Designing IP Addressing and Selecting Routing Protocols Part 2 — Transcript

Full transcript

1. 0:04reviewing enterprise
2. 0:06routing protocols
3. 0:10this segment describes considerations
4. 0:12for selecting the most appropriate
5. 0:14network routing protocol
6. 0:16so first routing protocol features are
7. 0:19discussed
8. 0:20followed by a description of various
9. 0:22routing protocols appropriate for the
10. 0:24enterprise use
11. 0:26the video discusses why certain
12. 0:28protocols are suitable
13. 0:30for specific modules in the enterprise
14. 0:32architecture
15. 0:33it concludes with a description of some
16. 0:36advanced
17. 0:37routing protocol deployment features
18. 0:39including redistribution
19. 0:41filtering and summarization
20. 0:48routing protocol features so you've got
21. 0:50the distance vector
22. 0:52versus the link state okay so there are
23. 0:56many ways to characterize
24. 0:57routing protocols including static
25. 1:01versus dynamic routing
26. 1:03interior versus exterior routing
27. 1:05protocols
28. 1:06you've got distance vector versus link
29. 1:09state
30. 1:10versus hybrid protocols okay
31. 1:13so you also have routing protocol
32. 1:15metrics
33. 1:17routing protocol convergence flat versus
34. 1:20hierarchical routing protocols
35. 1:23so basically on this slide
36. 1:26we will be talking about distance vector
37. 1:28and link state
38. 1:29comparison okay so there are protocols
39. 1:33that are classified to be distance
40. 1:34vector
41. 1:35which includes and eigrp
42. 1:40okay so this our distance vector
43. 1:43eigrp is considered to be the hybrid
44. 1:45protocol
45. 1:47okay now link state
46. 1:50such as ospf and isis these are types of
47. 1:54the link state routing protocols okay
48. 1:57now distance vector protocol
49. 2:00characteristics
50. 2:02well slow convergence easy
51. 2:05implementation and maintenance
52. 2:07and limited scalability
53. 2:10okay whereas for the link state protocol
54. 2:14it is fast convergence good scalability
55. 2:19less routing traffic overhead and more
56. 2:22knowledge
57. 2:22needed for implementation maintenance
58. 2:25again several protocols are running
59. 2:28under distance vector which includes
60. 2:30rip and eigrp and for the link state
61. 2:34well a popular ospf is being considered
62. 2:40now let's focus on the distance vector
63. 2:42protocol
64. 2:44in distance vector protocol routing
65. 2:46decisions
66. 2:47are made on a half by hub basis
67. 2:51okay each router relies on its neighbor
68. 2:54routers
69. 2:55to make the correct routing decision so
70. 2:58the router passes
71. 3:00only the results of this decision
72. 3:03which is its routing table to its
73. 3:05neighbors
74. 3:07distance vector protocols are typically
75. 3:09slower to converge
76. 3:10and do not scale well however they are
77. 3:13easy to implement and maintain
78. 3:16so again examples of distance of vector
79. 3:18protocols includes
80. 3:20rip okay both rip version one and
81. 3:22represent two
82. 3:24and the enhanced interior gateway
83. 3:26routing protocol or the eigrp now for
84. 3:32link state protocol
85. 3:34each router floods information about
86. 3:37itself
87. 3:38its link states either to all other
88. 3:41routers
89. 3:42in the network okay or
90. 3:45to part of the network we call it area
91. 3:49now each router makes its routing
92. 3:53decision based on all the received
93. 3:57information
94. 3:58and using the short test path first or
95. 4:01spf algorithm
96. 4:02also known as the dextras algo which
97. 4:05calculates the shortest path
98. 4:07to any destination the link state
99. 4:10protocols are fast to converge
100. 4:13have less routing traffic overhead and
101. 4:16scale well
102. 4:18however because of their complexity
103. 4:21the link state protocols are more
104. 4:23difficult to implement and maintain
105. 4:26the ip-link state protocols are ospf
106. 4:29and the integrated isis
107. 4:32so the third type of protocol that i
108. 4:34have mentioned earlier
109. 4:36is what you call hybrid interior gateway
110. 4:38protocol
111. 4:39which is the cisco's eigrp
112. 4:42now i have mentioned earlier that this
113. 4:44is part of the distance
114. 4:46vector okay now it is an advanced
