1 1 00:00:00,570 --> 00:00:03,378 Loops, in this lesson, we're going to discuss loops, 2 2 00:00:03,378 --> 00:00:06,970 specifically switching loops and routing loops. 3 3 00:00:06,970 --> 00:00:08,720 First we have switching loops. 4 4 00:00:08,720 --> 00:00:10,500 Switching loops, or bridging loops, 5 5 00:00:10,500 --> 00:00:12,660 are going to occur whenever there's more than one path 6 6 00:00:12,660 --> 00:00:15,040 between a source and destination device. 7 7 00:00:15,040 --> 00:00:17,100 As broadcast packets are set for switching 8 8 00:00:17,100 --> 00:00:18,810 through every single switch port, 9 9 00:00:18,810 --> 00:00:20,470 flooding can really occur quickly 10 10 00:00:20,470 --> 00:00:22,060 when broadcast messages are sent 11 11 00:00:22,060 --> 00:00:24,270 and then repeated back through another switch port 12 12 00:00:24,270 --> 00:00:25,870 in this looped architecture. 13 13 00:00:25,870 --> 00:00:28,360 This will create a broadcast storm for you. 14 14 00:00:28,360 --> 00:00:29,690 Since our modern networks are built 15 15 00:00:29,690 --> 00:00:31,920 with additional load balancing and fault tolerance, 16 16 00:00:31,920 --> 00:00:34,310 there's usually going to be multiple physical connections 17 17 00:00:34,310 --> 00:00:35,910 between each part of the network. 18 18 00:00:35,910 --> 00:00:38,030 And so naturally, we would have loops 19 19 00:00:38,030 --> 00:00:40,030 and switching loops would become inevitable 20 20 00:00:40,030 --> 00:00:42,700 if we didn't put some protections in place to prevent them. 21 21 00:00:42,700 --> 00:00:44,840 So, how are you going to prevent them? 22 22 00:00:44,840 --> 00:00:46,460 Well, to prevent a switching loop, 23 23 00:00:46,460 --> 00:00:49,950 you have to enable STP, the spanning tree protocol. 24 24 00:00:49,950 --> 00:00:52,860 To check if STP is enabled and configured on your switch, 25 25 00:00:52,860 --> 00:00:55,380 you need to use the command, show span tree, 26 26 00:00:55,380 --> 00:00:56,590 and then hit enter. 27 27 00:00:56,590 --> 00:00:58,910 For the network plus exam, you do not need to know 28 28 00:00:58,910 --> 00:01:01,570 how to configure a spanning tree protocol on a switch, 29 29 00:01:01,570 --> 00:01:02,990 but I'm showing you what it looks like 30 30 00:01:02,990 --> 00:01:05,110 just for demonstration purposes here. 31 31 00:01:05,110 --> 00:01:07,560 Now you can see here, the information for this switch 32 32 00:01:07,560 --> 00:01:10,920 in VLAN0001, which is our default VLAN. 33 33 00:01:10,920 --> 00:01:13,140 In this example, the route ID 34 34 00:01:13,140 --> 00:01:15,920 is set with a priority of one at a cost of two. 35 35 00:01:15,920 --> 00:01:20,610 The bridge ID is set with a priority of 61,441. 36 36 00:01:20,610 --> 00:01:22,900 As we look at the different interfaces on this switch, 37 37 00:01:22,900 --> 00:01:24,320 you can see there are two ports 38 38 00:01:24,320 --> 00:01:25,830 that are designated as the root port 39 39 00:01:25,830 --> 00:01:27,130 that will forward traffic. 40 40 00:01:27,130 --> 00:01:30,370 And all the other ports are set as designated ports. 41 41 00:01:30,370 --> 00:01:35,370 Eth1/11, 1/12, 1/15, and 1/16 are all in a blocking state. 42 42 00:01:37,267 --> 00:01:39,540 And this is what prevents a loop from occurring 43 43 00:01:39,540 --> 00:01:40,970 with this switch. 