is not needed. If every node on a given network or subnet is
1187 accessible through the same routers, then there is no reason to
1188 mention individual hosts in the routing tables. However, networks
1189 that include point-to-point lines sometimes require routers to keep
1190 track of routes to certain nodes. Whether this feature is required
1191 depends upon the addressing and routing approach used in the system.
1192 Thus, some implementations may choose not to support host routes. If
1193 host routes are not supported, they are to be dropped when they are
1194 received in response messages (see <a href="#section-3.7.2">section 3.7.2</a>).
1195
1196 The RIP-1 packet format does not distinguish among various types of
1197 address. Fields that are labeled "address" can contain any of the
1198 following:
1199
1200 host address subnet number network number zero (default route)
1201
1202 Entities which use RIP-1 are assumed to use the most specific
1203 information available when routing a datagram. That is, when routing
1204 a datagram, its destination address must first be checked against the
1205 list of node addresses. Then it must be checked to see whether it
1206 matches any known subnet or network number. Finally, if none of
1207 these match, the default route is used.
1208
1209 When a node evaluates information that it receives via RIP-1, its
1210 interpretation of an address depends upon whether it knows the subnet
1211 mask that applies to the net. If so, then it is possible to
1212 determine the meaning of the address. For example, consider net
1213 128.6. It has a subnet mask of 255.255.255.0. Thus 128.6.0.0 is a
1214 network number, 128.6.4.0 is a subnet number, and 128.6.4.1 is a node
1215 address. However, if the node does not know the subnet mask,
1216 evaluation of an address may be ambiguous. If there is a non-zero
1217 node part, there is no clear way to determine whether the address
1218 represents a subnet number or a node address. As a subnet number
1219 would be useless without the subnet mask, addresses are assumed to
1220 represent nodes in this situation. In order to avoid this sort of
1221 ambiguity, when using version 1, nodes must not send subnet routes to
1222 nodes that cannot be expected to know the appropriate subnet mask.
1223 Normally hosts only know the subnet masks for directly-connected
1224 networks. Therefore, unless special provisions have been made,
1225
1226
1227
1228 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1231
1232
1233 routes to a subnet must not be sent outside the network of which the
1234 subnet is a part. RIP-2 (see <a href="#section-4">section 4</a>) eliminates the subnet/host
1235 ambiguity by including the subnet mask in the routing entry.
1236
1237 This "subnet filtering" is carried out by the routers at the "border"
1238 of the subnetted network. These are routers which connect that
1239 network with some other network. Within the subnetted network, each
1240 subnet is treated as an individual network. Routing entries for each
1241 subnet are circulated by RIP. However, border routers send only a
1242 single entry for the network as a whole to nodes in other networks.
1243 This means that a border router will send different information to
1244 different neighbors. For neighbors connected to the subnetted
1245 network, it generates a list of all subnets to which it is directly
1246 connected, using the subnet number. For neighbors connected to other
1247 networks, it makes a single entry for the network as a whole, showing
1248 the metric associated with that network. This metric would normally
1249 be the smallest metric for the subnets to which the router is
1250 attached.
1251
1252 Similarly, border routers must not mention host routes for nodes
1253 within one of the directly-connected networks in messages to other
1254 networks. Those routes will be subsumed by the single entry for the
1255 network as a whole.
1256
1257 The router requirements RFC [<a href="#ref-11" title="&quot;Requirements for IP Version 4 Routers.&quot;">11</a>] specifies that all implementation of
1258 RIP should support host routes but if they do not then they must
1259 ignore any received host routes.
1260
1261 The special address 0.0.0.0 is used to describe a default route. A
1262 default route is used when it is not convenient to list every
1263 possible network in the RIP updates, and when one or more closely-
1264 connected routers in the system are prepared to handle traffic to the
1265 networks that are not listed explicitly. These routers should create
1266 RIP entries for the address 0.0.0.0, just as if it were a network to
1267 which they are connected. The decision as to how routers create
1268 entries for 0.0.0.0 is left to the implementor. Most commonly, the
1269 system administrator will be provided with a way to specify which
1270 routers should create entries for 0.0.0.0; however, other mechanisms
1271 are possible. For example, an implementor might decide that any
1272 router which speaks BGP should be declared to be a default router.
1273 It may be useful to allow the network administrator to choose the
1274 metric to be used in these entries. If there is more than one
1275 default router, this will make it possible to express a preference
1276 for one over the other. The entries for 0.0.0.0 are handled by RIP
1277 in exactly the same manner as if there were an actual network with
1278 this address. System administrators should take care to make sure
1279 that routes to 0.0.0.0 do not propagate further than is intended.
1280 Generally, each autonomous system has its own preferred default
1281
1282
1283
1284 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1287
1288
1289 router. Thus, routes involving 0.0.0.0 should generally not leave
1290 the boundary of an autonomous system. The mechanisms for enforcing
1291 this are not specified in this document.