115. 4:49distance vector
116. 4:50making it a hybrid okay so eigrp has
117. 4:54characteristics of both
118. 4:56distance vector and link state protocols
119. 4:59it combines distance vector behavior
120. 5:02with some link state characteristics and
121. 5:05some proprietary features
122. 5:08so eigrp routing protocols are running
123. 5:11only on cisco devices
124. 5:13so if you are using non-cisco devices so
125. 5:16it is recommended to use
126. 5:18ospf so eigrp
127. 5:21is a fast converging and scalable
128. 5:23routing protocol
129. 5:28now let's look at the interior versus
130. 5:30exterior
131. 5:32routing protocols so interior
132. 5:36okay or igps routing inside the
133. 5:39autonomous system
134. 5:41again autonomous system pertains to a
135. 5:43network
136. 5:44being maintained by a single
137. 5:47administrator so fast convergence and
138. 5:50easy configuration
139. 5:52low administrator influence on routing
140. 5:55decisions
141. 5:56now for exterior gateway protocols or
142. 5:59egps
143. 6:00routing between autonomous systems slow
144. 6:03convergence
145. 6:04and more complex configuration and high
146. 6:07administrator influence
147. 6:08on routing decisions so if you want to
148. 6:11learn more about
149. 6:12interior and exterior routing product
150. 6:14also go ahead and check the
151. 6:16supplementary videos
152. 6:20okay now where do we apply interior and
153. 6:23exterior
154. 6:24routing protocols so as what i've
155. 6:26mentioned earlier
156. 6:27okay interior or igps
157. 6:31are the routing protocols implemented
158. 6:34within the autonomous system
159. 6:36so if this is an enterprise a within
160. 6:38this network
161. 6:40we are using interior gateway protocol
162. 6:42example
163. 6:43eigrp okay assuming that this is
164. 6:47autonomous system two if you have
165. 6:50autonomous system three again these are
166. 6:52or this is maintained by a different
167. 6:56single administrator and within this
168. 6:58network
169. 6:59autonomous system three we are running
170. 7:02igp specifically ospf
171. 7:06now on ispc or
172. 7:09as1 okay so within the system
173. 7:13they are using igp isis
174. 7:17but connecting this different autonomous
175. 7:20systems
176. 7:21that is where we use the exterior
177. 7:24gateway routing protocol or egp
178. 7:26like the border gateway protocol or bgb
179. 7:32okay so interior versus exterior routing
180. 7:36protocols so
181. 7:36flat routing protocols propagate all
182. 7:39routing information
183. 7:40throughout the network so this
184. 7:44are to be classful routing protocols
185. 7:47okay so these are the old routing
186. 7:50protocols which
187. 7:51follows the subnet mask
188. 7:54okay this is not appropriate for large
189. 7:57networks
190. 7:58so example of routing protocols which
191. 8:01are class full and flat
192. 8:02are rip version one okay the old
193. 8:05igrp which now we are using eigrp
194. 8:10and rip version 2 which is classless
195. 8:14okay now for hierarchical routing
196. 8:18protocols
197. 8:18divide large networks into smaller areas
198. 8:22or subnets so these are said to be
199. 8:24classless routing protocols
200. 8:26and usually we are using eigrp
201. 8:30okay again eigrp is a proprietary
202. 8:33protocol from cisco
203. 8:34if you are using non-cisco devices so
204. 8:37better use ospf
205. 8:39okay or isis okay
206. 8:42so also for the hierarchical routing
207. 8:44protocols
208. 8:46so limited drought propagation between
209. 8:48areas
210. 8:51all right so flat versus hierarchical
211. 8:55routing protocols so again we mentioned
212. 8:59about
213. 8:59flat routing protocol earlier so flat
214. 9:02routing protocols have no means of
215. 9:03limiting route propagation
216. 9:05in major network that's within class a
217. 9:09class b or c network environment
218. 9:12these protocols are typically classful
219. 9:15distance
220. 9:15vector protocol so the key point
221. 9:18flat routing protocols propagate all
222. 9:21routing information
223. 9:23throughout the network whereas
224. 9:24hierarchical
225. 9:26routing protocols divide large networks
226. 9:29into smaller areas okay
227. 9:32so two examples of flat routing
228. 9:35protocols are
229. 9:36again reversion 1 and rip version 2.
