1 00:00:00,000 --> 00:00:03,824 [AUDIO LOGO] 2 00:00:03,824 --> 00:00:07,180 3 00:00:07,180 --> 00:00:10,730 So now that we understand the IS-IS hierarchy, 4 00:00:10,730 --> 00:00:14,710 now let's jump in and look at some good design practices 5 00:00:14,710 --> 00:00:17,650 that we can use with this both with our IP 6 00:00:17,650 --> 00:00:21,310 addressing and using this layer 2 hierarchy. 7 00:00:21,310 --> 00:00:25,500 Let's jump in and take a look at the details. 8 00:00:25,500 --> 00:00:27,110 So one of the primary things we want 9 00:00:27,110 --> 00:00:30,680 to be concerned with really with any protocol design-- 10 00:00:30,680 --> 00:00:34,220 this isn't necessarily specific for IS-IS-- 11 00:00:34,220 --> 00:00:38,240 but we need to have good addressing design. 12 00:00:38,240 --> 00:00:43,260 And this is really because we want to summarize. 13 00:00:43,260 --> 00:00:46,250 We don't want all of the routers and all of our network 14 00:00:46,250 --> 00:00:49,340 to know about all the routes, and therefore, we 15 00:00:49,340 --> 00:00:50,750 need to summarize. 16 00:00:50,750 --> 00:00:54,650 Now with a link-state protocol, it's 17 00:00:54,650 --> 00:00:58,370 important that we have a good design for our addressing 18 00:00:58,370 --> 00:01:02,300 to follow the protocol because we can only really 19 00:01:02,300 --> 00:01:06,210 summarize at these hierarchical boundaries. 20 00:01:06,210 --> 00:01:11,300 So, for example, in OSPF, we can summarize between areas. 21 00:01:11,300 --> 00:01:14,810 And really, in IS-IS, it's the same thing. 22 00:01:14,810 --> 00:01:17,960 We can summarize going from Level 2 into Level 1 23 00:01:17,960 --> 00:01:19,880 and Level 1 into Level 2. 24 00:01:19,880 --> 00:01:23,280 We can't summarize within the areas. 25 00:01:23,280 --> 00:01:25,430 That's why it becomes important here 26 00:01:25,430 --> 00:01:28,880 to make sure that our IP addressing design follows 27 00:01:28,880 --> 00:01:30,300 our network design. 28 00:01:30,300 --> 00:01:34,400 Now with IS-IS, we do have one extra thing we 29 00:01:34,400 --> 00:01:36,350 need to be concerned with, and that 30 00:01:36,350 --> 00:01:39,800 is to plan out our CLNP addressing. 31 00:01:39,800 --> 00:01:42,380 Now we'll be talking more about the addressing 32 00:01:42,380 --> 00:01:45,020 itself in an upcoming video here, 33 00:01:45,020 --> 00:01:47,480 but for now, just remember that this does need 34 00:01:47,480 --> 00:01:49,320 to be part of our planning. 35 00:01:49,320 --> 00:01:52,700 It's not a difficult task, but it does need to be planned out. 36 00:01:52,700 --> 00:01:56,420 Now again, we're almost always, except in the smallest 37 00:01:56,420 --> 00:01:59,900 of environments, we're going to want to use 38 00:01:59,900 --> 00:02:01,800 this two-layer hierarchy. 39 00:02:01,800 --> 00:02:04,680 And as we discussed in the previous video, 40 00:02:04,680 --> 00:02:08,509 this, of course, is to limit that LSP flooding, 41 00:02:08,509 --> 00:02:11,240 and it also, of course, as I just mentioned, 42 00:02:11,240 --> 00:02:14,840 gives us a place to actually do that summarization. 43 00:02:14,840 --> 00:02:18,080 And that's why it's so important that our IP addressing 44 00:02:18,080 --> 00:02:21,680 design follow our layer 2 hierarchy so 45 00:02:21,680 --> 00:02:26,060 that we can summarize at the places that we're able to. 46 00:02:26,060 --> 00:02:28,430 Now it's quite common in networking 47 00:02:28,430 --> 00:02:32,010 to use a three-layer architecture. 48 00:02:32,010 --> 00:02:35,600 Now, this isn't exactly the same as the hierarchy. 49 00:02:35,600 --> 00:02:38,000 The architecture is how we're going to be building out 50 00:02:38,000 --> 00:02:38,963 the physical network. 51 00:02:38,963 --> 00:02:40,380 So what we're going to do is we're 52 00:02:40,380 --> 00:02:44,030 going to take a look at how the three-layer architecture goes 53 00:02:44,030 --> 00:02:48,840 with the layer 2 hierarchy in a typical design. 