1 1 00:00:00,360 --> 00:00:01,560 As you might have guessed, 2 2 00:00:01,560 --> 00:00:04,440 packet captures can take up a lot of storage space 3 3 00:00:04,440 --> 00:00:06,720 and they can grow quickly in size. 4 4 00:00:06,720 --> 00:00:09,240 For example, if I'm conducting full packet capture 5 5 00:00:09,240 --> 00:00:10,980 on my home network each day, 6 6 00:00:10,980 --> 00:00:12,870 I would need several gigabytes of storage 7 7 00:00:12,870 --> 00:00:14,430 just for my small family 8 8 00:00:14,430 --> 00:00:17,010 because every single packet that goes in and out of my house 9 9 00:00:17,010 --> 00:00:18,180 would be captured. 10 10 00:00:18,180 --> 00:00:20,430 Every video game my son is playing online, 11 11 00:00:20,430 --> 00:00:22,200 every YouTube video he watches, 12 12 00:00:22,200 --> 00:00:24,420 every Netflix show my wife binges, 13 13 00:00:24,420 --> 00:00:26,880 all of that would be captured bit by bit. 14 14 00:00:26,880 --> 00:00:29,400 This is known as a full packet capture. 15 15 00:00:29,400 --> 00:00:31,890 Now, a full packet capture or FPC 16 16 00:00:31,890 --> 00:00:33,930 is going to capture the entire packet. 17 17 00:00:33,930 --> 00:00:35,820 This includes the header and the payload 18 18 00:00:35,820 --> 00:00:39,000 for all traffic that is entering or leaving your network. 19 19 00:00:39,000 --> 00:00:41,520 As I said, this would be a ton of information 20 20 00:00:41,520 --> 00:00:43,710 and quickly eat up all of our storage. 21 21 00:00:43,710 --> 00:00:45,210 Now, because full packet capture 22 22 00:00:45,210 --> 00:00:47,160 takes up so much storage space, 23 23 00:00:47,160 --> 00:00:49,140 a lot of times we won't collect it. 24 24 00:00:49,140 --> 00:00:52,290 Instead, we'll use something called flow analysis, 25 25 00:00:52,290 --> 00:00:54,420 and this relies on a flow collector. 26 26 00:00:54,420 --> 00:00:56,520 Now, a flow collector is a means of recording 27 27 00:00:56,520 --> 00:00:59,340 metadata and statistics about network traffic 28 28 00:00:59,340 --> 00:01:01,380 rather than recording each and every frame 29 29 00:01:01,380 --> 00:01:02,940 that's going over our network. 30 30 00:01:02,940 --> 00:01:05,280 This allows us to use flow analysis tools 31 31 00:01:05,280 --> 00:01:07,170 that provide network traffic statistics 32 32 00:01:07,170 --> 00:01:08,700 sampled by the collector. 33 33 00:01:08,700 --> 00:01:11,700 This way we can capture information about the flow 34 34 00:01:11,700 --> 00:01:15,450 instead of the data itself, saving us a lot of space. 35 35 00:01:15,450 --> 00:01:17,730 But we aren't going to have the contents 36 36 00:01:17,730 --> 00:01:19,260 of what was going over the network, 37 37 00:01:19,260 --> 00:01:21,480 just information about it. 38 38 00:01:21,480 --> 00:01:24,810 Now, this information can be stored inside of a database 39 39 00:01:24,810 --> 00:01:26,670 and queried later by our different tools 40 40 00:01:26,670 --> 00:01:28,620 and produce reports and graphs. 41 41 00:01:28,620 --> 00:01:31,170 This is the idea of flow analysis. 42 42 00:01:31,170 --> 00:01:32,940 Now, the great thing about flow analysis 43 43 00:01:32,940 --> 00:01:35,340 is it's going to allow us to highlight trends and patterns 44 44 00:01:35,340 --> 00:01:36,960 in the traffic being generated. 