1292
1293 <span class="h3"><a class="selflink" id="section-3.8" href="#section-3.8">3.8</a> Timers</span>
1294
1295 This section describes all events that are triggered by timers.
1296
1297 Every 30 seconds, the RIP process is awakened to send an unsolicited
1298 Response message containing the complete routing table (see <a href="#section-3.9">section</a>
1299 <a href="#section-3.9">3.9</a> on Split Horizon) to every neighboring router. When there are
1300 many routers on a single network, there is a tendency for them to
1301 synchronize with each other such that they all issue updates at the
1302 same time. This can happen whenever the 30 second timer is affected
1303 by the processing load on the system. It is undesirable for the
1304 update messages to become synchronized, since it can lead to
1305 unnecessary collisions on broadcast networks. Therefore,
1306 implementations are required to take one of two precautions:
1307
1308 - The 30-second updates are triggered by a clock whose rate is not
1309 affected by system load or the time required to service the
1310 previous update timer.
1311
1312 - The 30-second timer is offset by a small random time (+/- 0 to 5
1313 seconds) each time it is set. (Implementors may wish to consider
1314 even larger variation in the light of recent research results [<a href="#ref-10" title="&quot;The synchronization of Periodic Routing Messages,&quot;">10</a>])
1315
1316 There are two timers associated with each route, a "timeout" and a
1317 "garbage-collection" time. Upon expiration of the timeout, the route
1318 is no longer valid; however, it is retained in the routing table for
1319 a short time so that neighbors can be notified that the route has
1320 been dropped. Upon expiration of the garbage-collection timer, the
1321 route is finally removed from the routing table.
1322
1323 The timeout is initialized when a route is established, and any time
1324 an update message is received for the route. If 180 seconds elapse
1325 from the last time the timeout was initialized, the route is
1326 considered to have expired, and the deletion process described below
1327 begins for that route.
1328
1329 Deletions can occur for one of two reasons: the timeout expires, or
1330 the metric is set to 16 because of an update received from the
1331 current router (see <a href="#section-3.7.2">section 3.7.2</a> for a discussion of processing
1332 updates from other routers). In either case, the following events
1333 happen:
1334
1335
1336
1337
1338
1339
1340 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1343
1344
1345 - The garbage-collection timer is set for 120 seconds.
1346
1347 - The metric for the route is set to 16 (infinity). This causes the
1348 route to be removed from service.
1349
1350 - The route change flag is set to indicate that this entry has been
1351 changed.
1352
1353 - The output process is signalled to trigger a response.
1354
1355 Until the garbage-collection timer expires, the route is included in
1356 all updates sent by this router. When the garbage-collection timer
1357 expires, the route is deleted from the routing table.
1358
1359 Should a new route to this network be established while the garbage-
1360 collection timer is running, the new route will replace the one that
1361 is about to be deleted. In this case the garbage-collection timer
1362 must be cleared.
1363
1364 Triggered updates also use a small timer; however, this is best
1365 described in <a href="#section-3.9.1">section 3.9.1</a>.
1366
1367 <span class="h3"><a class="selflink" id="section-3.9" href="#section-3.9">3.9</a> Input Processing</span>
1368
1369 This section will describe the handling of datagrams received on the
1370 RIP port. Processing will depend upon the value in the command
1371 field.
1372
1373 See sections <a href="#section-4.6">4.6</a> and <a href="#section-5.1">5.1</a> for details on handling version numbers.
1374
1375 <span class="h4"><a class="selflink" id="section-3.9.1" href="#section-3.9.1">3.9.1</a> Request Messages</span>
1376
1377 A Request is used to ask for a response containing all or part of a
1378 router's routing table. Normally, Requests are sent as broadcasts
1379 (multicasts for RIP-2), from the RIP port, by routers which have just
1380 come up and are seeking to fill in their routing tables as quickly as
1381 possible. However, there may be situations (e.g., router monitoring)
1382 where the routing table of only a single router is needed. In this
1383 case, the Request should be sent directly to that router from a UDP
1384 port other than the RIP port. If such a Request is received, the
1385 router responds directly to the requestor's address and port.
1386
1387 The Request is processed entry by entry. If there are no entries, no
1388 response is given. There is one special case. If there is exactly
1389 one entry in the request, and it has an address family identifier of
1390 zero and a metric of infinity (i.e., 16), then this is a request to
1391 send the entire routing table. In that case, a call is made to the
1392 output process to send the routing table to the requesting
1393
1394
1395
1396 <span class="grey">Malkin 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="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1399
1400
1401 address/port. Except for this special case, processing is quite
1402 simple. Examine the list of RTEs in the Request one by one. For
1403 each entry, look up the destination in the router's routing database
1404 and, if there is a route, put that route's metric in the metric field
1405 of the RTE. If there is no explicit route to the specified
1406 destination, put infinity in the metric field. Once all the entries
1407 have been filled in, change the command from Request to Response and
1408 send the datagram back to the requestor.