54 00:02:48,840 --> 00:02:51,290 So just to go over a design here a little bit. 55 00:02:51,290 --> 00:02:54,680 As we mentioned, the devices should 56 00:02:54,680 --> 00:02:59,030 be grouped into areas based on the IP address and scheme 57 00:02:59,030 --> 00:03:02,360 again because of where we need to do our summarization. 58 00:03:02,360 --> 00:03:05,940 And let's talk about how this fits into the architecture. 59 00:03:05,940 --> 00:03:09,050 So typically, we will have a core layer. 60 00:03:09,050 --> 00:03:11,720 In this case, the core layer is going 61 00:03:11,720 --> 00:03:14,580 to be these routers up here at the top. 62 00:03:14,580 --> 00:03:18,990 These are going to be acting as the backbone area. 63 00:03:18,990 --> 00:03:21,960 Now again, as we mentioned in our previous video, 64 00:03:21,960 --> 00:03:24,290 you don't have to have a dedicated backbone 65 00:03:24,290 --> 00:03:26,300 area in IS-IS. 66 00:03:26,300 --> 00:03:30,920 However, for good scalability, it's generally a best practice. 67 00:03:30,920 --> 00:03:34,580 These are generally going to be level two-only devices. 68 00:03:34,580 --> 00:03:37,700 As you can see in our Area 100 here, 69 00:03:37,700 --> 00:03:41,120 we're only using these Level 2 connections up 70 00:03:41,120 --> 00:03:42,920 here in Area 100. 71 00:03:42,920 --> 00:03:47,040 All of the other areas would then connect to this area. 72 00:03:47,040 --> 00:03:50,150 In this case, we're only picturing Area 10. 73 00:03:50,150 --> 00:03:51,920 But if there were more areas, they 74 00:03:51,920 --> 00:03:54,470 would just be connected in here to Area 100, 75 00:03:54,470 --> 00:03:56,160 acting as our backbone. 76 00:03:56,160 --> 00:03:58,460 And again, this is just a design decision. 77 00:03:58,460 --> 00:04:00,290 As we've mentioned, you do not have 78 00:04:00,290 --> 00:04:05,300 to have a dedicated backbone area in IS-IS. 79 00:04:05,300 --> 00:04:07,820 Then, of course, we have the distribution layer. 80 00:04:07,820 --> 00:04:10,670 In our design, that would be these routers right here. 81 00:04:10,670 --> 00:04:13,850 These are the routers that, of course, connect up 82 00:04:13,850 --> 00:04:14,990 to the core layer. 83 00:04:14,990 --> 00:04:17,029 We can see that in our diagram. 84 00:04:17,029 --> 00:04:21,410 These are going to be using Level 2 to go to the core 85 00:04:21,410 --> 00:04:25,220 but Level 1 inside of the area. 86 00:04:25,220 --> 00:04:28,850 We can, of course, see here that Level 1 is all 87 00:04:28,850 --> 00:04:30,500 of these connections down here. 88 00:04:30,500 --> 00:04:33,470 And to go back to our design and the whole idea 89 00:04:33,470 --> 00:04:36,630 that the IP addressing needs to follow our design, 90 00:04:36,630 --> 00:04:39,860 this is where we would do that summarization. 91 00:04:39,860 --> 00:04:42,410 This is the boundary between the Level 1 92 00:04:42,410 --> 00:04:46,170 and the Level 2 environments for this particular area. 93 00:04:46,170 --> 00:04:50,190 And again, if we had more than Area 10, this, of course, 94 00:04:50,190 --> 00:04:53,240 would be the same in all the other areas. 95 00:04:53,240 --> 00:04:56,990 Last but not least, of course, we have the access layer. 96 00:04:56,990 --> 00:04:59,030 That would, of course, be these devices 97 00:04:59,030 --> 00:05:00,510 down here at the bottom. 98 00:05:00,510 --> 00:05:03,950 And this is where all of the other devices 99 00:05:03,950 --> 00:05:06,720 connect in to the rest of our network. 100 00:05:06,720 --> 00:05:10,770 So this is the connection point for everything else. 101 00:05:10,770 --> 00:05:16,220 So this is where PCs, printers, servers, everything, 102 00:05:16,220 --> 00:05:19,190 connections to other routing domains. 103 00:05:19,190 --> 00:05:22,750 Everything comes in at the access layer. 104 00:05:22,750 --> 00:05:26,900 Keep in mind that this, of course, is not set in stone. 105 00:05:26,900 --> 00:05:29,650 But this is a very typical design 106 00:05:29,650 --> 00:05:34,150 to use our two-tier hierarchy in IS-IS 107 00:05:34,150 --> 00:05:37,900 in a three-tier architecture for our network, 108 00:05:37,900 --> 00:05:40,630 where we have core distribution and access. 