Network Monitoring — Transcript

Full transcript

1. 0:04hello there
2. 0:05good day on this video lecture
3. 0:09i'm going to discuss network monitoring
4. 0:15okay so for
5. 0:16the topic outline so on this video i'll
6. 0:19be covering the introduction about
7. 0:21network monitoring the monitor
8. 0:25types of information
9. 0:28network monitoring configurations
10. 0:31network monitoring methods performance
11. 0:35monitoring
12. 0:36this includes performance indicators
13. 0:39performance monitoring functions
14. 0:42fault monitoring problems
15. 0:45of fault monitoring and fault monitoring
16. 0:49functions
17. 0:50and the accounting monitoring so this is
18. 0:54part
19. 0:54and continuation of the f-cups that we
20. 0:57have
21. 0:58discussed on the previous video on the
22. 1:00network management overview
23. 1:03so let's start so
24. 1:07network monitoring is basically
25. 1:09concerned
26. 1:10with observing and
27. 1:14analyzing the status and behavior of the
28. 1:16end systems
29. 1:18so when you say and systems this
30. 1:20pertains to computers
31. 1:21servers printers right
32. 1:25um voip these are the components that
33. 1:28can be found on the end segment of the
34. 1:30network
35. 1:31okay so also intermediate systems
36. 1:35so intermediate systems pertains to
37. 1:38the devices that is located are situated
38. 1:40between
39. 1:41the end systems or end devices so this
40. 1:44includes
41. 1:45uh routers switches firewalls
42. 1:48okay so those are intermediate systems
43. 1:51layer two or layer three switches
44. 1:54okay and sub networks so pertaining to
45. 1:57the vlans
46. 1:58that make up the network to be managed
47. 2:01okay so what are the issues in
48. 2:05network monitoring okay so first
49. 2:08is what to monitor so a systems
50. 2:11administrator
51. 2:13so one of our duties and responsibility
52. 2:16of course is to monitor the network
53. 2:20okay but what do we have to monitor in
54. 2:23the network
55. 2:24so define what is to be monitored okay
56. 2:28so next would be how do we how do we
57. 2:31monitor it
58. 2:32how to obtain information from the
59. 2:34managed resources
60. 2:36when you say managed resources this
61. 2:38pertains to the managed objects
62. 2:40okay and this managed objects are
63. 2:43the ant systems intermediate systems and
64. 2:45the subnets
65. 2:47okay and
66. 2:51what to do with the monitor information
67. 2:52so how the monitor information is used
68. 2:55in various management functional levels
69. 2:58so basically
70. 2:59on what to monitor being a systems
71. 3:01administrator so what do we want to
72. 3:03monitor on the network
73. 3:04so basically we want to monitor the
74. 3:06performance of the network
75. 3:08the behavior of the network right the
76. 3:12activities on the network okay
77. 3:16so we are not monitoring
78. 3:19the persons or the personal involved
79. 3:22within the organizations but their
80. 3:23activities over the infrastructure
81. 3:26okay and how do we monitor it so
82. 3:29basically we are using a third party
83. 3:30software
84. 3:31okay so we have a lot of applications
85. 3:33that we can use
86. 3:35for monitoring monitoring the network
87. 3:36monitoring the
88. 3:38bandwidth monitoring the throughput so
89. 3:41basically in here
90. 3:43we are using mostly application software
91. 3:45or programs
92. 3:47okay and what do we do with the monitor
93. 3:50information so after gathering all those
94. 3:52information
95. 3:53after the observation and analysis so
96. 3:57what do we do with this information okay
97. 4:00so basically this information can be
98. 4:02used
99. 4:04for future planning okay so maybe
100. 4:08at the current situation you are
101. 4:09encountering some difficulties or
102. 4:11problems regarding your network
103. 4:13and the result of your observation and
104. 4:15analysis
105. 4:16might help you improve the performance
106. 4:18of the network in the future
107. 4:21okay so basically what to do with the
108. 4:23monitor information is
109. 4:24that would help you the team and
110. 4:27organization
111. 4:28to improve the service provided on the
112. 4:31infrastructure
113. 4:33okay next
114. 4:36monitor types of information okay so
115. 4:39this talks about
116. 4:40the static information so when you say
117. 4:43static information
118. 4:44these are the information that could be
119. 4:46hardly changes
120. 4:47okay so current configuration
121. 4:50information so example if we're talking
122. 4:52about the router
123. 4:53so the number and identification of
124. 4:55ports on the router is considered to be
125. 4:56static information
126. 4:58okay so the number of lan or the
127. 5:01internet interfaces on the router
128. 5:04is considered to be static information
129. 5:07also if we're talking about these
130. 5:09switches the number of ports available
131. 5:11on the switch
132. 5:12is considered to be static information
133. 5:15all right so next would be dynamic
134. 5:19information
135. 5:20so dynamic information changes
136. 5:23frequently
137. 5:24okay so information related to events in
138. 5:27the network
139. 5:28so the change of state of a device
140. 5:32for instance the router serial 0 0
141. 5:350 goes down and then after a second it
142. 5:38went up
143. 5:39okay so this our dynamic information the
144. 5:41change of state from
145. 5:42on to off or from up down
146. 5:45these are considered to be dynamic
147. 5:47information okay
148. 5:49so the transition or transmission and
149. 5:51reception of packets
150. 5:53are considered to be dynamic information
151. 5:56okay so this would
152. 5:57uh basically depend on the number of