230. 9:40take note however that rip version 2
231. 9:43is already a classless protocol
232. 9:46now this figure here illustrates a flat
233. 9:49network
234. 9:50and hierarchical network
235. 9:53now for the hierarchical routing
236. 9:54protocols to solve the problems
237. 9:57associated with flat
238. 9:58routing protocols additional features
239. 10:01are implemented in hierarchical routing
240. 10:03protocols to support
241. 10:04large networks for example some support
242. 10:08an area based design so hierarchical
243. 10:12routing protocols are typically classed
244. 10:14as
245. 10:14link state protocols so classless
246. 10:18means that routing updates include
247. 10:20subnet masks
248. 10:21in their routing updates therefore the
249. 10:24routing protocol supports
250. 10:26vlsm or the variable length subnet mask
251. 10:32routing protocol convergence
252. 10:36so different routing protocols need
253. 10:38different amounts
254. 10:39of time to converge in a given network
255. 10:43so although the convergence depends on
256. 10:45networks
257. 10:46topology and structure pure distance of
258. 10:49vector protocols are slower to converge
259. 10:53than link state protocols the use of
260. 10:56periodic updates
261. 10:58and the hold down mechanism are the main
262. 11:01reasons
263. 11:03for slow convergence as a result
264. 11:06the fast converging protocols should be
265. 11:09used
266. 11:10when the network's convergence time is
267. 11:13crucial
268. 11:18now as shown here in the figure
269. 11:20different routing protocols
270. 11:22need different amounts of time to
271. 11:23converge in a given network
272. 11:26so link state protocols usually converge
273. 11:29much
274. 11:30more quickly because they instantly
275. 11:32propagate
276. 11:33routing updates okay so whenever the
277. 11:37change
278. 11:37occurs in a link state okay that's ospf
279. 11:41a link state update is flooded
280. 11:44throughout the network or within the
281. 11:45entire network
282. 11:47so there is no need to wait for the
283. 11:49whole down timer to expire
284. 11:52or for the next periodic update as with
285. 11:54distance
286. 11:55vector protocols
287. 12:00now let's talk about this hybrid eigrp
288. 12:04or the enhanced interior gateway routing
289. 12:06protocol from cisco
290. 12:08so eigrp is a special case because
291. 12:12it incorporates the distance vector
292. 12:14principle of metric propagation
293. 12:16it means it sends only the best routes
294. 12:19to the neighbors
295. 12:21however it does not have periodic
296. 12:24updates
297. 12:25nor does it implement the principle of
298. 12:28hold downs
299. 12:29so the most distinct feature of eigrp is
300. 12:33that
301. 12:34it stores all feasible backup routes in
302. 12:37its topology
303. 12:38table so when a backup route exists
304. 12:41for a lost destination the switch
305. 12:45over to the best backup route
306. 12:49is almost immediate and involves
307. 12:52no action from other routers therefore
308. 12:56very fast convergence can be achieved
309. 12:59with proper
310. 13:00eigrp deployment okay
311. 13:05eigrp characteristics so as i mentioned
312. 13:08earlier
313. 13:09eigrp is proprietary okay it's a cisco
314. 13:12proprietary protocol for routing ipb4
315. 13:16eigrp can also be configured for routing
316. 13:19ipv6 okay internetwork packet exchange
317. 13:23ipx and apple talk traffic
318. 13:27now eigrp is an enhanced version
319. 13:30of the old igrp which is pure
320. 13:34distance vector protocol so eigrp
321. 13:37however
322. 13:38is a hybrid routing protocol
323. 13:41it is a distance vector protocol with
324. 13:44additional link state