109 00:05:40,630 --> 00:05:43,210 110 00:05:43,210 --> 00:05:47,800 Just some final things to keep in mind in your IS-IS designs. 111 00:05:47,800 --> 00:05:50,620 Most large environments are going 112 00:05:50,620 --> 00:05:53,770 to be using those wide metrics that we discussed. 113 00:05:53,770 --> 00:05:58,300 Remember, they're required for IPv6 anyway. 114 00:05:58,300 --> 00:06:00,730 So again, as I mentioned when we talked about them 115 00:06:00,730 --> 00:06:02,830 the first time, you're probably going 116 00:06:02,830 --> 00:06:05,387 to want wide metrics in any case. 117 00:06:05,387 --> 00:06:06,970 So you probably just want to turn them 118 00:06:06,970 --> 00:06:08,260 on right from the beginning. 119 00:06:08,260 --> 00:06:09,700 Other considerations. 120 00:06:09,700 --> 00:06:16,180 IOS does not dynamically set the metric on the interfaces. 121 00:06:16,180 --> 00:06:20,320 All of these interfaces actually default to a cost of 10, 122 00:06:20,320 --> 00:06:24,440 which unfortunately makes it act a little bit like a hop count. 123 00:06:24,440 --> 00:06:27,520 So if you do have varying speeds in your environment 124 00:06:27,520 --> 00:06:30,970 for your interfaces, which most environments do, 125 00:06:30,970 --> 00:06:33,610 you're going to have to set this manually. 126 00:06:33,610 --> 00:06:36,890 And when I say you, I, of course, mean the administrator. 127 00:06:36,890 --> 00:06:39,280 So this does need to be manually configured 128 00:06:39,280 --> 00:06:42,250 on most of our IOS devices. 129 00:06:42,250 --> 00:06:46,000 One final thing to consider with our IS-IS designs 130 00:06:46,000 --> 00:06:51,730 is IS-IS does use the LSPs very efficiently. 131 00:06:51,730 --> 00:06:54,460 What that ultimately means is this 132 00:06:54,460 --> 00:06:59,140 allows us to have around 1,000 routers in an area 133 00:06:59,140 --> 00:07:00,970 without too much trouble. 134 00:07:00,970 --> 00:07:05,200 This would actually be pushing the limits in an OSPF design. 135 00:07:05,200 --> 00:07:09,340 But since IS-IS does handle this a little bit more efficiently, 136 00:07:09,340 --> 00:07:12,160 it does make this a little more scalable 137 00:07:12,160 --> 00:07:15,820 than OSPF, which, again, is one of the reasons we 138 00:07:15,820 --> 00:07:20,520 like this in a service provider environment. 139 00:07:20,520 --> 00:07:23,030 In this video, we jumped in, and we 140 00:07:23,030 --> 00:07:26,510 looked at a little bit of IS-IS design. 141 00:07:26,510 --> 00:07:29,000 Specifically, what we focused on is 142 00:07:29,000 --> 00:07:34,370 how we take a two-tier hierarchy like IS-IS uses, 143 00:07:34,370 --> 00:07:39,650 and we fit that into our three-tier network architecture 144 00:07:39,650 --> 00:07:42,770 that we generally use in our networking designs 145 00:07:42,770 --> 00:07:46,190 where we have core distribution and access. 146 00:07:46,190 --> 00:07:49,590 And when we overlay IS-IS on top of that, 147 00:07:49,590 --> 00:07:54,320 we end up with the core usually acting as a backbone area. 148 00:07:54,320 --> 00:07:57,050 Although, again, just one final reminder. 149 00:07:57,050 --> 00:07:59,990 IS-IS does not need a backbone area, 150 00:07:59,990 --> 00:08:03,920 but we do typically use the core as a backbone. 151 00:08:03,920 --> 00:08:06,360 Then we have our distribution layer, 152 00:08:06,360 --> 00:08:09,620 which is where we do our summarization, acting 153 00:08:09,620 --> 00:08:14,900 as the border between our Level 1 and our Level 2 routing. 154 00:08:14,900 --> 00:08:17,420 Then, of course, we have the access layer 155 00:08:17,420 --> 00:08:22,670 where everything else connects in to the IS-IS environment. 156 00:08:22,670 --> 00:08:24,880 I hope this has been informative for you, 157 00:08:24,880 --> 00:08:28,050 and I'd like to thank you for viewing. 158 00:08:28,050 --> 00:08:32,000