45 45 00:01:36,960 --> 00:01:39,090 It allows us to get alerts based on anomalies 46 46 00:01:39,090 --> 00:01:41,130 that we might see and different patterns and triggers 47 47 00:01:41,130 --> 00:01:42,150 that we might see. 48 48 00:01:42,150 --> 00:01:43,770 We can also have visualization tools 49 49 00:01:43,770 --> 00:01:45,720 that allow us to quickly create a map 50 50 00:01:45,720 --> 00:01:48,570 of different network connections and the flow patterns, 51 51 00:01:48,570 --> 00:01:50,820 and we can identify different traffic patterns 52 52 00:01:50,820 --> 00:01:53,910 that might reveal bad behavior, malware in transit, 53 53 00:01:53,910 --> 00:01:56,490 tunneling, and other bad things out there. 54 54 00:01:56,490 --> 00:01:59,370 Now, to do this, we can use several different tools. 55 55 00:01:59,370 --> 00:02:01,770 For instance, there is NetFlow, Zeek, 56 56 00:02:01,770 --> 00:02:04,080 and Multi Router Traffic Grapher. 57 57 00:02:04,080 --> 00:02:05,730 Let's talk about each of these. 58 58 00:02:05,730 --> 00:02:07,740 First, we have NetFlow. 59 59 00:02:07,740 --> 00:02:09,570 NetFlow is a Cisco-developed means 60 60 00:02:09,570 --> 00:02:11,250 of reporting network flow information 61 61 00:02:11,250 --> 00:02:12,990 to a structured database. 62 62 00:02:12,990 --> 00:02:14,640 This is one of the first ones out there, 63 63 00:02:14,640 --> 00:02:16,770 and eventually it actually became the standard 64 64 00:02:16,770 --> 00:02:18,210 that everyone started using 65 65 00:02:18,210 --> 00:02:22,500 under the name IPFIX or IP Flow Information Export. 66 66 00:02:22,500 --> 00:02:25,320 Now, this allows us to define a particular traffic flow 67 67 00:02:25,320 --> 00:02:26,610 based on the different packets 68 68 00:02:26,610 --> 00:02:28,350 that share the same characteristics. 69 69 00:02:28,350 --> 00:02:31,590 For instance, maybe it's the same source and destination IP 70 70 00:02:31,590 --> 00:02:34,290 signifying that there's a session between two hosts. 71 71 00:02:34,290 --> 00:02:35,790 That would be one data flow, 72 72 00:02:35,790 --> 00:02:37,770 and we can get information about it. 73 73 00:02:37,770 --> 00:02:39,510 Now, some of the information we can capture 74 74 00:02:39,510 --> 00:02:42,630 inside of NetFlow is going to be information about the packets 75 75 00:02:42,630 --> 00:02:44,130 that go over the devices, 76 76 00:02:44,130 --> 00:02:45,960 things like the network protocol interface 77 77 00:02:45,960 --> 00:02:49,290 that's being used, the version and type of IP being used, 78 78 00:02:49,290 --> 00:02:51,480 the source and destination IP address, 79 79 00:02:51,480 --> 00:02:53,280 the source and destination port, 80 80 00:02:53,280 --> 00:02:55,710 or the IP's type of service. 81 81 00:02:55,710 --> 00:02:57,570 All of this information can be gathered 82 82 00:02:57,570 --> 00:03:00,120 and then we can actually analyze it and look at it visually 83 83 00:03:00,120 --> 00:03:01,680 using different tools. 84 84 00:03:01,680 --> 00:03:04,050 For example, here on the screen, you can see that I'm using 85 85 00:03:04,050 --> 00:03:08,190 the SolarWinds tool to show the NetFlow data of a network, 86 86 00:03:08,190 --> 00:03:10,440 or you can look at it in a text-based manner. 