153. 6:00packet that traverses over the
154. 6:02network so this is considered to be
155. 6:04dynamic
156. 6:05or ever changing information okay so
157. 6:08this would depend
158. 6:09on the number of packet reverses over
159. 6:12the
160. 6:13infrastructure okay so the next one
161. 6:16would be
162. 6:17statistical the statistical information
163. 6:21was derived from
164. 6:23dynamic information okay so example are
165. 6:26the and
166. 6:26the average number of packets
167. 6:28transmitted per unit time
168. 6:30so the number of packets dropped
169. 6:33from the network the number of packets
170. 6:36um
171. 6:37prohibited and denied by the access
172. 6:40lists
173. 6:41okay so configured on the router so
174. 6:43these are all
175. 6:44statistical information okay
176. 6:48so next would be the organization of the
177. 6:51management information base
178. 6:53so when you say mib or the management
179. 6:56information base
180. 6:57this is basically the database
181. 6:59containing the information
182. 7:01pertinent to network management
183. 7:04and organic organization of the mib
184. 7:08includes of course the statistical
185. 7:11database
186. 7:12the dynamic database and this the
187. 7:14statistical or the static sorry static
188. 7:16database in here
189. 7:18okay so again when we say statistical
190. 7:21these are
191. 7:22information derived from dynamic
192. 7:25okay and static database these are the
193. 7:28configurations
194. 7:29fixed on the devices okay
195. 7:33so this includes
196. 7:36this components here okay the
197. 7:39configuration database the sensor
198. 7:41database
199. 7:42for the dynamic databases so you've got
200. 7:45the event the state
201. 7:46okay like what we have mentioned earlier
202. 7:49and the statistical database are of
203. 7:50course
204. 7:52the the statistics or the data derived
205. 7:55from
206. 7:55dynamic databases all right
207. 7:58so the mib is basically the storage
208. 8:02of the configured information on the
209. 8:05device
210. 8:08okay so let's talk about monitoring
211. 8:10system components
212. 8:11so the first one would be monitoring
213. 8:13application
214. 8:15we have mentioned earlier on how do we
215. 8:17monitor or how do we come up
216. 8:19with the information okay so derived
217. 8:23from the
218. 8:23infrastructure okay so
219. 8:26or what tools are we gonna use to
220. 8:29capture the information or the data
221. 8:32or the packet that we need during the
222. 8:34monitoring so of course
223. 8:36we will be using the monitoring
224. 8:38application
225. 8:39so monitoring application includes the
226. 8:42functions of monitoring
227. 8:44that are visible to the user okay
228. 8:47example performance
229. 8:49fault accounting so these are all
230. 8:52functions of monitoring that is or that
231. 8:55are visible to the end users
232. 8:58okay so basically performance you'll see
233. 9:00it and you feel it
234. 9:02okay so by by using tools
235. 9:06like um the speed test
236. 9:09okay so for the internet connectivity so
237. 9:11we can see and compare
238. 9:13okay so having a benchmark the
239. 9:15performance of the network
240. 9:18all right so aside from monitoring
241. 9:20application we also have the
242. 9:22manager function so manager function
243. 9:25performs the basic monitoring function
244. 9:28of retrieving information
245. 9:32okay so from the previous discussion on
246. 9:34the management
247. 9:36overview or the network management
248. 9:38overview we've talked about
249. 9:40the manager agent paradigm
250. 9:43all right so on the manager agent
251. 9:47paradigm so we've talked about the
252. 9:49duties and responsibilities or the
253. 9:51function of
254. 9:51each okay where in the manager is
255. 9:54basically the application residing on
256. 9:57the server where the administrator is in
257. 10:00okay
258. 10:01and the agent is basically
259. 10:05uh the operating systems or the software
260. 10:09residing on the managed object that
261. 10:12captures
262. 10:12information locally and for word data
263. 10:16on the manager okay so the
264. 10:20manager function you also have the agent
265. 10:22function
266. 10:23okay so which captures data or gathers
267. 10:26records
268. 10:27okay so from the managed object or mos
269. 10:30so every managed devices on the
270. 10:32infrastructure has an agent
271. 10:35and the agent function is of course to
272. 10:37fetch data from
273. 10:39the local device and forward it to the
274. 10:41manager
275. 10:42so the manager function is to perform
276. 10:44basic monitoring function of retrieving
277. 10:46information
278. 10:47so it receives data from the agent
279. 10:51all right so also we kept on mentioning
280. 10:55managed object
281. 10:56or managed objects so what are this so
282. 10:59managed objects
283. 11:00pertains to a management information
284. 11:04that represents resources and their
285. 11:07activities
286. 11:08so managed objects or mo's are
287. 11:11routers switches firewalls workstations
288. 11:14servers
289. 11:15these are all managed objects okay so
290. 11:18the device that we are monitoring and
291. 11:20controlling
292. 11:21are said to be managed object all right
293. 11:24next would be a monitoring agent so the
294. 11:27monitoring agent generates
295. 11:29summaries and statistical analysis of
296. 11:31this management information
297. 11:34okay and we use it for
298. 11:38future planning that's what i've said on
299. 11:40the introduction
300. 11:41all right so next would be the
301. 11:44functional architecture
302. 11:45of network monitoring the manager agent
303. 11:49paradigm or model that we have presented