44 44 00:01:40,970 --> 00:01:43,590 Now the bottom line here, when it comes to switching loops, 45 45 00:01:43,590 --> 00:01:45,470 is that if you suspect a switching loop, 46 46 00:01:45,470 --> 00:01:48,230 it's likely an issue with how you configure your STP 47 47 00:01:48,230 --> 00:01:49,640 and you need to escalate this work 48 48 00:01:49,640 --> 00:01:51,900 to a network administrator or network engineer 49 49 00:01:51,900 --> 00:01:54,850 to troubleshoot it, reconfigure it and repair it. 50 50 00:01:54,850 --> 00:01:56,990 Next, let's talk about routing loops. 51 51 00:01:56,990 --> 00:01:59,040 A routing loop is formed when an error occurs 52 52 00:01:59,040 --> 00:02:01,120 in the operation of your routing algorithm. 53 53 00:02:01,120 --> 00:02:02,770 And this creates a circular route 54 54 00:02:02,770 --> 00:02:04,890 amongst a group of network devices. 55 55 00:02:04,890 --> 00:02:07,410 Routing loops are caused by incorrect configurations 56 56 00:02:07,410 --> 00:02:08,790 of your routing protocols, 57 57 00:02:08,790 --> 00:02:10,210 where data packets get sent 58 58 00:02:10,210 --> 00:02:12,120 between different hosts of different networks. 59 59 00:02:12,120 --> 00:02:13,960 And they get caught in this endless loop, 60 60 00:02:13,960 --> 00:02:16,600 traveling in a circle between the different network routers 61 61 00:02:16,600 --> 00:02:18,370 with incorrect route entries. 62 62 00:02:18,370 --> 00:02:20,170 Routing loops, unlike switching loops 63 63 00:02:20,170 --> 00:02:22,800 are not caused by physical circular connections, 64 64 00:02:22,800 --> 00:02:25,970 but instead by the logical layer 3 circular connections 65 65 00:02:25,970 --> 00:02:28,140 that can exist within your routing tables. 66 66 00:02:28,140 --> 00:02:30,490 Remember, we want multiple physical connections 67 67 00:02:30,490 --> 00:02:31,590 between our routers, 68 68 00:02:31,590 --> 00:02:34,510 because this gives us higher redundancy and fault tolerance. 69 69 00:02:34,510 --> 00:02:37,690 So our routing protocols have methods in place to protect us 70 70 00:02:37,690 --> 00:02:39,490 from physical loops causing issues. 71 71 00:02:39,490 --> 00:02:41,700 Things like weighted connections based on hops 72 72 00:02:41,700 --> 00:02:43,056 or the speed of those connections. 73 73 00:02:43,056 --> 00:02:45,940 For example, distance vector routing protocols, 74 74 00:02:45,940 --> 00:02:49,410 use a Time to Live, or TTL, in the data gram header 75 75 00:02:49,410 --> 00:02:50,700 of the IP packets. 76 76 00:02:50,700 --> 00:02:52,920 And this will help us avoid routing loops. 77 77 00:02:52,920 --> 00:02:55,500 So, if a packet reaches the Time to Live of zero, 78 78 00:02:55,500 --> 00:02:56,870 it's going to be dropped by the router 79 79 00:02:56,870 --> 00:02:58,090 and it's not going to be forwarded. 80 80 00:02:58,090 --> 00:03:00,480 And that effectively ends your routing loop. 81 81 00:03:00,480 --> 00:03:02,320 Now, another method to prevent routing loops 82 82 00:03:02,320 --> 00:03:04,030 is known as split-horizon. 83 83 00:03:04,030 --> 00:03:06,370 If you can figure split-horizon, you're going to ensure 84 84 00:03:06,370 --> 00:03:07,920 that you have this router configuration 85 85 00:03:07,920 --> 00:03:10,020 that stops a route from being advertised back 86 86 00:03:10,020 --> 00:03:11,750 in the direction that it came from. 87 87 00:03:11,750 --> 00:03:14,110 This split-horizon mechanism ensures that a router 88 88 00:03:14,110 --> 00:03:16,620 can not send back alert route to the same router 89 89 00:03:16,620 --> 00:03:17,920 that it learned it from. 90 90 00:03:17,920 --> 00:03:20,610 To setup split-horizon on a Cisco router, for example, 91 91 00:03:20,610 --> 00:03:23,830 you're simply going to enter the command ip split-horizon 92 92 00:03:23,830 --> 00:03:25,980 and hit enter at the command line interface. 