1 00:00:00,000 --> 00:00:00,994 2 00:00:00,994 --> 00:00:04,473 [AUDIO LOGO] 3 00:00:04,473 --> 00:00:11,950 4 00:00:11,950 --> 00:00:15,460 So we'll start off here on Trouble Ticket 1. 5 00:00:15,460 --> 00:00:17,470 And on Trouble Ticket 1, we're going 6 00:00:17,470 --> 00:00:20,200 to be using BGP Topology 1. 7 00:00:20,200 --> 00:00:22,600 Actually, all of these tickets are 8 00:00:22,600 --> 00:00:25,600 going to use BGP Topology 1, so we 9 00:00:25,600 --> 00:00:27,650 won't point it out every time. 10 00:00:27,650 --> 00:00:29,260 But that is the topology that you 11 00:00:29,260 --> 00:00:31,030 want to have convenient when we're 12 00:00:31,030 --> 00:00:32,650 going through these tickets. 13 00:00:32,650 --> 00:00:35,560 So what's the actual essence of this ticket? 14 00:00:35,560 --> 00:00:38,650 Well, the tickets going to be listed down here at the bottom. 15 00:00:38,650 --> 00:00:42,430 And the ticket is that router 5 is supposed 16 00:00:42,430 --> 00:00:46,990 to have a neighborship to router 1 and router 2, 17 00:00:46,990 --> 00:00:49,270 but they are down. 18 00:00:49,270 --> 00:00:55,690 So router 5 is supposed to be peered up with R1 and R2. 19 00:00:55,690 --> 00:00:58,450 But according to our ticket, this isn't working. 20 00:00:58,450 --> 00:01:00,700 Now, another point of information 21 00:01:00,700 --> 00:01:03,310 that you would hopefully be able to get out of documentation 22 00:01:03,310 --> 00:01:06,460 if you didn't know this, but in our environment, 23 00:01:06,460 --> 00:01:10,870 both of these routers are in fact route reflectors. 24 00:01:10,870 --> 00:01:14,380 That's why router 5 would only be peering with router 1 25 00:01:14,380 --> 00:01:15,670 and router 2. 26 00:01:15,670 --> 00:01:18,050 So as we get in and look at this, 27 00:01:18,050 --> 00:01:21,250 we'll see that it's not neighbors with routers 3 and 4 28 00:01:21,250 --> 00:01:24,590 either, but it's also not supposed to be. 29 00:01:24,590 --> 00:01:26,980 So let's go ahead and jump in. 30 00:01:26,980 --> 00:01:31,040 We'll, of course, always start with verifying the ticket. 31 00:01:31,040 --> 00:01:34,480 So that means we start on router 5. 32 00:01:34,480 --> 00:01:38,780 So let's start off with the good old show ip bgp summary. 33 00:01:38,780 --> 00:01:42,580 And what we see is router 1 is actually 34 00:01:42,580 --> 00:01:44,680 listed as being active. 35 00:01:44,680 --> 00:01:48,160 Router 2 is listed as being idle. 36 00:01:48,160 --> 00:01:52,950 If we do the same command again, we see that they now flip-flop. 37 00:01:52,950 --> 00:01:55,110 So both of these routers are cycling 38 00:01:55,110 --> 00:01:58,580 between active and idle. 39 00:01:58,580 --> 00:02:01,970 Now, one of the first things we want to check with BGP 40 00:02:01,970 --> 00:02:04,070 is that we have reachability. 41 00:02:04,070 --> 00:02:07,920 Now, we know that we have a route to the destination, 42 00:02:07,920 --> 00:02:11,585 in this case, 10.1.1.1 and 10.2.2.2. 43 00:02:11,585 --> 00:02:14,330 The reason we know that is it wouldn't even 44 00:02:14,330 --> 00:02:18,230 be trying to go active if it didn't have the route. 