1 00:00:00,360 --> 00:00:09,960 The border will become the D.R. if the D.R. files designated Rodders are chosen on a per segment basis. 2 00:00:10,380 --> 00:00:18,270 So in this example, on this Ethernet segment or two may be chosen as a designated rider, but on this 3 00:00:18,270 --> 00:00:26,670 segment between our four and or seven of form may be chosen as a designated rider on a per physical 4 00:00:26,670 --> 00:00:27,820 segment basis. 5 00:00:27,840 --> 00:00:29,850 A designated route is chosen. 6 00:00:30,210 --> 00:00:35,730 So just because our four is not a designated rider on this segment does not mean that our four is not 7 00:00:35,730 --> 00:00:37,560 a designated rider on this segment. 8 00:00:37,680 --> 00:00:43,290 Every Ethernet segment requires the election and maintenance of a designated rider. 9 00:00:43,770 --> 00:00:50,550 So from an update point of view, if this link goes down or one is updating its designated rider using 10 00:00:50,550 --> 00:00:59,370 this multicast address 224006 or to the designated rider updates all of its neighbors on that segment. 11 00:00:59,370 --> 00:01:08,640 So are three or six or five and or four get the update about the change of foal then floods that update 12 00:01:08,640 --> 00:01:14,070 to or seven or seven will then flood that update to any neighbors that it may have not shown in this 13 00:01:14,070 --> 00:01:14,880 diagram. 14 00:01:14,880 --> 00:01:18,360 The same with or five or six or three and so forth. 15 00:01:18,630 --> 00:01:23,460 The update will be flooded throughout the topology, just as in this example where all four receive 16 00:01:23,460 --> 00:01:26,910 the update from our two and then sends it to our seven. 17 00:01:27,600 --> 00:01:33,780 So just to sum up, designated riders or dealers are elected firstly based on the highest priority. 18 00:01:34,260 --> 00:01:37,590 The range is from 0 to 55. 19 00:01:38,250 --> 00:01:46,170 The default value is one zero disables the ability for this rider to become a designated rata or backup 20 00:01:46,170 --> 00:01:47,250 designated rider. 21 00:01:47,640 --> 00:01:53,490 If the priorities are the same, the rider with the highest rata ID will become the designated rider. 22 00:01:53,730 --> 00:01:58,230 It's important to realize that preemption doesn't exist with designated riders. 23 00:01:59,370 --> 00:02:04,920 So in this example, if the priority of router two was set to ten and the priority of right of five 24 00:02:04,920 --> 00:02:10,650 was set to five and an election took place or two would become the designated Rada. 25 00:02:11,070 --> 00:02:19,440 However, if r t went down, rather five would become the designated rider, and for instance, R three 26 00:02:19,440 --> 00:02:21,570 would become the backup designated rider. 27 00:02:22,200 --> 00:02:27,360 Let's assume that the priorities of the other riders are set to one, but rather three has the highest 28 00:02:27,360 --> 00:02:31,290 rate of ID, so rather three becomes the designated rider. 29 00:02:31,970 --> 00:02:39,320 When R2 comes back again, it will not become a designated rider or backup designated rider. 30 00:02:39,500 --> 00:02:46,250 In other words, it's not preemptive because our five is already the designated rider and all three 31 00:02:46,250 --> 00:02:52,940 is the backup designated rider or two will just accept that fact and will become known as a DH or other, 32 00:02:52,940 --> 00:02:55,760 or draw the depending which term you prefer. 33 00:02:56,150 --> 00:03:01,580 It will not try and preempt another election and try and become the designated rider. 34 00:03:02,030 --> 00:03:07,580 All other riders, including our two, will become DE or others in a topology like this. 35 00:03:08,870 --> 00:03:12,170 Let's talk about the shortest path first algorithm or SPF algorithm. 36 00:03:12,950 --> 00:03:19,490 The SBF algorithm places each router at the root of a tree and calculates the shortest path to each 37 00:03:19,490 --> 00:03:26,390 node using the Dijkstra's algorithm based on the cumulative cost that is required to reach that destination. 38 00:03:26,630 --> 00:03:33,440 So as an example, if rather one wants to get to a network behind router two, it's going to determine 39 00:03:33,440 --> 00:03:39,350 the best route based on the cost, which uses a Formula ten to the eight divided by bandwidth. 