1409
1410 Note that there is a difference in metric handling for specific and
1411 whole-table requests. If the request is for a complete routing
1412 table, normal output processing is done, including Split Horizon (see
1413 <a href="#section-3.9">section 3.9</a> on Split Horizon). If the request is for specific
1414 entries, they are looked up in the routing table and the information
1415 is returned as is; no Split Horizon processing is done. The reason
1416 for this distinction is the expectation that these requests are
1417 likely to be used for different purposes. When a router first comes
1418 up, it multicasts a Request on every connected network asking for a
1419 complete routing table. It is assumed that these complete routing
1420 tables are to be used to update the requestor's routing table. For
1421 this reason, Split Horizon must be done. It is further assumed that
1422 a Request for specific networks is made only by diagnostic software,
1423 and is not used for routing. In this case, the requester would want
1424 to know the exact contents of the routing table and would not want
1425 any information hidden or modified.
1426
1427 <span class="h4"><a class="selflink" id="section-3.9.2" href="#section-3.9.2">3.9.2</a> Response Messages</span>
1428
1429 A Response can be received for one of several different reasons:
1430
1431 - response to a specific query
1432 - regular update (unsolicited response)
1433 - triggered update caused by a route change
1434
1435 Processing is the same no matter why the Response was generated.
1436
1437 Because processing of a Response may update the router's routing
1438 table, the Response must be checked carefully for validity. The
1439 Response must be ignored if it is not from the RIP port. The
1440 datagram's IPv4 source address should be checked to see whether the
1441 datagram is from a valid neighbor; the source of the datagram must be
1442 on a directly-connected network. It is also worth checking to see
1443 whether the response is from one of the router's own addresses.
1444 Interfaces on broadcast networks may receive copies of their own
1445 broadcasts/multicasts immediately. If a router processes its own
1446 output as new input, confusion is likely so such datagrams must be
1447 ignored.
1448
1449
1450
1451
1452 <span class="grey">Malkin Standards Track [Page 26]</span></pre>
1453 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-27" ></span>
1454 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1455
1456
1457 Once the datagram as a whole has been validated, process the RTEs in
1458 the Response one by one. Again, start by doing validation.
1459 Incorrect metrics and other format errors usually indicate
1460 misbehaving neighbors and should probably be brought to the
1461 administrator's attention. For example, if the metric is greater
1462 than infinity, ignore the entry but log the event. The basic
1463 validation tests are:
1464
1465 - is the destination address valid (e.g., unicast; not net 0 or 127)
1466 - is the metric valid (i.e., between 1 and 16, inclusive)
1467
1468 If any check fails, ignore that entry and proceed to the next.
1469 Again, logging the error is probably a good idea.
1470
1471 Once the entry has been validated, update the metric by adding the
1472 cost of the network on which the message arrived. If the result is
1473 greater than infinity, use infinity. That is,
1474
1475 metric = MIN (metric + cost, infinity)
1476
1477 Now, check to see whether there is already an explicit route for the
1478 destination address. If there is no such route, add this route to
1479 the routing table, unless the metric is infinity (there is no point
1480 in adding a route which is unusable). Adding a route to the routing
1481 table consists of:
1482
1483 - Setting the destination address to the destination address in the
1484 RTE
1485
1486 - Setting the metric to the newly calculated metric (as described
1487 above)
1488
1489 - Set the next hop address to be the address of the router from which
1490 the datagram came
1491
1492 - Initialize the timeout for the route. If the garbage-collection
1493 timer is running for this route, stop it (see <a href="#section-3.6">section 3.6</a> for a
1494 discussion of the timers)
1495
1496 - Set the route change flag
1497
1498 - Signal the output process to trigger an update (see <a href="#section-3.8.1">section 3.8.1</a>)
1499
1500 If there is an existing route, compare the next hop address to the
1501 address of the router from which the datagram came. If this datagram
1502 is from the same router as the existing route, reinitialize the
1503 timeout. Next, compare the metrics. If the datagram is from the
1504 same router as the existing route, and the new metric is different
1505
1506
1507
1508 <span class="grey">Malkin Standards Track [Page 27]</span></pre>
1509 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-28" ></span>
1510 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1511
1512
1513 than the old one; or, if the new metric is lower than the old one; do
1514 the following actions:
1515
1516 - Adopt the route from the datagram (i.e., put the new metric in and
1517 adjust the next hop address, if necessary).
1518
1519 - Set the route change flag and signal the output process to trigger
1520 an update
1521
1522 - If the new metric is infinity, start the deletion process
1523 (described above); otherwise, re-initialize the timeout
1524
1525 If the new metric is infinity, the deletion process begins for the
1526 route, which is no longer used for routing packets. Note that the
1527 deletion process is started only when the metric is first set to
1528 infinity. If the metric was already infinity, then a new deletion
1529 process is not started.