304. 11:51on the network management overview
305. 11:53now let's have some recap in here okay
306. 11:56so the manager agent model
307. 11:57is basically the relationship between
308. 12:01the manager and the agent okay
309. 12:04so the monitoring application is
310. 12:07basically a manager function okay
311. 12:11and on a managed object there is an
312. 12:13agent
313. 12:14and of course an agent function so the
314. 12:17communications between the server
315. 12:20having the management application or the
316. 12:22manager
317. 12:24to the agent so basically this is
318. 12:27true or via the use of the simple
319. 12:30network management protocol or snmp
320. 12:32so we'll be talking about snmp in the
321. 12:35next and upcoming videos
322. 12:38all right so for the model of
323. 12:41summarization
324. 12:42okay so again in here is the server this
325. 12:45is where the systems administrator is
326. 12:47seated
327. 12:47okay monitoring the performance of the
328. 12:49network via the application
329. 12:52and we have here the monitoring agent
330. 12:55whose primary
331. 12:57purpose is to summarize okay
332. 13:00so the information forwarded by these
333. 13:02agents
334. 13:03so there might be two or more devices
335. 13:07or a lot of mo's connected to the
336. 13:10um to the monitoring agent okay
337. 13:14so that's it next
338. 13:17how about the network monitoring
339. 13:19configuration
340. 13:21okay so letter a here
341. 13:24manage resources in a manager system
342. 13:27okay so there is a possibility that we
343. 13:30are actually monitoring
344. 13:32okay so a device where the application
345. 13:35where the manager
346. 13:37is also installed for example
347. 13:40you want to monitor the performance of
348. 13:42your server
349. 13:43so as a systems administrator you are
350. 13:46seated in front of the server and you
351. 13:48are using definitely
352. 13:49a monitoring application okay now this
353. 13:52monitoring application performs a
354. 13:54manager function
355. 13:55which is to receive information from
356. 13:59the agent okay so the agent is on the
357. 14:02managed object
358. 14:03now if it happens that you are
359. 14:05monitoring itself or monitoring the
360. 14:08the server okay so where the application
361. 14:10and the agent is stored
362. 14:12so that is this model here okay
363. 14:16so that would be possible okay so you
364. 14:18are monitoring your own device
365. 14:20okay so the agent is installed on the
366. 14:22server the manager is also installed on
367. 14:25the server
368. 14:26all right next
369. 14:29resources in the agent system so
370. 14:32with this type of model the monitoring
371. 14:34application or the manager function
372. 14:37or simply say manager is
373. 14:40stored or installed in a computer
374. 14:44okay different from where the agent is
375. 14:48installed
376. 14:49so for example this is a server here and
377. 14:51i have here a router
378. 14:53okay and i manage the router
379. 14:56using the subnet or the internet okay
380. 15:00so that is this resources in the agent
381. 15:02system
382. 15:03okay or also it is also possible to
383. 15:08monitor okay or to control
384. 15:11any devices are managed object outside
385. 15:15the network or it could be over the
386. 15:16internet
387. 15:18and we have here an agent function
388. 15:21which is to summarize all the
389. 15:23information gathered from demos
390. 15:27okay and the fourth one would be a proxy
391. 15:30monitor agent so again when say proxy
392. 15:33this is the software that we use so that
393. 15:36this
394. 15:37proprietary software can communicate
395. 15:39with an open systems and vice versa
396. 15:42all right next
397. 15:46network monitoring methods so how do we
398. 15:49monitor
399. 15:50the network so one is via
400. 15:53polling okay so when you say polling
401. 15:56this is a request response interaction
402. 15:59between the manager and the agent
403. 16:01so polling is basically initiated by the
404. 16:05manager
405. 16:06going to the agent okay so
406. 16:09a manager sends a request to an agent
407. 16:12which processes the request and responds
408. 16:14with information from hmib
409. 16:18all right so for example if i am the
410. 16:20systems administrator and i
411. 16:22am seated in a computer where i control
412. 16:26and configure for instance a router okay
413. 16:29so i am using puti puti is basically
414. 16:33my manager okay and on the router
415. 16:37this router uses an ios so basically the
416. 16:40ios
417. 16:41captures data okay from its mib
418. 16:45okay or from its virtual database and
419. 16:48that data will be forwarded to the puti
420. 16:51to present it to the administrator so
421. 16:53that is an example of polling
422. 16:55okay for instance the administrator
423. 16:58wishes to display or to know the routing
424. 17:01table of the router
425. 17:03so we simply execute the command show ip
426. 17:05route
427. 17:06on the putsy application okay
428. 17:09so that is an example of polling where
429. 17:12in the manager
430. 17:14send a request to the agent and the
431. 17:16agent will respond
432. 17:18with the request of the manager so
433. 17:20that's polling
434. 17:21okay so all the show commands that we
435. 17:24use
436. 17:24in cisco routers or cisco switches
437. 17:28or in firewall are all examples of
438. 17:32polling all right so
439. 17:35a manager may use polling to learn about
440. 17:38the configuration it is managing
441. 17:40to obtain periodically an update or
442. 17:44or an update of conditions or
443. 17:46investigate an
444. 17:47area in detail after being altered
445. 17:51to a problem so if the action is