45 00:02:18,230 --> 00:02:21,150 So we could verify the route if we like, 46 00:02:21,150 --> 00:02:23,900 but we should already know that it has one, 47 00:02:23,900 --> 00:02:27,150 or again, it wouldn't even try to go active. 48 00:02:27,150 --> 00:02:30,050 It would be sitting in idle the whole time, 49 00:02:30,050 --> 00:02:33,570 but that doesn't mean we actually have reachability. 50 00:02:33,570 --> 00:02:37,150 So that should be probably one of the first things we check. 51 00:02:37,150 --> 00:02:39,750 Let's do a ping 10.1.1.1. 52 00:02:39,750 --> 00:02:42,300 And we'll go ahead and source that from our loopback 53 00:02:42,300 --> 00:02:45,360 since these are supposed to be peering using loopbacks. 54 00:02:45,360 --> 00:02:47,310 We see that that's successful. 55 00:02:47,310 --> 00:02:50,390 56 00:02:50,390 --> 00:02:52,790 And we can also ping, too. 57 00:02:52,790 --> 00:02:55,890 So we have connections between the devices. 58 00:02:55,890 --> 00:02:57,860 This is not a layer 1, a layer 2, 59 00:02:57,860 --> 00:03:00,740 or a layer issue from the perspective 60 00:03:00,740 --> 00:03:03,360 of at least being able to ping. 61 00:03:03,360 --> 00:03:05,240 So now would probably be a good idea 62 00:03:05,240 --> 00:03:07,770 to go look at what the other routers are seeing here. 63 00:03:07,770 --> 00:03:09,650 So let's go to router 1. 64 00:03:09,650 --> 00:03:12,500 We'll do our show ip bgp summary here. 65 00:03:12,500 --> 00:03:17,420 And in relation to router 5, we have an idle. 66 00:03:17,420 --> 00:03:19,790 So let's check its configuration. 67 00:03:19,790 --> 00:03:22,880 Let's just do a show on router bgp. 68 00:03:22,880 --> 00:03:25,760 Looking at router 5 here, we see that it's 69 00:03:25,760 --> 00:03:28,820 using the neighbor-group RRC, which stands 70 00:03:28,820 --> 00:03:30,690 for Route Reflector Client. 71 00:03:30,690 --> 00:03:33,140 And if we look at our neighbor-group RRC, 72 00:03:33,140 --> 00:03:35,660 it does appear to be configured properly. 73 00:03:35,660 --> 00:03:37,640 It's configuring the remote-as. 74 00:03:37,640 --> 00:03:40,460 We have our update source set to Loopback0. 75 00:03:40,460 --> 00:03:42,260 And then under our address family, 76 00:03:42,260 --> 00:03:45,020 it is configured with a Route Reflector Client 77 00:03:45,020 --> 00:03:47,010 and next-hop-self. 78 00:03:47,010 --> 00:03:50,580 So fundamentally, this looks correct. 79 00:03:50,580 --> 00:03:53,150 Let's go ahead and verify router 5's configuration, 80 00:03:53,150 --> 00:03:55,430 as well, since we didn't look at that. 81 00:03:55,430 --> 00:03:58,310 We'll just do a show run section bgp. 82 00:03:58,310 --> 00:04:01,130 This is a fairly straightforward configuration. 83 00:04:01,130 --> 00:04:02,960 It is definitely set to be neighbored 84 00:04:02,960 --> 00:04:06,050 with 10.1.1.1 and 10.2.2.2. 85 00:04:06,050 --> 00:04:08,900 They both have the remote as configured properly, 86 00:04:08,900 --> 00:04:11,690 and they're both set to Loopback0. 87 00:04:11,690 --> 00:04:13,970 So far, everything's looking good. 88 00:04:13,970 --> 00:04:15,630 So why don't we do this? 89 00:04:15,630 --> 00:04:18,829 Why don't we jump into Wireshark and take a look 90 00:04:18,829 --> 00:04:22,590 if these devices are actually communicating with each other? 