1530
1531 If the new metric is the same as the old one, it is simplest to do
1532 nothing further (beyond re-initializing the timeout, as specified
1533 above); but, there is a heuristic which could be applied. Normally,
1534 it is senseless to replace a route if the new route has the same
1535 metric as the existing route; this would cause the route to bounce
1536 back and forth, which would generate an intolerable number of
1537 triggered updates. However, if the existing route is showing signs
1538 of timing out, it may be better to switch to an equally-good
1539 alternative route immediately, rather than waiting for the timeout to
1540 happen. Therefore, if the new metric is the same as the old one,
1541 examine the timeout for the existing route. If it is at least
1542 halfway to the expiration point, switch to the new route. This
1543 heuristic is optional, but highly recommended.
1544
1545 Any entry that fails these tests is ignored, as it is no better than
1546 the current route.
1547
1548 <span class="h3"><a class="selflink" id="section-3.10" href="#section-3.10">3.10</a> Output Processing</span>
1549
1550 This section describes the processing used to create response
1551 messages that contain all or part of the routing table. This
1552 processing may be triggered in any of the following ways:
1553
1554 - By input processing, when a Request is received (this Response is
1555 unicast to the requestor; see <a href="#section-3.7.1">section 3.7.1</a>)
1556
1557 - By the regular routing update (broadcast/multicast every 30
1558 seconds) router.
1559
1560 - By triggered updates (broadcast/multicast when a route changes)
1561
1562
1563
1564 <span class="grey">Malkin Standards Track [Page 28]</span></pre>
1565 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-29" ></span>
1566 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1567
1568
1569 When a Response is to be sent to all neighbors (i.e., a regular or
1570 triggered update), a Response message is directed to the router at
1571 the far end of each connected point-to-point link, and is broadcast
1572 (multicast for RIP-2) on all connected networks which support
1573 broadcasting. Thus, one Response is prepared for each directly-
1574 connected network, and sent to the appropriate address (direct or
1575 broadcast/multicast). In most cases, this reaches all neighboring
1576 routers. However, there are some cases where this may not be good
1577 enough. This may involve a network that is not a broadcast network
1578 (e.g., the ARPANET), or a situation involving dumb routers. In such
1579 cases, it may be necessary to specify an actual list of neighboring
1580 routers and send a datagram to each one explicitly. It is left to
1581 the implementor to determine whether such a mechanism is needed, and
1582 to define how the list is specified.
1583
1584 <span class="h4"><a class="selflink" id="section-3.10.1" href="#section-3.10.1">3.10.1</a> Triggered Updates</span>
1585
1586 Triggered updates require special handling for two reasons. First,
1587 experience shows that triggered updates can cause excessive load on
1588 networks with limited capacity or networks with many routers on them.
1589 Therefore, the protocol requires that implementors include provisions
1590 to limit the frequency of triggered updates. After a triggered
1591 update is sent, a timer should be set for a random interval between 1
1592 and 5 seconds. If other changes that would trigger updates occur
1593 before the timer expires, a single update is triggered when the timer
1594 expires. The timer is then reset to another random value between 1
1595 and 5 seconds. A triggered update should be suppressed if a regular
1596 update is due by the time the triggered update would be sent.
1597
1598 Second, triggered updates do not need to include the entire routing
1599 table. In principle, only those routes which have changed need to be
1600 included. Therefore, messages generated as part of a triggered
1601 update must include at least those routes that have their route
1602 change flag set. They may include additional routes, at the
1603 discretion of the implementor; however, sending complete routing
1604 updates is strongly discouraged. When a triggered update is
1605 processed, messages should be generated for every directly-connected
1606 network. Split Horizon processing is done when generating triggered
1607 updates as well as normal updates (see <a href="#section-3.9">section 3.9</a>). If, after Split
1608 Horizon processing for a given network, a changed route will appear
1609 unchanged on that network (e.g., it appears with an infinite metric),
1610 the route need not be sent. If no routes need be sent on that
1611 network, the update may be omitted. Once all of the triggered
1612 updates have been generated, the route change flags should be
1613 cleared.
1614
1615
1616
1617
1618
1619
1620 <span class="grey">Malkin Standards Track [Page 29]</span></pre>
1621 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-30" ></span>
1622 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1623
1624
1625 If input processing is allowed while output is being generated,
1626 appropriate interlocking must be done. The route change flags should
1627 not be changed as a result of processing input while a triggered
1628 update message is being generated.
1629
1630 The only difference between a triggered update and other update
1631 messages is the possible omission of routes that have not changed.
1632 The remaining mechanisms, described in the next section, must be
1633 applied to all updates.
1634
1635 <span class="h4"><a class="selflink" id="section-3.10.2" href="#section-3.10.2">3.10.2</a> Generating Response Messages</span>
1636
1637 This section describes how a Response message is generated for a
1638 particular directly-connected network:
1639
1640 Set the version number to either 1 or 2. The mechanism for deciding
1641 which version to send is implementation specific; however, if this is
1642 the Response to a Request, the Response version should match the
1643 Request version. Set the command to Response. Set the bytes labeled
1644 "must be zero" to zero. Start filling in RTEs. Recall that there is
1645 a limit of 25 RTEs to a Response; if there are more, send the current
1646 Response and start a new one. There is no defined limit to the
1647 number of datagrams which make up a Response.