446. 17:55initiated by
447. 17:56the manager we call it polling
448. 17:59all right okay
449. 18:02so how about if for instance the router
450. 18:06interface serial zero zero goes down and
451. 18:10the router sends a notification
452. 18:13onto the manager okay so is it polling
453. 18:17again with this example okay
454. 18:20so the agent initiated the action and we
455. 18:24call it event reporting
456. 18:26all right so event reporting the
457. 18:29information flow is initiated from the
458. 18:31agent to the manager
459. 18:32when you say polling it is from the
460. 18:34manager
461. 18:35to agent all right so the counterpart of
462. 18:38polling is event reporting
463. 18:41so an agent may generate report
464. 18:43periodically
465. 18:44to give the manager its current status
466. 18:47or
467. 18:47whenever a significant event happens
468. 18:50okay so like for inside for instance
469. 18:54due to intermittent uh interfaces
470. 18:58okay so your serial zeros like zero zero
471. 19:01goes down after a second it went up
472. 19:05okay so there is a notification from
473. 19:07your screen
474. 19:08okay so about the status of the
475. 19:10interface and we call it event reporting
476. 19:14okay so when you're printing something
477. 19:16okay for instance
478. 19:18and the printer runs out of paper so
479. 19:20there is a notification in your screen
480. 19:23be a sort of a message informing you
481. 19:26that
482. 19:27the printer is running out of paper so
483. 19:30that is an example of reporting
484. 19:32okay or event reporting all right
485. 19:36so this is good for detecting problem as
486. 19:39soon as they occur
487. 19:40okay so because this is real time
488. 19:43all right you could even configure your
489. 19:45device to have some event reporting by
490. 19:48your your cell phone or via a mobile app
491. 19:51on your gadgets all right so that
492. 19:55would ease the troubleshooting
493. 20:01okay so next would be performance
494. 20:04monitoring
495. 20:06okay so performance monitoring is
496. 20:09measuring the performance of the network
497. 20:12or performance monitoring
498. 20:15that's absolutely required in network
499. 20:17management
500. 20:19okay so the objective
501. 20:22is to detect and fix problems that cause
502. 20:26performance degradation so you've got
503. 20:28slowing performance
504. 20:30okay the performance is worse than
505. 20:34the usual all right so you've got
506. 20:38performance monitoring in there okay and
507. 20:41what do we do with that well the
508. 20:43performance monitoring
509. 20:46can be used to better plan network
510. 20:48upgrades
511. 20:50okay so basically if you're going to
512. 20:52implement a network
513. 20:54so on its first run or initial run the
514. 20:57network is doing good
515. 20:58okay and then after some time okay so of
516. 21:01course the network
517. 21:03ages down okay you've got the number of
518. 21:06um
519. 21:07connections also increases okay so what
520. 21:10will happen is
521. 21:11basically there would be performance
522. 21:13degradation due to
523. 21:14aging technology okay or due to an
524. 21:18increase
525. 21:19in the number of users okay
526. 21:22so to to better upgrade
527. 21:26or to better plan and do some network
528. 21:28upgrades
529. 21:29that would help okay via performance
530. 21:32monitoring
531. 21:33okay or we could address the problem via
532. 21:35performance monitoring
533. 21:37so what are the problems in selecting
534. 21:39and using appropriate indicators or
535. 21:40metrics
536. 21:42okay so basically the first one would be
537. 21:46too many indicators and use okay
538. 21:49so it's just like you enabled all the
539. 21:52features that you have
540. 21:54on the application okay so for instance
541. 21:59when you installed microsoft office okay
542. 22:02so you have
543. 22:02or you are given these options what
544. 22:05specific
545. 22:06feature okay needs to be installed
546. 22:09on your microsoft office package so you
547. 22:12have the option to choose let's say
548. 22:15word powerpoint excel okay but you also
549. 22:18have this
550. 22:19choose all or select all and then you
551. 22:21install it
552. 22:22okay now the thing is you have a lot of
553. 22:25applications
554. 22:26installed on your computer but you are
555. 22:28not actually using them all
556. 22:30right so that is similar with having
557. 22:34too many indicators in use or by having
558. 22:39too many components that we want to
559. 22:42monitor
560. 22:43and yet some of it we are we are not
561. 22:46using it
562. 22:47okay the meaning of most indicators
563. 22:52are not yet clearly understood okay so
564. 22:55we simply
565. 22:56choose to enable them all but we don't
566. 22:59know
567. 23:00where to use it or how what's the
568. 23:02significance of it
569. 23:03or what's the importance of it on your
570. 23:05organization
571. 23:06okay so be careful when you install
572. 23:09performance
573. 23:10monitoring applications because some of
574. 23:12it might not be needed in the
575. 23:14organization
576. 23:15okay so take note that whenever we check
577. 23:18the features
578. 23:19on the applications so that would cause
579. 23:22traffic
580. 23:23and that traffic would cause performance
581. 23:25degradation
582. 23:26so you choose only indicators which
583. 23:30the organization might need most
584. 23:34all right next would be some indicators
585. 23:36are supported by
586. 23:38some manufacturers only okay so
587. 23:41why there is too many indicators in use
588. 23:44because
589. 23:44of the bullet number three so if i am
590. 23:47using apple then i'll be using