91 00:04:22,590 --> 00:04:26,720 So here's a packet capture of the traffic on that Ethernet 92 00:04:26,720 --> 00:04:28,100 segment. 93 00:04:28,100 --> 00:04:29,720 And here, we have a packet. 94 00:04:29,720 --> 00:04:32,810 Notice that it's actually coming from router 5's 95 00:04:32,810 --> 00:04:40,130 physical address 172.16.35.5 going to router 1. 96 00:04:40,130 --> 00:04:44,120 And also, notice that it's an ICMP message. 97 00:04:44,120 --> 00:04:46,190 Now, the reason this is important 98 00:04:46,190 --> 00:04:50,300 is look over under the Info section here, 99 00:04:50,300 --> 00:04:54,470 it says communication administratively filtered. 100 00:04:54,470 --> 00:04:56,630 So we're actually getting a message 101 00:04:56,630 --> 00:05:01,620 from router 5 telling us that it's rejecting our traffic. 102 00:05:01,620 --> 00:05:04,430 So every time that router 1 tries 103 00:05:04,430 --> 00:05:08,120 to send traffic to router 5 and vice versa, 104 00:05:08,120 --> 00:05:09,960 notice what's happening. 105 00:05:09,960 --> 00:05:12,080 We're getting a lot of send messages. 106 00:05:12,080 --> 00:05:14,430 We're getting SYN-ACK messages. 107 00:05:14,430 --> 00:05:16,880 And then we're getting communication administratively 108 00:05:16,880 --> 00:05:17,660 filtered. 109 00:05:17,660 --> 00:05:20,930 So something is stopping this communication 110 00:05:20,930 --> 00:05:24,650 from forming a TCP three-way handshake. 111 00:05:24,650 --> 00:05:27,050 One side is going SYN-SYN-ACK. 112 00:05:27,050 --> 00:05:30,890 The other side is just going SYN and staying there. 113 00:05:30,890 --> 00:05:33,050 It's never even getting a SYN-ACK, 114 00:05:33,050 --> 00:05:35,810 instead it's getting that the communication is 115 00:05:35,810 --> 00:05:37,430 administratively filtered. 116 00:05:37,430 --> 00:05:42,282 This sounds a lot like an ACL. 117 00:05:42,282 --> 00:05:46,730 So let's jump into router 5 and see if it does in fact 118 00:05:46,730 --> 00:05:49,070 have any ACLs configured on that router 119 00:05:49,070 --> 00:05:52,970 because it really looks like this traffic is being blocked 120 00:05:52,970 --> 00:05:54,650 based on our Wireshark output. 121 00:05:54,650 --> 00:05:57,200 122 00:05:57,200 --> 00:06:00,310 Let's do a show ip interface gigabit 1. 123 00:06:00,310 --> 00:06:04,780 And notice right here, we have an inbound access list 124 00:06:04,780 --> 00:06:07,270 called ACL filter. 125 00:06:07,270 --> 00:06:09,680 Let's take a look at that access list. 126 00:06:09,680 --> 00:06:11,830 If we look at this access list, we 127 00:06:11,830 --> 00:06:13,450 see that it's permitting some things, 128 00:06:13,450 --> 00:06:18,830 like web traffic, FTP traffic, SMTP traffic, and so on. 129 00:06:18,830 --> 00:06:22,810 However, what it's not doing is permitting BGP. 130 00:06:22,810 --> 00:06:24,610 It's denying everything else. 131 00:06:24,610 --> 00:06:26,500 And based on the number of matches 132 00:06:26,500 --> 00:06:29,950 we have there particularly in a lab, that would probably 133 00:06:29,950 --> 00:06:32,980 indicate that we're getting a lot of denied traffic 134 00:06:32,980 --> 00:06:34,520 on this interface. 