87 87 00:03:10,440 --> 00:03:13,020 For instance, here you can see the different flows 88 88 00:03:13,020 --> 00:03:16,110 that are there going from one all the way down to flow 15, 89 89 00:03:16,110 --> 00:03:18,930 and as I increase the ability to look into flow 15, 90 90 00:03:18,930 --> 00:03:21,090 you can see the source and destination address, 91 91 00:03:21,090 --> 00:03:23,220 the source port, the destination port, 92 92 00:03:23,220 --> 00:03:24,930 and some basic information about it, 93 93 00:03:24,930 --> 00:03:27,210 but you don't see the content of any of those packets 94 94 00:03:27,210 --> 00:03:28,290 that have been going through. 95 95 00:03:28,290 --> 00:03:29,310 Now because of this, 96 96 00:03:29,310 --> 00:03:31,470 NetFlow is going to provide you with metadata 97 97 00:03:31,470 --> 00:03:33,600 but you're not going to have the full packet captures, 98 98 00:03:33,600 --> 00:03:35,190 so you're not going to have a complete record 99 99 00:03:35,190 --> 00:03:36,510 of what's happening. 100 100 00:03:36,510 --> 00:03:38,850 Now, if you want to be able to have the best of both worlds, 101 101 00:03:38,850 --> 00:03:40,740 you can use something like Zeek. 102 102 00:03:40,740 --> 00:03:43,800 Zeke is a hybrid tool that passively monitors your network 103 103 00:03:43,800 --> 00:03:44,850 like a sniffer, 104 104 00:03:44,850 --> 00:03:47,250 but it's only going to log full packet capture 105 105 00:03:47,250 --> 00:03:49,620 based on data of potential interest. 106 106 00:03:49,620 --> 00:03:52,080 So essentially it's going to be sampling that data 107 107 00:03:52,080 --> 00:03:53,670 just like NetFlow does, 108 108 00:03:53,670 --> 00:03:55,590 but when it finds something that's quite interesting, 109 109 00:03:55,590 --> 00:03:57,330 it's going to log the entire thing, 110 110 00:03:57,330 --> 00:03:59,550 so we can then go back and look at that. 111 111 00:03:59,550 --> 00:04:01,080 This helps reduce our storage 112 112 00:04:01,080 --> 00:04:02,700 and our processing requirements, 113 113 00:04:02,700 --> 00:04:05,190 and it gives us an ability to have all this data 114 114 00:04:05,190 --> 00:04:07,110 into one single format. 115 115 00:04:07,110 --> 00:04:09,480 One of the great things about Zeek is that it performs 116 116 00:04:09,480 --> 00:04:10,980 normalization of the data 117 117 00:04:10,980 --> 00:04:13,620 and then it stores that data as a tab-delimited 118 118 00:04:13,620 --> 00:04:15,780 or JavaScript Object Notation language, 119 119 00:04:15,780 --> 00:04:17,850 JSON-formatted text files. 120 120 00:04:17,850 --> 00:04:18,780 This allows you to use it 121 121 00:04:18,780 --> 00:04:20,760 with lots of other different tools. 122 122 00:04:20,760 --> 00:04:22,860 For instance, here you can see on the screen 123 123 00:04:22,860 --> 00:04:24,420 that since this file was normalized, 124 124 00:04:24,420 --> 00:04:27,240 I could import it into another tool for visualization, 125 125 00:04:27,240 --> 00:04:29,010 searching, and analysis. 126 126 00:04:29,010 --> 00:04:32,160 In this image, I've imported the Zeek log into Splunk 127 127 00:04:32,160 --> 00:04:34,860 and from there I'm able to search for specific information 128 128 00:04:34,860 --> 00:04:37,290 during my analysis of a potential incident. 