40 00:03:39,770 --> 00:03:42,890 So as an example, this is at1 link router. 41 00:03:42,890 --> 00:03:44,810 One wants to send traffic to router two. 42 00:03:45,200 --> 00:03:52,100 Will it take this RT one link or will it use the path via router three using a ten meg and ten meg link? 43 00:03:52,460 --> 00:03:58,580 Now rip using hop count would send the traffic directly to our t, but what would OSPF do? 44 00:03:59,530 --> 00:04:03,430 OSPF by default uses a reference bandwidth of ten to the eight. 45 00:04:03,880 --> 00:04:08,260 You can change this and you need to when you've got gig and ten gig links. 46 00:04:08,950 --> 00:04:13,960 OSPF has been around for many years and in the early days there were no speeds such as gig, ten gig 47 00:04:13,960 --> 00:04:14,890 and so forth. 48 00:04:15,310 --> 00:04:20,200 So today, if you've got gig and ten gig links, you're going to want to change this reference bandwidth. 49 00:04:20,860 --> 00:04:23,020 But for now, let's assume that we're using the default. 50 00:04:23,020 --> 00:04:28,690 So the cost of a link is ten to the eight divided by the bandwidth in bits per second. 51 00:04:29,480 --> 00:04:32,630 Now, one kilobits per second equals 1000 bits per second. 52 00:04:33,050 --> 00:04:36,290 One megabits per second is 1 million bits per second. 53 00:04:36,770 --> 00:04:40,070 Ten megabits per second is 10 million bits per second. 54 00:04:40,310 --> 00:04:47,420 So the cost of a ten megabits per second link is ten to the eight divided by 10 million, which gives 55 00:04:47,420 --> 00:04:51,720 you a cost of ten using the same formula for at1 link. 56 00:04:51,740 --> 00:05:00,200 So for this link between R one and r t, the cost is ten to the eight divided by 154 for 000, which 57 00:05:00,200 --> 00:05:01,880 gives you a cost of 64. 58 00:05:01,910 --> 00:05:10,550 So the cost using this link would be 60 for the cost if the traffic was sent via r two would be as follows 59 00:05:10,550 --> 00:05:11,690 ten to the eight. 60 00:05:12,560 --> 00:05:15,380 But what about 10 million, which is ten? 61 00:05:15,620 --> 00:05:20,270 But they are two lengths of ten, meg, so the total cost is 20. 62 00:05:20,660 --> 00:05:26,830 So the cost via the PT one is 64, but the cost via the ten meg links is 20. 63 00:05:26,840 --> 00:05:34,110 So OSPF is going to choose the link via R2 to send traffic to all three because the cost is lower. 64 00:05:34,130 --> 00:05:38,360 Based on this formula ten to the eight divided by bandwidth. 65 00:05:38,960 --> 00:05:43,730 So the default formula used for costing is ten to the eight divided by bandwidth. 66 00:05:44,240 --> 00:05:50,120 Now you can change the reference bandwidth by using this command within the OSPF process or the cost 67 00:05:50,120 --> 00:05:54,670 reference bandwidth and then specifying a value in megabits per second. 68 00:05:54,680 --> 00:05:56,120 The default is 100. 69 00:05:56,150 --> 00:05:59,390 In other words, the reference bandwidth is 100 megabits per second. 70 00:05:59,750 --> 00:06:03,140 100 megabits per second has a cost in OSPF of one. 71 00:06:03,590 --> 00:06:09,710 If you change this to 1000 as an example, a gigabit link would be seen as the reference bandwidth. 72 00:06:10,100 --> 00:06:15,740 Or if you change it to 100,100 gig link would be used as a reference bandwidth. 73 00:06:15,950 --> 00:06:22,190 You need to do this command on all routers where you have interfaces with bandwidths greater than fast 74 00:06:22,190 --> 00:06:22,910 Ethernet. 75 00:06:23,360 --> 00:06:26,480 You can also change the cost on an interface. 76 00:06:26,600 --> 00:06:33,230 So rather than OSPF calculating the cost of the link based on the bandwidth, you can set the OSPF cost 77 00:06:33,230 --> 00:06:37,790 by tapping the command IP OSPF cost and then specifying a value. 78 00:06:38,450 --> 00:06:44,530 It's very important that the bandwidth statements on your router interfaces be configured correctly 79 00:06:44,540 --> 00:06:48,140 because that's what OSPF deems the speed of the link to be. 80 00:06:48,680 --> 00:06:50,810 Your copy does the same thing. 81 00:06:50,840 --> 00:06:56,930 It uses bandwidth as part of the calculation when working out the shortest route to the destination.