1648
1649 To fill in the RTEs, examine each route in the routing table. If a
1650 triggered update is being generated, only entries whose route change
1651 flags are set need be included. If, after Split Horizon processing,
1652 the route should not be included, skip it. If the route is to be
1653 included, then the destination address and metric are put into the
1654 RTE. Routes must be included in the datagram even if their metrics
1655 are infinite.
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676 <span class="grey">Malkin Standards Track [Page 30]</span></pre>
1677 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-31" ></span>
1678 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1679
1680
1681 <span class="h2"><a class="selflink" id="section-4" href="#section-4">4</a>. Protocol Extensions</span>
1682
1683 This section does not change the RIP protocol per se. Rather, it
1684 provides extensions to the message format which allows routers to
1685 share important additional information.
1686
1687 The same header format is used for RIP-1 and RIP-2 messages (see
1688 <a href="#section-3.4">section 3.4</a>). The format for the 20-octet route entry (RTE) for
1689 RIP-2 is:
1690
1691 0 1 2 3 3
1692 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
1693 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1694 | Address Family Identifier (2) | Route Tag (2) |
1695 +-------------------------------+-------------------------------+
1696 | IP Address (4) |
1697 +---------------------------------------------------------------+
1698 | Subnet Mask (4) |
1699 +---------------------------------------------------------------+
1700 | Next Hop (4) |
1701 +---------------------------------------------------------------+
1702 | Metric (4) |
1703 +---------------------------------------------------------------+
1704
1705 The Address Family Identifier, IP Address, and Metric all have the
1706 meanings defined in <a href="#section-3.4">section 3.4</a>. The Version field will specify
1707 version number 2 for RIP messages which use authentication or carry
1708 information in any of the newly defined fields.
1709
1710 <span class="h3"><a class="selflink" id="section-4.1" href="#section-4.1">4.1</a> Authentication</span>
1711
1712 Since authentication is a per message function, and since there is
1713 only one 2-octet field available in the message header, and since any
1714 reasonable authentication scheme will require more than two octets,
1715 the authentication scheme for RIP version 2 will use the space of an
1716 entire RIP entry. If the Address Family Identifier of the first (and
1717 only the first) entry in the message is 0xFFFF, then the remainder of
1718 the entry contains the authentication. This means that there can be,
1719 at most, 24 RIP entries in the remainder of the message. If
1720 authentication is not in use, then no entries in the message should
1721 have an Address Family Identifier of 0xFFFF. A RIP message which
1722 contains an authentication entry would begin with the following
1723 format:
1724
1725
1726
1727
1728
1729
1730
1731
1732 <span class="grey">Malkin Standards Track [Page 31]</span></pre>
1733 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-32" ></span>
1734 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1735
1736
1737 0 1 2 3 3
1738 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
1739 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1740 | Command (1) | Version (1) | unused |
1741 +---------------+---------------+-------------------------------+
1742 | 0xFFFF | Authentication Type (2) |
1743 +-------------------------------+-------------------------------+
1744 ~ Authentication (16) ~
1745 +---------------------------------------------------------------+
1746
1747 Currently, the only Authentication Type is simple password and it is
1748 type 2. The remaining 16 octets contain the plain text password. If
1749 the password is under 16 octets, it must be left-justified and padded
1750 to the right with nulls (0x00).
1751
1752 <span class="h3"><a class="selflink" id="section-4.2" href="#section-4.2">4.2</a> Route Tag</span>
1753
1754 The Route Tag (RT) field is an attribute assigned to a route which
1755 must be preserved and readvertised with a route. The intended use of
1756 the Route Tag is to provide a method of separating "internal" RIP
1757 routes (routes for networks within the RIP routing domain) from
1758 "external" RIP routes, which may have been imported from an EGP or
1759 another IGP.
1760
1761 Routers supporting protocols other than RIP should be configurable to
1762 allow the Route Tag to be configured for routes imported from
1763 different sources. For example, routes imported from EGP or BGP
1764 should be able to have their Route Tag either set to an arbitrary
1765 value, or at least to the number of the Autonomous System from which
1766 the routes were learned.
1767
1768 Other uses of the Route Tag are valid, as long as all routers in the
1769 RIP domain use it consistently. This allows for the possibility of a
1770 BGP-RIP protocol interactions document, which would describe methods
1771 for synchronizing routing in a transit network.
1772
1773 <span class="h3"><a class="selflink" id="section-4.3" href="#section-4.3">4.3</a> Subnet mask</span>
1774
1775 The Subnet Mask field contains the subnet mask which is applied to
1776 the IP address to yield the non-host portion of the address. If this
1777 field is zero, then no subnet mask has been included for this entry.