129 129 00:04:37,290 --> 00:04:38,370 The third tool we have 130 130 00:04:38,370 --> 00:04:42,390 is the Multi Router Traffic Grapher or MRTG. 131 131 00:04:42,390 --> 00:04:44,040 Now, if you took Network Plus, 132 132 00:04:44,040 --> 00:04:45,330 you've been introduced to this before 133 133 00:04:45,330 --> 00:04:47,250 because it was on that syllabus as well. 134 134 00:04:47,250 --> 00:04:50,100 The MRTG is a tool that's used to create graphs 135 135 00:04:50,100 --> 00:04:51,900 that show network traffic flows 136 136 00:04:51,900 --> 00:04:53,220 through the network interfaces 137 137 00:04:53,220 --> 00:04:54,840 of different routers and switches 138 138 00:04:54,840 --> 00:04:57,570 by polling those appliances using SNMP, 139 139 00:04:57,570 --> 00:04:59,970 the Simple Network Management Protocol. 140 140 00:04:59,970 --> 00:05:02,250 This looks like this graph here on the screen. 141 141 00:05:02,250 --> 00:05:04,230 Now, what is useful about this? 142 142 00:05:04,230 --> 00:05:06,810 Well, you can see patterns emerging. 143 143 00:05:06,810 --> 00:05:10,170 For instance, in the top graph on the router firewall, 144 144 00:05:10,170 --> 00:05:12,420 where do you see a big spike in traffic? 145 145 00:05:12,420 --> 00:05:14,520 You can see that there's a big spike in traffic 146 146 00:05:14,520 --> 00:05:16,590 between 2:00 and 4:00 am. 147 147 00:05:16,590 --> 00:05:17,730 Now, is that normal? 148 148 00:05:17,730 --> 00:05:20,040 Maybe and maybe not, but it's something we should go 149 149 00:05:20,040 --> 00:05:22,440 and further analyze, because if we see a big spike 150 150 00:05:22,440 --> 00:05:25,350 between 2:00 and 4:00 am, that could be something normal, 151 151 00:05:25,350 --> 00:05:27,330 like we are doing offsite backups, 152 152 00:05:27,330 --> 00:05:29,430 and therefore we're going to do that during off-peak hours, 153 153 00:05:29,430 --> 00:05:30,870 and you'll see a big spike in traffic 154 154 00:05:30,870 --> 00:05:33,000 as we're sending our data to our offsite facility. 155 155 00:05:33,000 --> 00:05:34,680 Maybe that's a reasonable explanation, 156 156 00:05:34,680 --> 00:05:37,620 or maybe the server has been infected with malware 157 157 00:05:37,620 --> 00:05:40,320 and at 2:00 to 4:00 am when we're all home sleeping, 158 158 00:05:40,320 --> 00:05:42,690 they're sending all the data back to the bad guy 159 159 00:05:42,690 --> 00:05:45,660 who's now collecting our data as part of data exfiltration. 160 160 00:05:45,660 --> 00:05:47,250 We don't know just by looking at this 161 161 00:05:47,250 --> 00:05:48,990 which of those two cases it is, 162 162 00:05:48,990 --> 00:05:50,910 but by looking at this we can identify 163 163 00:05:50,910 --> 00:05:53,250 that we should zero in between 2:00 and 4:00 am 164 164 00:05:53,250 --> 00:05:55,740 and figure out what was that additional traffic flow 165 165 00:05:55,740 --> 00:05:57,180 that happened and why. 166 166 00:05:57,180 --> 00:05:59,460 So based on this, we might make a hypothesis 167 167 00:05:59,460 --> 00:06:01,590 that we think data exfiltration is happening. 168 168 00:06:01,590 --> 00:06:03,840 In that case, we would want to set up a network sniffer 169 169 00:06:03,840 --> 00:06:06,510 in front of this server to see what traffic is leaving 170 170 00:06:06,510 --> 00:06:08,130 and where it's going to. 171 171 00:06:08,130 --> 00:06:09,750 Based on that, we can then dive deeper 172 172 00:06:09,750 --> 00:06:11,880 to see if it was something that was known good, 173 173 00:06:11,880 --> 00:06:15,273 such as a backup, or known bad, such as a bad guy.