591. 23:50monitoring applications for apple so
592. 23:53i'll
593. 23:53if i'm using linux i'll be using
594. 23:55monitoring application for you
595. 23:57for linux okay so
596. 24:00that would lead to many indicators
597. 24:03in use okay next would be
598. 24:07frequently the indicators are accurately
599. 24:11measured but incorrectly interpreted by
600. 24:13human
601. 24:13or the management application there's a
602. 24:15problem here okay
603. 24:17so we've gathered the data
604. 24:20accurately and correctly but when we
605. 24:23interpreted it
606. 24:24well problem arises all right
607. 24:28so we've got an incorrect interpretation
608. 24:30of the results
609. 24:32okay next would be
610. 24:36the calculation of indicators takes too
611. 24:38much time
612. 24:40why performance aggregation okay
613. 24:43so that is because of traffic
614. 24:47why there is too much traffic because
615. 24:49you have too many indicators
616. 24:51checked on your monitoring application
617. 24:54okay so it's a chain reaction
618. 25:01okay so next is how do we measure
619. 25:04performance indicators
620. 25:06okay so it's either service oriented
621. 25:09okay so when say service oriented we've
622. 25:12talked about
623. 25:13availability so availability pertains to
624. 25:16the percentage of time that the network
625. 25:18system or
626. 25:19component or an application is available
627. 25:22for a user
628. 25:23okay so we measured service oriented
629. 25:27via the availability of the device
630. 25:30the availability of the resources okay
631. 25:34next would be the response time so how
632. 25:37long it takes for a response to appear
633. 25:39at a user's terminal after the user's
634. 25:42accident calls for it okay response time
635. 25:46for example you have encountered an
636. 25:48error on your system or on your computer
637. 25:51and you seek help from the help desk
638. 25:54how long would it take for the help test
639. 25:56to address your problem
640. 25:57response time okay
641. 26:01or how fast are they to resolve the
642. 26:04issues that you have
643. 26:05response time next would be accuracy
644. 26:08the percentage of time that no errors in
645. 26:11the transmission
646. 26:12and delivery of information because if
647. 26:16there is an
648. 26:16error there is a tendency to retransmit
649. 26:19the data
650. 26:20okay so again service oriented
651. 26:24network performance indicators are
652. 26:27measured using availability
653. 26:29response time and accuracy
654. 26:32so the second is efficiency oriented
655. 26:37okay so efficient efficiency oriented
656. 26:41we talked about throughput okay
657. 26:45throughput is the rate at which
658. 26:46application oriented events
659. 26:48file transfers occur okay so throughput
660. 26:52is related to bandwidth okay
661. 26:55but if we are after the efficiency we're
662. 26:58not talking about bandwidth
663. 27:00but we measure it instead using
664. 27:02throughput
665. 27:03now what's the difference between
666. 27:04throughput and bandwidth
667. 27:06bandwidth is basically the
668. 27:09frequency is the difference between the
669. 27:11highest and the lowest frequency
670. 27:13so in say bandwidth we're talking about
671. 27:15the capacity
672. 27:16okay the capacity of the internet what
673. 27:18is your maximum speed
674. 27:20all right what is your minimum speed
675. 27:22minimum is zero maximum is 100 for
676. 27:25instance
677. 27:25so that range from 0 to 100 is said to
678. 27:29be the bandwidth
679. 27:30all right when you say throughput this
680. 27:32is the actual transfer of data
681. 27:35so if you have 0 to 100 mb and
682. 27:38you're getting 90 mb then that is the
683. 27:41throughput
684. 27:42all right next would be utilization
685. 27:46so the percentage of the theoretical
686. 27:49capacity of a resource
687. 27:50example transmission line switch cpu
688. 27:54that is being used
689. 27:56okay so we can
690. 27:59monitor this via an application or
691. 28:02if you are using your your pc you can
692. 28:05monitor the cpu utilization the ram
693. 28:07utilization
694. 28:08okay via the task manager and that is an
695. 28:11efficiency oriented
696. 28:12performance indicator okay
697. 28:16so we measure the performance either
698. 28:19via efficiency oriented or service
699. 28:22oriented so of course that would depend
700. 28:25okay next
701. 28:28now from the figure here okay
702. 28:31so what is or what are the elements of
703. 28:34response time
704. 28:36okay so how long would it take for a
705. 28:39device to respond
706. 28:40on your inquiry okay is it real time
707. 28:44no there's no such thing as real time in
708. 28:46the network
709. 28:47the mere fact that your data traverses
710. 28:49on the cable
711. 28:51on this interfaces it causes delay
712. 28:55okay so the truth is for
713. 28:58every response that we're getting okay
714. 29:00so there
715. 29:01is an integration of delays on it all
716. 29:04right
717. 29:05so the elements of response time these
718. 29:07are delays
719. 29:08okay so if you observe here now from the
720. 29:12equation
721. 29:13okay so that's what i'm saying earlier
722. 29:14when your data traverses on the cable
723. 29:17there is delay and it gets into the
724. 29:19hardware there is delay
725. 29:21okay so going back again to the cable
726. 29:23another delay and so on
727. 29:25okay so therefore response time
728. 29:28is equals to the inbound terminal delay
729. 29:32plus the inbound queueing time all right
730. 29:36plus the inbound service time and so on
731. 29:39so basically
732. 29:40there is always delay for every response
733. 29:43time