325. 13:46protocol features so eigrp features
326. 13:49includes the following
327. 13:51so it uses a triggered updates
328. 13:54eigrp take note has no periodic updates
329. 13:58uses a topology table to keep all routes
330. 14:02received from its neighbors not only the
331. 14:05best routes
332. 14:07establishes adjacencies with neighboring
333. 14:09routers using the hello protocols
334. 14:12just like in ospf okay uses multicast
335. 14:16rather than broadcast for communication
336. 14:19it supports also vlsm
337. 14:22supports manual route summarization
338. 14:25eigrp summarizes on major network
339. 14:28boundaries by default
340. 14:30okay but this features can be turned off
341. 14:34and summarization can be configured at
342. 14:36any point in the network
343. 14:39so it can be used to create
344. 14:41hierarchically structured
345. 14:42large networks and supports
346. 14:45an equal load balancing
347. 14:49okay so eigrp
348. 14:52uses the diffusing update algorithm
349. 14:56to determine the best path ospf uses
350. 14:59dixtras algorithm
351. 15:04now let's talk about the open shortest
352. 15:06path
353. 15:07first routing protocol or the ospf
354. 15:10ospf is a standardized protocol
355. 15:13for routing ipb4 okay so
356. 15:16it was developed uh in the year 1988
357. 15:21by the ng uh internet engineering task
358. 15:23force right
359. 15:24etf to replace rip in larger
360. 15:28more diverse media networks
361. 15:31okay so in 1998 minor changes
362. 15:35in ospf version 2 addressed some of the
363. 15:38ospf version 1's problem while
364. 15:40maintaining backward
365. 15:42compatibility so ospf
366. 15:45was developed for use in large scalable
367. 15:48networks
368. 15:49in which rips inherent limitations
369. 15:52failed to satisfy requirements
370. 15:55so ospf is a superior to reap
371. 15:59in all aspects including the following
372. 16:02okay so you've got it converges
373. 16:05much faster it supports
374. 16:08vlsm manual summarization and
375. 16:11hierarchical structures
376. 16:13it has improved metric calculation
377. 16:16for best path okay and it does not have
378. 16:21a half count limitations on rip version
379. 16:24one
380. 16:25and two so at its inception
381. 16:28ospf support or supported the largest
382. 16:34networks
383. 16:36ospf hierarchical design so although
384. 16:39ospf
385. 16:40was developed for large networks each
386. 16:43implementation requires
387. 16:45proper design and planning this is
388. 16:47especially important for networks
389. 16:49with 50 or more routers
390. 16:52okay so the concept of multiple separate
391. 16:55areas inside one domain or aes
392. 16:57was implemented in ospf to reduce the
393. 17:00amount of routing traffic
394. 17:02and make networks more scalable
395. 17:05so in ospf there must always be
396. 17:09one backbone area which is
397. 17:12area 0 to which all
398. 17:15other non-backbone areas must be
399. 17:18directly connected
400. 17:19so you've got area one here area two
401. 17:22this
402. 17:23non-backbone area should be directly
403. 17:25connected
404. 17:26to area zero or backbone via
405. 17:30the area border router or abr
406. 17:34okay now a router is a member of an ospf
407. 17:38area when at least one of each
408. 17:40interfaces
409. 17:42operates in that area so router that
410. 17:45resides
411. 17:46on boundaries between the backbone and
412. 17:49the non-backbone area are called
413. 17:51area border routers or abrs and have at
414. 17:54least
415. 17:55one of each interface connected to each
416. 17:58area
417. 17:59so the boundary between the areas is
418. 18:02within the abr itself
419. 18:04so if external routes are propagated