1778
1779 On an interface where a RIP-1 router may hear and operate on the
1780 information in a RIP-2 routing entry the following rules apply:
1781
1782 1) information internal to one network must never be advertised into
1783 another network,
1784
1785
1786
1787
1788 <span class="grey">Malkin Standards Track [Page 32]</span></pre>
1789 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-33" ></span>
1790 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1791
1792
1793 2) information about a more specific subnet may not be advertised
1794 where RIP-1 routers would consider it a host route, and
1795
1796 3) supernet routes (routes with a netmask less specific than the
1797 "natural" network mask) must not be advertised where they could be
1798 misinterpreted by RIP-1 routers.
1799
1800 <span class="h3"><a class="selflink" id="section-4.4" href="#section-4.4">4.4</a> Next Hop</span>
1801
1802 The immediate next hop IP address to which packets to the destination
1803 specified by this route entry should be forwarded. Specifying a
1804 value of 0.0.0.0 in this field indicates that routing should be via
1805 the originator of the RIP advertisement. An address specified as a
1806 next hop must, per force, be directly reachable on the logical subnet
1807 over which the advertisement is made.
1808
1809 The purpose of the Next Hop field is to eliminate packets being
1810 routed through extra hops in the system. It is particularly useful
1811 when RIP is not being run on all of the routers on a network. A
1812 simple example is given in <a href="#appendix-A">Appendix A</a>. Note that Next Hop is an
1813 "advisory" field. That is, if the provided information is ignored, a
1814 possibly sub-optimal, but absolutely valid, route may be taken. If
1815 the received Next Hop is not directly reachable, it should be treated
1816 as 0.0.0.0.
1817
1818 <span class="h3"><a class="selflink" id="section-4.5" href="#section-4.5">4.5</a> Multicasting</span>
1819
1820 In order to reduce unnecessary load on those hosts which are not
1821 listening to RIP-2 messages, an IP multicast address will be used for
1822 periodic broadcasts. The IP multicast address is 224.0.0.9. Note
1823 that IGMP is not needed since these are inter-router messages which
1824 are not forwarded.
1825
1826 On NBMA networks, unicast addressing may be used. However, if a
1827 response addressed to the RIP-2 multicast address is received, it
1828 should be accepted.
1829
1830 In order to maintain backwards compatibility, the use of the
1831 multicast address will be configurable, as described in <a href="#section-5.1">section 5.1</a>.
1832 If multicasting is used, it should be used on all interfaces which
1833 support it.
1834
1835 <span class="h3"><a class="selflink" id="section-4.6" href="#section-4.6">4.6</a> Queries</span>
1836
1837 If a RIP-2 router receives a RIP-1 Request, it should respond with a
1838 RIP-1 Response. If the router is configured to send only RIP-2
1839 messages, it should not respond to a RIP-1 Request.
1840
1841
1842
1843
1844 <span class="grey">Malkin Standards Track [Page 33]</span></pre>
1845 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-34" ></span>
1846 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1847
1848
1849 <span class="h2"><a class="selflink" id="section-5" href="#section-5">5</a>. Compatibility</span>
1850
1851 RFC [<a href="#ref-1" title="&quot;Routing Information Protocol&quot;">1</a>] showed considerable forethought in its specification of the
1852 handling of version numbers. It specifies that RIP messages of
1853 version 0 are to be discarded, that RIP messages of version 1 are to
1854 be discarded if any Must Be Zero (MBZ) field is non-zero, and that
1855 RIP messages of any version greater than 1 should not be discarded
1856 simply because an MBZ field contains a value other than zero. This
1857 means that the new version of RIP is totally backwards compatible
1858 with existing RIP implementations which adhere to this part of the
1859 specification.
1860
1861 <span class="h3"><a class="selflink" id="section-5.1" href="#section-5.1">5.1</a> Compatibility Switch</span>
1862
1863 A compatibility switch is necessary for two reasons. First, there
1864 are implementations of RIP-1 in the field which do not follow RFC [<a href="#ref-1" title="&quot;Routing Information Protocol&quot;">1</a>]
1865 as described above. Second, the use of multicasting would prevent
1866 RIP-1 systems from receiving RIP-2 updates (which may be a desired
1867 feature in some cases). This switch should be configurable on a
1868 per-interface basis.
1869
1870 The switch has four settings: RIP-1, in which only RIP-1 messages are
1871 sent; RIP-1 compatibility, in which RIP-2 messages are broadcast;
1872 RIP-2, in which RIP-2 messages are multicast; and "none", which
1873 disables the sending of RIP messages. It is recommended that the
1874 default setting be either RIP-1 or RIP-2, but not RIP-1
1875 compatibility. This is because of the potential problems which can
1876 occur on some topologies. RIP-1 compatibility should only be used
1877 when all of the consequences of its use are well understood by the
1878 network administrator.