734. 29:44right so it's just that as systems
735. 29:47administrator how do we minimize
736. 29:50delays okay
737. 29:53next performance monitoring functions
738. 29:57okay so this includes performance
739. 29:59measurement
740. 30:00and performance analysis now let's talk
741. 30:04about performance measurement
742. 30:06so this is the actual gathering of
743. 30:08statistics
744. 30:09about network traffic and timing okay
745. 30:13so we do it via observation
746. 30:16okay and of course with the help of
747. 30:18tools we get it
748. 30:20okay so typically performed by agents
749. 30:23within network devices and
750. 30:26we these agents forward the data
751. 30:29to the manager okay so by uh snmp
752. 30:34and of course the systems administrator
753. 30:37can view and analyze okay so that data
754. 30:42okay so example includes the amount of
755. 30:46data in and out of the node
756. 30:47or the number of connections traffic per
757. 30:50connection
758. 30:51so that's performance measurement okay
759. 30:54next would be performance analysis so
760. 30:57analyzing the guard data
761. 30:59and presenting it so example you've got
762. 31:02the total the average the min max
763. 31:04histogram
764. 31:05something like the statistical
765. 31:06information about the gathered data
766. 31:09we call it performance analysis okay
767. 31:12and usually we compare it or we do a
768. 31:15benchmark of it on the previous records
769. 31:17that we have
770. 31:18okay and we are expecting
771. 31:22okay so we've got the improvement
772. 31:25or better service provided to the end
773. 31:27users
774. 31:29all right next is
775. 31:33synthetic traffic generation so what is
776. 31:35this so generating artificial traffic
777. 31:38load
778. 31:38to test okay your network so for
779. 31:41instance
780. 31:42if i'm going to design a network that is
781. 31:44good for for
782. 31:46100 workstations and i only have 10 here
783. 31:49so i could use the synthetic traffic
784. 31:51generator to test
785. 31:53whether the network is capable of having
786. 31:56100 workstations though i only have 10.
787. 31:59okay so something like doing a
788. 32:01simulation test
789. 32:03okay it permits the network to be
790. 32:05observed under controlled
791. 32:07load okay
792. 32:11all right so next would be a typical
793. 32:14performance related questions
794. 32:16okay so performance measurements can be
795. 32:19used
796. 32:20to answer a number of questions so these
797. 32:23questions could help us
798. 32:25improve the performance okay so
799. 32:29if we are asked or invited okay so to
800. 32:32ask questions
801. 32:33on how to improve the performance
802. 32:36of the network well these questions
803. 32:39could help you
804. 32:40okay number one why is the response so
805. 32:43slow
806. 32:45so if you are the i.t personal and i am
807. 32:48the
808. 32:48end users and they ask you this question
809. 32:50how would you answer it
810. 32:52why is the response so slow are you
811. 32:55going
812. 32:56to give me answers like because the
813. 32:57internet is low
814. 32:59or maybe we have um increased the number
815. 33:03of users on the network
816. 33:04on the network that is why the response
817. 33:06is so slow
818. 33:08okay and you ask the same question okay
819. 33:11same questions
820. 33:12were asked after a year
821. 33:15and how would you address it are you
822. 33:17going to give us the same answer
823. 33:20okay so this is a very loaded question
824. 33:22so why is the response so slow
825. 33:24okay next would be why is the
826. 33:27transmission rate so high so this is
827. 33:29somehow technical
828. 33:31okay retransmission happens because
829. 33:34the data or there is a drop on the
830. 33:36transmission of packets
831. 33:38okay so if there is a drop in the
832. 33:40transmission of packet the tendency is
833. 33:42the devices will retransmit it
834. 33:44especially if you are transmitting a tcp
835. 33:47packets
836. 33:48okay so that requires an acknowledgement
837. 33:51all right
838. 33:53next is traffic evenly distributed among
839. 33:56network users
840. 33:58or are there source destination pairs
841. 34:01with unusually heavy traffic
842. 34:04okay so for instance in an organization
843. 34:07basically
844. 34:08we give um traffic or we give bandwidth
845. 34:12okay to every department depending on
846. 34:16their
847. 34:16functions and the organization okay so
848. 34:20allowing or assigning bandwidth to
849. 34:24marketing
850. 34:25is different compared to assigning
851. 34:28bandwidth to accounting personnel
852. 34:31all right so you you assign bandwidth
853. 34:35depending on its function okay
854. 34:38in the organization all right
855. 34:42next what is the percentage for its type
856. 34:45of packet transmitted
857. 34:47so there should be a higher percent
858. 34:50of a successful transmission than
859. 34:52failure
860. 34:53okay next what is the channel
861. 34:56utilization and throughput
862. 34:58so we discussed throughput and defined
863. 35:00throughput earlier
864. 35:02and the channel utilization okay so
865. 35:05we're talking about the ratio here okay
866. 35:07so out of
867. 35:08say you'll have 30 mbps
868. 35:12okay for instance okay so that's your
869. 35:16your bandwidth what are you getting from
870. 35:18it
871. 35:19are you getting 30 also are you getting
872. 35:22just 20
873. 35:23okay so utilization
874. 35:27is it should be higher okay
875. 35:30and the throughput should also be higher
876. 35:32it should be close to
877. 35:33the bandwidth all right
878. 35:36next what is the effect of traffic load
879. 35:38on utilization
880. 35:40throughput and time delays okay