420. 18:07into ospf aes
421. 18:10the router does redistributes
422. 18:13the routes and it is called autonomous
423. 18:15system border routers or asbrs
424. 18:18okay so careful design
425. 18:21and correct mapping of areas to the
426. 18:24network topology are important because
427. 18:26manual summarization of routes can only
428. 18:29be performed
429. 18:30on abrs or asbrs
430. 18:34again when you say asbrs these are
431. 18:37routers
432. 18:38connected to non-ospf networks example
433. 18:42the isps
434. 18:44so traffic sent from one non-backbone
435. 18:48area to another always crosses the
436. 18:50backbone
437. 18:51for example the area one 1
438. 18:55abr must forward traffic from area 1
439. 18:59distant for area 2 into the backbone
440. 19:04the area 2 abr receives the traffic from
441. 19:07the backbone
442. 19:08and forwards it to the appropriate
443. 19:10destination
444. 19:11inside area 2.
445. 19:16ospf characteristics ospf
446. 19:20is a link state protocol that has the
447. 19:22following characteristics for deployment
448. 19:25in the enterprise networks so first fast
449. 19:28convergence
450. 19:30okay so ospf achieves fast convergence
451. 19:34times
452. 19:34using triggered link state updates that
453. 19:37include
454. 19:38one or more link state advertisements or
455. 19:40lsas
456. 19:42so lsas describe the state of the link
457. 19:45on specific routers and are propagated
458. 19:49and changed within an area so therefore
459. 19:53all routers in the same area have
460. 19:56identical
461. 19:57topology tables each router has a
462. 20:00complete
463. 20:01view of all the links and devices in the
464. 20:04area
465. 20:05so depending on their type lsas are
466. 20:08usually
467. 20:08changed by abrs when they cross
468. 20:12into another area
469. 20:15so the next one is it's a very good
470. 20:18scalability okay ospf
471. 20:22multiple area structure provides good
472. 20:25scalability however
473. 20:27ospf's strict area implementation rules
474. 20:31require proper design to support
475. 20:34other scalability features such as
476. 20:36manual summarization
477. 20:38okay on abrs and asbr's
478. 20:42stop areas totally stubby areas
479. 20:45and not so stubby areas or nssa
480. 20:48okay next is reduced bandwidth
481. 20:52usage along with the area structure
482. 20:56the use of triggered not periodic
483. 20:59okay so we're talking about the
484. 21:00triggered updates and
485. 21:02manual summarization reduces the
486. 21:05bandwidth used
487. 21:06by ospf by limiting the volume of link
488. 21:09state
489. 21:10propagation okay
490. 21:13and the last one would be of course the
491. 21:15vlsm support
492. 21:17so because ospf is a classless routing
493. 21:19protocol it supports
494. 21:20vlsm to achieve better use of ip address
495. 21:23space
496. 21:26so to know more about ospf you can go
497. 21:28ahead and check the
498. 21:30supplementary video okay so that
499. 21:32discusses
500. 21:34the details about this routing protocol
501. 21:38next would be the isis
502. 21:42so isis was developed by digital
503. 21:45equipment corporation or dec
504. 21:47as a dynamic link state routing protocol
505. 21:49for the open systems interconnection
506. 21:52osi protocol suite okay
507. 21:55so the osi suite uses connectionless
508. 21:58network systems or service
509. 22:00so we call it clns to provide
510. 22:03connectionless delivery of data
511. 22:05and the actual layer 3 protocol is
512. 22:07connectionless network protocol or clnp
513. 22:11now clnp is the osi suite solution
514. 22:14for connection less delivery of data
515. 22:17similar to ip and tcp ip suite
516. 22:20so isis uses clns address