1879
1880 For completeness, routers should also implement a receive control
1881 switch which would determine whether to accept, RIP-1 only, RIP-2
1882 only, both, or none. It should also be configurable on a per-
1883 interface basis. It is recommended that the default be compatible
1884 with the default chosen for sending updates.
1885
1886 <span class="h3"><a class="selflink" id="section-5.2" href="#section-5.2">5.2</a> Authentication</span>
1887
1888 The following algorithm should be used to authenticate a RIP message.
1889 If the router is not configured to authenticate RIP-2 messages, then
1890 RIP-1 and unauthenticated RIP-2 messages will be accepted;
1891 authenticated RIP-2 messages shall be discarded. If the router is
1892 configured to authenticate RIP-2 messages, then RIP-1 messages and
1893 RIP-2 messages which pass authentication testing shall be accepted;
1894 unauthenticated and failed authentication RIP-2 messages shall be
1895 discarded. For maximum security, RIP-1 messages should be ignored
1896
1897
1898
1899
1900 <span class="grey">Malkin Standards Track [Page 34]</span></pre>
1901 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-35" ></span>
1902 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1903
1904
1905 when authentication is in use (see <a href="#section-4.1">section 4.1</a>); otherwise, the
1906 routing information from authenticated messages will be propagated by
1907 RIP-1 routers in an unauthenticated manner.
1908
1909 Since an authentication entry is marked with an Address Family
1910 Identifier of 0xFFFF, a RIP-1 system would ignore this entry since it
1911 would belong to an address family other than IP. It should be noted,
1912 therefore, that use of authentication will not prevent RIP-1 systems
1913 from seeing RIP-2 messages. If desired, this may be done using
1914 multicasting, as described in sections <a href="#section-4.5">4.5</a> and <a href="#section-5.1">5.1</a>.
1915
1916 <span class="h3"><a class="selflink" id="section-5.3" href="#section-5.3">5.3</a> Larger Infinity</span>
1917
1918 While on the subject of compatibility, there is one item which people
1919 have requested: increasing infinity. The primary reason that this
1920 cannot be done is that it would violate backwards compatibility. A
1921 larger infinity would obviously confuse older versions of rip. At
1922 best, they would ignore the route as they would ignore a metric of
1923 16. There was also a proposal to make the Metric a single octet and
1924 reuse the high three octets, but this would break any implementations
1925 which treat the metric as a 4-octet entity.
1926
1927 <span class="h3"><a class="selflink" id="section-5.4" href="#section-5.4">5.4</a> Addressless Links</span>
1928
1929 As in RIP-1, addressless links will not be supported by RIP-2.
1930
1931 <span class="h2"><a class="selflink" id="section-6" href="#section-6">6</a>. Interaction between version 1 and version 2</span>
1932
1933 Because version 1 packets do not contain subnet information, the
1934 semantics employed by routers on networks that contain both version 1
1935 and version 2 networks should be limited to that of version 1.
1936 Otherwise it is possible either to create blackhole routes (i.e.,
1937 routes for networks that do not exist) or to create excessive routing
1938 information in a version 1 environment.
1939
1940 Some implementations attempt to automatically summarize groups of
1941 adjacent routes into single entries, the goal being to reduce the
1942 total number of entries. This is called auto-summarization.
1943
1944 Specifically, when using both version 1 and version 2 within a
1945 network, a single subnet mask should be used throughout the network.
1946 In addition, auto-summarization mechanisms should be disabled for
1947 such networks, and implementations must provide mechanisms to disable
1948 auto-summarization.
1949
1950
1951
1952
1953
1954
1955
1956 <span class="grey">Malkin Standards Track [Page 35]</span></pre>
1957 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-36" ></span>
1958 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
1959
1960
1961 <span class="h2"><a class="selflink" id="section-7" href="#section-7">7</a>. Security Considerations</span>
1962
1963 The basic RIP protocol is not a secure protocol. To bring RIP-2 in
1964 line with more modern routing protocols, an extensible authentication
1965 mechanism has been incorporated into the protocol enhancements. This
1966 mechanism is described in sections <a href="#section-4.1">4.1</a> and <a href="#section-5.2">5.2</a>. Security is further
1967 enhanced by the mechanism described in [<a href="#ref-3" title="&quot;RIP-II MD5 Authentication&quot;">3</a>].
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012 <span class="grey">Malkin Standards Track [Page 36]</span></pre>
2013 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-37" ></span>
2014 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
2015
2016
2017 Appendix A
2018
2019 This is a simple example of the use of the next hop field in a rip
2020 entry.
2021
2022 ----- ----- ----- ----- ----- -----
2023 |IR1| |IR2| |IR3| |XR1| |XR2| |XR3|
2024 --+-- --+-- --+-- --+-- --+-- --+--
2025 | | | | | |
2026 --+-------+-------+---------------+-------+-------+--
2027 &lt;-------------RIP-2-------------&gt;
2028
2029 Assume that IR1, IR2, and IR3 are all "internal" routers which are
2030 under one administration (e.g. a campus) which has elected to use
2031 RIP-2 as its IGP. XR1, XR2, and XR3, on the other hand, are under
2032 separate administration (e.g. a regional network, of which the campus
2033 is a member) and are using some other routing protocol (e.g. OSPF).