881. 35:44so basically with this
882. 35:47with every traffic on the network
883. 35:50your performance has the tendency to
884. 35:52decrease
885. 35:53okay so if you want to address really
886. 35:56the performance you have to improve
887. 35:58the bandwidth you have to improve the
888. 36:00throughput
889. 36:02the utilization and that would diminish
890. 36:05delays
891. 36:07okay next when does traffic load start
892. 36:10to degrade system performance
893. 36:13okay so for example in school okay
894. 36:16so if you have a class at 7 30 in the
895. 36:18morning okay so
896. 36:19when you came in there early at six
897. 36:21o'clock okay so basically you're gonna
898. 36:23have a good
899. 36:25internet access a better internet access
900. 36:28but as the number of students increases
901. 36:31in the campus
902. 36:32so you will experience some degradation
903. 36:36in terms of internet access that is
904. 36:38because
905. 36:39a number of users okay a bulk of users
906. 36:43are trying to access the same resource
907. 36:45that you are accessing
908. 36:47okay so you should be able to know
909. 36:50the time when the performance
910. 36:53aggregation
911. 36:54happens on the network and also you
912. 36:56should be able to identify
913. 36:58okay the the peak okay i mean
914. 37:02during what time would be the
915. 37:05a very good or excellent internet access
916. 37:08every sunday
917. 37:09all right or maybe 4 30 to 9 we've got
918. 37:12a better internet access that is because
919. 37:15less users
920. 37:16are in the campus during that time okay
921. 37:19so a systems administrator
922. 37:21you should know okay the time
923. 37:25when to start the
924. 37:28allocation of the resources okay so to
925. 37:31anticipate the
926. 37:32[Music]
927. 37:33volume of users in the network to avoid
928. 37:36performance degradation okay so we have
929. 37:39some sort of reallocation of resources
930. 37:41maybe
931. 37:42right and the last one would be
932. 37:46what is the maximum capacity of the
933. 37:48channel under normal operating
934. 37:49conditions
935. 37:50and how many active users are necessary
936. 37:53to reach this
937. 37:54maximum okay so again as network
938. 37:57administrator
939. 37:58or a systems administrator you should
940. 37:59know that
941. 38:01this channel or that this network
942. 38:04or that this internet could only
943. 38:08allow 100 people to access okay
944. 38:11so if you have more than 100 then
945. 38:15performance duplication starts or begins
946. 38:19okay and a systems administrator you
947. 38:21should have
948. 38:23the solution to that okay so again the
949. 38:26possible solution is
950. 38:27there might be a need for reallocation
951. 38:30of resources all right
952. 38:35okay so next would be fault monitoring
953. 38:39okay so to detect faults as quickly as
954. 38:42possible
955. 38:43after they occur and to identify the
956. 38:45cause of the fault so that correctional
957. 38:47action may be taken
958. 38:49so a systems administrator well it's our
959. 38:51duty to
960. 38:52[Music]
961. 38:54maintain okay a fault free
962. 38:57network okay so
963. 39:00problems on fault monitoring so this
964. 39:03includes fault detection problem
965. 39:05so there is an observable fault example
966. 39:07deadlock
967. 39:08okay so device not monitorable deadlock
968. 39:12is unobservable i mean you don't know
969. 39:15when it will be coming
970. 39:16okay so we all know that that luck
971. 39:18happens only if
972. 39:20the resources requested by the
973. 39:22application cannot be granted
974. 39:24okay so there is also partial or
975. 39:27partially observable faults
976. 39:29inefficient to pinpoint what a problem
977. 39:31is
978. 39:32okay or uncertainty in observation
979. 39:35so not clear what really the problem is
980. 39:39okay so that's a fault detection problem
981. 39:42you don't know the source of the problem
982. 39:44so how can we solve it
983. 39:46next would be fault isolation problems
984. 39:49so multiple potential causes
985. 39:52too many related observations sometimes
986. 39:55this is our
987. 39:56problem we have a lot of information we
988. 39:58have a lot of
989. 39:59a lot of inputs and we have a hard time
990. 40:02to
991. 40:02digest okay so we have too many related
992. 40:06observations that is why
993. 40:08so that takes a slow in solving the
994. 40:11problem
995. 40:12okay interferences between diagnosis
996. 40:15and local recovery procedures and
997. 40:18absence of automated testing tools so we
998. 40:20do it manually
999. 40:21so you've got fault isolation problem
1000. 40:25all right now considering this diagram
1001. 40:28here
1002. 40:29okay we only have one link
1003. 40:32okay connecting to the server now what
1004. 40:35will happen if this t1 link fails
1005. 40:39okay so it's dc or disconnected
1006. 40:44okay so what it says here is that
1007. 40:48whenever you you create connections to
1008. 40:51the server for
1009. 40:52instance there might be a sort of
1010. 40:54redundancy
1011. 40:55okay so that when the primary link goes
1012. 40:57down so there is an alternate or backup
1013. 40:59link
1014. 41:00okay so that will capture
1015. 41:03okay so the the connections
1016. 41:08with the server or with the devices
1017. 41:11all right so if the
1018. 41:15primary link goes up so there is also a
1019. 41:18chance that
1020. 41:19it will go back to the original state i
1021. 41:22mean
1022. 41:22so the backup will go on the backup
1023. 41:25state
1024. 41:26and the primary will go on the primary
1025. 41:29duty
1026. 41:30all right
1027. 41:34okay so propagation of failures to a