517. 22:24to identify the routers and build the
518. 22:27link state databases or lsdb
519. 22:33next would be the border gateway
520. 22:34protocol this is an example of egp
521. 22:38okay so bgp is an egp that is primarily
522. 22:42used to interconnect autonomous systems
523. 22:45bgp is a successor to egp
524. 22:49the exterior gateway protocol okay so
525. 22:52note
526. 22:52the dual use of egp acronym here
527. 22:55okay so because egp is obsolete
528. 23:00bgp is currently the only egp in use
529. 23:04all right so the key point here for bgp
530. 23:06is that
531. 23:07the main goal of bgp is to provide an
532. 23:11inter-domain routing system
533. 23:13that guarantees a loop free exchange of
534. 23:16routing information
535. 23:18between autonomous systems so bgp
536. 23:21routers exchange information about path
537. 23:24to destination networks
538. 23:28bgp network implementation so in this
539. 23:32figure here
540. 23:33okay bgp is used to interconnect
541. 23:37multiple autonomous systems because of
542. 23:40the multiple connection between
543. 23:41autonomous systems
544. 23:43and the need for path manipulation
545. 23:46the use of static routing is excluded
546. 23:51so aes 65000 is multi-homed
547. 23:55to three isps okay as65500
548. 24:01okay as65000
549. 24:05and as64500
550. 24:12internal bgp so when bgp is running
551. 24:15between routers
552. 24:16within 1as it is called internal
553. 24:20bgp or ibgp ibgp
554. 24:23is run within an aes to exchange bgp
555. 24:27information
556. 24:28so that all internal bgp speakers have
557. 24:31the same bgp routing information
558. 24:33about outside autonomous systems and so
559. 24:37that information can be passed to other
560. 24:40autonomous systems
561. 24:42as long as they can reach each other
562. 24:45routers that run
563. 24:46ibgp do not have to be directly
564. 24:48connected to each other
565. 24:51static routes or routes learned from an
566. 24:54igp
567. 24:54running within the aes provides rich
568. 24:57ability
569. 24:59when bgp runs between routers in
570. 25:01different autonomous systems
571. 25:03it is called external bgp or ebgp
572. 25:08routers that run ebgp are usually
573. 25:11connected directly to each other
574. 25:17recommend that enterprise routing
575. 25:19protocol comparison
576. 25:21okay so basically this would yield to
577. 25:25two of the most popular
578. 25:28interior gateway routing protocol eig rp
579. 25:32and ospf
580. 25:34so they are both fast convergence
581. 25:38they are very good uh scalability okay
582. 25:43this two both uses vlsm
583. 25:47okay multiple network layer protocol
584. 25:50support
585. 25:51well eigrp yes okay ospf
586. 25:55no mix vendor devices
587. 26:00pro spf yes you can use
588. 26:03ospf routing protocol on
589. 26:06any brand of routers
590. 26:09unlike eigrp which is supported only on
591. 26:12cisco
592. 26:12equipments this is proprietary here
593. 26:18now to summarize reviewing the
594. 26:20enterprise routing protocols
595. 26:23so protocols with hierarchical and link
596. 26:25state attributes supports the fastest
597. 26:27network convergence
598. 26:29eigrp and ospf are the recommended igps
599. 26:33for the enterprise
600. 26:34eigrp is a cisco proprietary for routing
601. 26:37ipv4
602. 26:38ipv6 ipx and apple talk traffic
603. 26:42ospf is a standardized protocol for
604. 26:45routing ipv4
605. 26:47developed to replace rip in a larger
606. 26:49scale
607. 26:50more diverse media networks it can also
608. 26:53support ipv6
609. 26:55bgp is a representative
610. 26:59egp it is primarily used to interconnect
611. 27:03autonomous systems
612. 27:04or to connect enterprises to an isp