93 93 00:03:25,980 --> 00:03:28,670 If you believe split-horizon has been disabled by accident, 94 94 00:03:28,670 --> 00:03:31,640 you can look to see if the no ip split-horizon command 95 95 00:03:31,640 --> 00:03:33,250 was previously issued. 96 96 00:03:33,250 --> 00:03:35,560 Now, route poisoning is another method we can use 97 97 00:03:35,560 --> 00:03:38,030 to avoid routing loops inside of our networks. 98 98 00:03:38,030 --> 00:03:40,080 If a router detects that one of its connected routes 99 99 00:03:40,080 --> 00:03:41,970 has failed, the router is going to poison 100 100 00:03:41,970 --> 00:03:43,800 that route by increasing its metric 101 101 00:03:43,800 --> 00:03:45,530 to an infinitely high number. 102 102 00:03:45,530 --> 00:03:47,810 This happens automatically inside your routers. 103 103 00:03:47,810 --> 00:03:50,200 So there really isn't something you have to do or configure, 104 104 00:03:50,200 --> 00:03:52,010 it'll just happen for you. 105 105 00:03:52,010 --> 00:03:54,190 Finally, we have hold-down timers. 106 106 00:03:54,190 --> 00:03:56,470 Hold-down timers are used to prevent bad routes 107 107 00:03:56,470 --> 00:03:57,730 from being restored and passed 108 108 00:03:57,730 --> 00:03:59,420 to other routers by accident. 109 109 00:03:59,420 --> 00:04:00,850 Now hold-down timers are used 110 110 00:04:00,850 --> 00:04:03,490 with distant vector routing protocols like RIP. 111 111 00:04:03,490 --> 00:04:04,640 The router is going to be configured, 112 112 00:04:04,640 --> 00:04:07,130 so that will not advertise or accept any routes 113 113 00:04:07,130 --> 00:04:08,940 that are in a hold-down state. 114 114 00:04:08,940 --> 00:04:12,100 This occurs for a set period known as the hold-down period. 115 115 00:04:12,100 --> 00:04:15,590 By default, the hold-down timer is set at 180 seconds 116 116 00:04:15,590 --> 00:04:17,590 or three minutes when you're using RIP 117 117 00:04:17,590 --> 00:04:18,920 as your routing protocol. 118 118 00:04:18,920 --> 00:04:20,240 It should be configured that way, 119 119 00:04:20,240 --> 00:04:22,080 automatically, by your router. 120 120 00:04:22,080 --> 00:04:24,300 As I said, most of the routing loop issues 121 121 00:04:24,300 --> 00:04:26,708 can be solved by simply using the right routing protocols, 122 122 00:04:26,708 --> 00:04:28,830 ensuring they're configured properly. 123 123 00:04:28,830 --> 00:04:31,190 If you're adding a bunch of static routes into your router, 124 124 00:04:31,190 --> 00:04:33,940 be really careful because this is how most routing loops 125 125 00:04:33,940 --> 00:04:35,170 are going to be created. 126 126 00:04:35,170 --> 00:04:37,640 Remember, statically-created routes are given a metric 127 127 00:04:37,640 --> 00:04:41,010 of 1 by default, making it extremely, highly trusted 128 128 00:04:41,010 --> 00:04:42,040 by the router. 129 129 00:04:42,040 --> 00:04:44,250 The only type of route that the router will trust more 130 130 00:04:44,250 --> 00:04:45,432 than your statically assigned route 131 131 00:04:45,432 --> 00:04:47,700 is a directly connected route. 132 132 00:04:47,700 --> 00:04:51,380 Any routes that it learns through OSPF or RIP or BGP 133 133 00:04:51,380 --> 00:04:54,260 or EIGRP or any other routing protocol 134 134 00:04:54,260 --> 00:04:56,930 will not overwrite that static route that you created. 135 135 00:04:56,930 --> 00:04:58,550 So you have to be careful with static routes 136 136 00:04:58,550 --> 00:05:00,800 in order to prevent a potential routing loop.