1028. 41:36higher layers so basically
1029. 41:38being a systems administrator you should
1030. 41:40know where each component
1031. 41:43of a network device works okay so for
1032. 41:45instance router router basically works
1033. 41:47on the layer three
1034. 41:48okay so it just works in the layer two
1035. 41:50okay so if there would be a problem for
1036. 41:53instance
1037. 41:54okay so on the connections on the router
1038. 41:57so basically that would be propagated
1039. 41:59onto the application layer okay so if
1040. 42:02there would be problem with the
1041. 42:03connections here
1042. 42:04that can be visualized on the
1043. 42:06application layer so we could have
1044. 42:07something like
1045. 42:08if we can initially ping the connection
1046. 42:10then after some time we've got request
1047. 42:12timeout
1048. 42:13all right or if there would be a problem
1049. 42:17on the lower layer
1050. 42:18so that would be propagating up to the
1051. 42:20highest layer
1052. 42:22all right so that would aid
1053. 42:25troubleshooting
1054. 42:26okay so how would you start
1055. 42:27troubleshooting in here so are you going
1056. 42:29to use
1057. 42:31a bottom up approach or a top down
1058. 42:33approach
1059. 42:35okay so usually if there will be a
1060. 42:36problem on the hardware
1061. 42:38that propagates onto the higher layers
1062. 42:43so you should know okay the techniques
1063. 42:46okay that is being used on
1064. 42:49troubleshooting
1065. 42:50network troubleshooting specifically
1066. 42:55okay so next would be default monitoring
1067. 42:58functions
1068. 43:00so let's start with logs or lagging
1069. 43:02facilities so record important events
1070. 43:04and errors so lag should be accessible
1071. 43:09by managers via polling so we have
1072. 43:12introduced and defined
1073. 43:13polling okay from the previous slides
1074. 43:16okay
1075. 43:16so there should be a lagging facilities
1076. 43:19okay
1077. 43:20event reporting sending events error to
1078. 43:24managers
1079. 43:25sending alarms to manager
1080. 43:28to warm possible problems so we're
1081. 43:30talking we're talking about
1082. 43:31notifications here okay so notifications
1083. 43:36that would be initiated by the agent to
1084. 43:38the manager
1085. 43:40you could also use diagnosis uh
1086. 43:42diagnostic function like the
1087. 43:44connectivity test
1088. 43:46so traceroute okay or tracer
1089. 43:50right trace route on the routers and
1090. 43:52tracer on
1091. 43:54the computers you can also use the
1092. 43:56response time
1093. 43:57test okay
1094. 44:00the live nest test or the ping
1095. 44:04okay the protocol integrity test
1096. 44:08you bring one two seven zero zero one so
1097. 44:11that is the test where your protocol
1098. 44:12stack is
1099. 44:14properly working okay you could also
1100. 44:17use the loopback test okay or
1101. 44:20same with the protocol integrity test
1102. 44:22something like one two seven zero zero
1103. 44:24one
1104. 44:24it's a look back all right
1105. 44:28next how about accounting so accounting
1106. 44:31monitoring
1107. 44:32keeping track of the users usage of
1108. 44:35network resources
1109. 44:38so communication facilities
1110. 44:41computer hardware software and systems
1111. 44:44and
1112. 44:45services okay so how do you monitor the
1113. 44:48utilization for instance
1114. 44:50of an employee okay so
1115. 44:53we use application okay so as always
1116. 44:56for every monitoring activities so we
1117. 44:59use
1118. 44:59applications so you we don't want this
1119. 45:02to
1120. 45:03perform manually all right so
1121. 45:06usage may be
1122. 45:08[Music]
1123. 45:09need to be broken down by account by
1124. 45:13project
1125. 45:13or by individual user for appropriate
1126. 45:16accounting purposes
1127. 45:19okay so before we end up let's wrap up
1128. 45:23so let's summarize what we have
1129. 45:25discussed and learned from this
1130. 45:27video so first we talked about network
1131. 45:29monitoring
1132. 45:31okay and it is the most basic aspect
1133. 45:34of network management okay
1134. 45:37or part of the system's administration
1135. 45:41or systems administrators duty and
1136. 45:43responsibility
1137. 45:45okay we also talked about the purpose of
1138. 45:48network monitoring
1139. 45:50that is to gather information about the
1140. 45:51status and behavior of the network
1141. 45:53elements
1142. 45:54okay so information to be gathered
1143. 45:58includes
1144. 45:58static okay dynamic and statistical
1145. 46:01information
1146. 46:03monitoring methods we've talked about
1147. 46:05polling and event reporting
1148. 46:06okay polling initiated by the manager
1149. 46:10and event reporting initiated by the
1150. 46:12agent
1151. 46:13also we've talked about the monitoring
1152. 46:15functions performance monitoring
1153. 46:18fault monitoring and accounting
1154. 46:20monitoring
1155. 46:22okay so if you will observe okay so
1156. 46:25network monitoring is more on the
1157. 46:27performance
1158. 46:28fault and accounting where is security
1159. 46:32and configuration we've got f cops right
1160. 46:35fcaps fault is there configuration is
1161. 46:39missing
1162. 46:40all right so accounting is there
1163. 46:44you've got performance is there and
1164. 46:47security is not in there
1165. 46:49okay so the configuration
1166. 46:52and security will be covered on the next
1167. 46:55video
1168. 46:56okay so that is network control all
1169. 46:58right
1170. 46:59so that ends up the video so thank you
1171. 47:02for sticking around and thanks for
1172. 47:04watching have a great day
1173. 47:10[Music]
1174. 47:17you