135 00:06:34,520 --> 00:06:37,420 We also actually want to see if this access list is applied 136 00:06:37,420 --> 00:06:42,040 to gigabit 2 because router 5 has two interfaces that it 137 00:06:42,040 --> 00:06:43,670 could be talking on. 138 00:06:43,670 --> 00:06:45,400 But as we can see, that access list 139 00:06:45,400 --> 00:06:48,530 is applied inbound on this interface, as well. 140 00:06:48,530 --> 00:06:52,190 So this definitely appears to be the problem. 141 00:06:52,190 --> 00:06:55,450 So it does appear that we may have actually found 142 00:06:55,450 --> 00:06:57,170 an isolated the problem. 143 00:06:57,170 --> 00:07:00,530 So then the question always comes up, how should we fix it? 144 00:07:00,530 --> 00:07:02,480 Well, what are our choices? 145 00:07:02,480 --> 00:07:06,460 Choice number one, we could definitely just remove the ACL. 146 00:07:06,460 --> 00:07:08,560 That would certainly solve the problem 147 00:07:08,560 --> 00:07:10,840 assuming the ACL is the problem. 148 00:07:10,840 --> 00:07:14,500 We could modify the ACL to actually support 149 00:07:14,500 --> 00:07:18,920 BGP to allow that traffic, or in this case, 150 00:07:18,920 --> 00:07:22,180 we could even just change the direction of the ACL. 151 00:07:22,180 --> 00:07:24,730 It's applied inbound right now. 152 00:07:24,730 --> 00:07:27,940 If it was applied outbound, it wouldn't filter 153 00:07:27,940 --> 00:07:29,710 the router's own traffic. 154 00:07:29,710 --> 00:07:32,770 But that would also probably fundamentally change 155 00:07:32,770 --> 00:07:34,600 what the ACL is there for. 156 00:07:34,600 --> 00:07:38,330 So from a troubleshooting and design perspective, 157 00:07:38,330 --> 00:07:40,930 I would look at this and say, well, obviously, 158 00:07:40,930 --> 00:07:42,580 the ACL is there for a reason. 159 00:07:42,580 --> 00:07:43,930 Somebody wanted it there. 160 00:07:43,930 --> 00:07:46,570 It's supposed to be filtering traffic in the direction 161 00:07:46,570 --> 00:07:48,100 that it's doing so. 162 00:07:48,100 --> 00:07:51,500 So the best fix, at least I think so, 163 00:07:51,500 --> 00:07:56,540 would be to add BGP to the ACL to allow it to come through. 164 00:07:56,540 --> 00:07:58,250 So that's what we're going to do. 165 00:07:58,250 --> 00:08:01,163 Let's go ahead and update that ACL. 166 00:08:01,163 --> 00:08:03,830 So first, let's just get a quick reminder of what it looks like. 167 00:08:03,830 --> 00:08:05,770 So this is the current status. 168 00:08:05,770 --> 00:08:08,270 We'll go into Configuration mode. 169 00:08:08,270 --> 00:08:10,090 We'll go into our access list. 170 00:08:10,090 --> 00:08:12,440 And then we'll create a new statement. 171 00:08:12,440 --> 00:08:15,260 And it really doesn't matter too much where we put this. 172 00:08:15,260 --> 00:08:17,540 However, it's a routing protocol. 173 00:08:17,540 --> 00:08:19,310 I would generally put it near the top. 174 00:08:19,310 --> 00:08:26,110 So we will say line 5 is going to be to permit tcp from anyone 175 00:08:26,110 --> 00:08:30,130 to anyone if it equals bgp. 176 00:08:30,130 --> 00:08:33,789 Now, one downside to this is if we do a show ip access 177 00:08:33,789 --> 00:08:38,260 list again, is now we have all of our numbers nice and neat 178 00:08:38,260 --> 00:08:41,500 by tens except for this 5 that we just 179 00:08:41,500 --> 00:08:43,640 