2034 XR1, XR2, and XR3 exchange routing information among themselves such
2035 that they know that the best routes to networks N1 and N2 are via
2036 XR1, to N3, N4, and N5 are via XR2, and to N6 and N7 are via XR3. By
2037 setting the Next Hop field correctly (to XR2 for N3/N4/N5, to XR3 for
2038 N6/N7), only XR1 need exchange RIP-2 routes with IR1/IR2/IR3 for
2039 routing to occur without additional hops through XR1. Without the
2040 Next Hop (for example, if RIP-1 were used) it would be necessary for
2041 XR2 and XR3 to also participate in the RIP-2 protocol to eliminate
2042 extra hops.
2043
2044 References
2045
2046 [<a id="ref-1">1</a>] Hedrick, C., "Routing Information Protocol", STD 34, <a href="./rfc1058">RFC 1058</a>,
2047 Rutgers University, June 1988.
2048
2049 [<a id="ref-2">2</a>] Malkin, G., and F. Baker, "RIP Version 2 MIB Extension", <a href="./rfc1389">RFC</a>
2050 <a href="./rfc1389">1389</a>, January 1993.
2051
2052 [<a id="ref-3">3</a>] Baker, F., and R. Atkinson, "RIP-II MD5 Authentication", <a href="./rfc2082">RFC</a>
2053 <a href="./rfc2082">2082</a>, January 1997.
2054
2055 [<a id="ref-4">4</a>] Bellman, R. E., "Dynamic Programming", Princeton University
2056 Press, Princeton, N.J., 1957.
2057
2058 [<a id="ref-5">5</a>] Bertsekas, D. P., and Gallaher, R. G., "Data Networks",
2059 Prentice-Hall, Englewood Cliffs, N.J., 1987.
2060
2061 [<a id="ref-6">6</a>] Braden, R., and Postel, J., "Requirements for Internet Gateways",
2062 STD 4, <a href="./rfc1009">RFC 1009</a>, June 1987.
2063
2064
2065
2066
2067
2068 <span class="grey">Malkin Standards Track [Page 37]</span></pre>
2069 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-38" ></span>
2070 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
2071
2072
2073 [<a id="ref-7">7</a>] Boggs, D. R., Shoch, J. F., Taft, E. A., and Metcalfe, R. M.,
2074 "Pup: An Internetwork Architecture", IEEE Transactions on
2075 Communications, April 1980.
2076
2077 [<a id="ref-8">8</a>] Ford, L. R. Jr., and Fulkerson, D. R., "Flows in Networks",
2078 Princeton University Press, Princeton, N.J., 1962.
2079
2080 [<a id="ref-9">9</a>] Xerox Corp., "Internet Transport Protocols", Xerox System
2081 Integration Standard XSIS 028112, December 1981.
2082
2083 [<a id="ref-10">10</a>] Floyd, S., and V. Jacobson, "The synchronization of Periodic
2084 Routing Messages," ACM Sigcom '93 symposium, September 1993.
2085
2086 [<a id="ref-11">11</a>] Baker, F., "Requirements for IP Version 4 Routers." <a href="./rfc1812">RFC 1812</a>,
2087 June 1995.
2088
2089 Author's Address
2090
2091 Gary Scott Malkin
2092 Bay Networks
2093 8 Federal Street
2094 Billerica, MA 01821
2095
2096 Phone: (978) 916-4237
2097 EMail: [email protected]
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124 <span class="grey">Malkin Standards Track [Page 38]</span></pre>
2125 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-39" ></span>
2126 <span class="grey"><a href="./rfc2453">RFC 2453</a> RIP Version 2 November 1998</span>
2127
2128
2129 Full Copyright Statement
2130
2131 Copyright (C) The Internet Society (1998). All Rights Reserved.
2132
2133 This document and translations of it may be copied and furnished to
2134 others, and derivative works that comment on or otherwise explain it
2135 or assist in its implementation may be prepared, copied, published
2136 and distributed, in whole or in part, without restriction of any
2137 kind, provided that the above copyright notice and this paragraph are
2138 included on all such copies and derivative works. However, this
2139 document itself may not be modified in any way, such as by removing
2140 the copyright notice or references to the Internet Society or other
2141 Internet organizations, except as needed for the purpose of
2142 developing Internet standards in which case the procedures for
2143 copyrights defined in the Internet Standards process must be
2144 followed, or as required to translate it into languages other than
2145 English.
2146
2147 The limited permissions granted above are perpetual and will not be
2148 revoked by the Internet Society or its successors or assigns.
2149
2150 This document and the information contained herein is provided on an
2151 "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
2152 TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
2153 BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
2154 HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
2155 MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180 Malkin Standards Track [Page 39]
2181 </pre>