had to stick at the top. 180 00:08:43,640 --> 00:08:46,090 So one cool feature we can use if we want to, 181 00:08:46,090 --> 00:08:47,560 and this is, of course, completely 182 00:08:47,560 --> 00:08:53,410 optional, we can say ip access-list sequence, 183 00:08:53,410 --> 00:08:56,530 put in the name, and then tell it what line number we 184 00:08:56,530 --> 00:08:58,160 want it to start with. 185 00:08:58,160 --> 00:09:00,520 So we'll say start with 10, and then 186 00:09:00,520 --> 00:09:02,680 what number you want it to increment by. 187 00:09:02,680 --> 00:09:04,420 We'll, again, say 10. 188 00:09:04,420 --> 00:09:08,150 And what this will do is remember the list 189 00:09:08,150 --> 00:09:11,560 so that when we look at it, it's now starting at 10 190 00:09:11,560 --> 00:09:13,690 and going through 70. 191 00:09:13,690 --> 00:09:18,040 By the way, also note that our tcp any any 192 00:09:18,040 --> 00:09:23,470 equals bgp line line 10 already has 31 matches. 193 00:09:23,470 --> 00:09:24,850 I'm going to guess. 194 00:09:24,850 --> 00:09:28,070 That means our neighborships are probably up. 195 00:09:28,070 --> 00:09:30,860 So let's do our show ip bgp summary again, 196 00:09:30,860 --> 00:09:32,620 and we can see that they've been up 197 00:09:32,620 --> 00:09:34,990 for about a minute and a half, give or take here. 198 00:09:34,990 --> 00:09:37,550 And we've learned eight prefixes. 199 00:09:37,550 --> 00:09:40,270 If we do a show ip bgp, we're actually 200 00:09:40,270 --> 00:09:43,510 getting prefixes coming from BGP. 201 00:09:43,510 --> 00:09:46,870 So this gives us verification that we have, in fact, 202 00:09:46,870 --> 00:09:49,230 solved this ticket. 203 00:09:49,230 --> 00:09:53,630 So in this video, we had a ticket involving BGP. 204 00:09:53,630 --> 00:09:55,700 But at the end of the day, this one 205 00:09:55,700 --> 00:09:58,730 may not have actually been directly related to BGP. 206 00:09:58,730 --> 00:10:01,850 It was just affecting BGP traffic, therefore, 207 00:10:01,850 --> 00:10:04,050 preventing our neighborships from coming up. 208 00:10:04,050 --> 00:10:07,190 But this is the kind of thing we have to worry about, 209 00:10:07,190 --> 00:10:10,040 particularly with inbound ACLs. 210 00:10:10,040 --> 00:10:13,740 We have to make sure we account for routing protocols. 211 00:10:13,740 --> 00:10:16,430 Now, one thing you might note, and you 212 00:10:16,430 --> 00:10:18,870 may have to go back and look at the ACL again, 213 00:10:18,870 --> 00:10:25,190 but it did not allow IS-IS, which is actually our IGP here. 214 00:10:25,190 --> 00:10:30,620 Just a quick note on that, IS-IS runs on CLNS. 215 00:10:30,620 --> 00:10:32,780 It doesn't run on IP. 216 00:10:32,780 --> 00:10:36,350 Therefore, an IP access-list is not 217 00:10:36,350 --> 00:10:39,330 going to effect IS-IS traffic. 218 00:10:39,330 --> 00:10:41,940 That's why we didn't have to allow it. 219 00:10:41,940 --> 00:10:45,390 But either way, this gets our BGP neighbors up and running, 220 00:10:45,390 --> 00:10:48,350 and this solves this ticket. 221 00:10:48,350 --> 00:10:50,510 I hope this has been informative for you, 222 00:10:50,510 --> 00:10:53,590 and I'd like to thank you for viewing. 223 00:10:53,590 --> 00:10:57,000