1 00:00:00,970 --> 00:00:01,920 Hello and welcome. 2 00:00:02,130 --> 00:00:09,690 So last lecture about the VPC theoretical part and this is the direct connect basically and the high 3 00:00:09,690 --> 00:00:13,140 performance low latency high bandwidth connection. 4 00:00:13,140 --> 00:00:25,080 That is not Internet based between your VBC and E.W. us and your on premise headquarter branch offices 5 00:00:25,080 --> 00:00:29,870 or data centers basically or remote location that you need to connect your viewers. 6 00:00:30,110 --> 00:00:30,650 OK. 7 00:00:31,020 --> 00:00:34,290 So how does that look like. 8 00:00:34,300 --> 00:00:42,870 So here is your the PC and here is your remote location on the customer network or the on premise site 9 00:00:43,330 --> 00:00:49,500 that you would like to connect to your PC and you don't want. 10 00:00:49,490 --> 00:00:55,250 You are not interested in going over the Internet so the VPN connection here is not what you are targeting 11 00:00:55,260 --> 00:00:56,550 you are targeting. 12 00:00:56,550 --> 00:01:04,450 On the other hand a higher speed higher bandwidth less latency and higher performance connection. 13 00:01:04,680 --> 00:01:10,080 And of course the Internet can give you that cannot guarantee high speed or high performance all the 14 00:01:10,080 --> 00:01:17,780 time and cannot guarantee latency and hence its not the expected performance that you need. 15 00:01:17,790 --> 00:01:20,570 Why would you need that connection you could have. 16 00:01:20,610 --> 00:01:27,330 You may have a lot of traffic that you would like to exchange between your network and your VBC and 17 00:01:27,330 --> 00:01:30,960 it could be also that you are running some applications that are time sensitive. 18 00:01:30,960 --> 00:01:32,280 Basically they are not. 19 00:01:32,370 --> 00:01:40,410 They cannot tolerate the fluctuating latencies and light and jitters and all that on the Internet. 20 00:01:40,590 --> 00:01:43,660 But of course nothing nothing is for free on the Internet. 21 00:01:43,680 --> 00:01:50,850 Its quick to deploy it is cost effective but there are no guarantees on the performance on that day 22 00:01:50,870 --> 00:01:57,120 connect it takes more time there is a lead time to submit requests to submit requests to the partner 23 00:01:57,120 --> 00:02:05,640 providers that support that connect connection in order to lay down the connections to your premises 24 00:02:05,670 --> 00:02:13,530 and on the other side as we see now connected in the connect locations of AWOS and that to be mapped 25 00:02:13,590 --> 00:02:15,880 then to your VTC it takes time. 26 00:02:16,080 --> 00:02:21,200 So to hire performance and all that but if you have an immediate need you would like to do some testing. 27 00:02:21,360 --> 00:02:26,750 And you cannot wait then definitely VPN is the way to go because you don't have any other option. 28 00:02:27,270 --> 00:02:33,850 Or if you can your product can wait for a few weeks or few months even sometimes then be it. 29 00:02:33,870 --> 00:02:40,260 That would be the direction that we need to understand some new thing called E.W. direct connect location 30 00:02:40,260 --> 00:02:40,550 here. 31 00:02:40,590 --> 00:02:41,920 So what is it a mistake. 32 00:02:42,000 --> 00:02:43,080 Good question. 33 00:02:43,320 --> 00:02:49,710 Basically lets say in the U.S. I mean you have Equinix you have AT&T you have horizon and you have other 34 00:02:49,710 --> 00:02:55,770 companies that can provide you sort of wide area network links so basically and to connect something 35 00:02:55,770 --> 00:03:05,540 in Dallas to something in lets say Austin in Texas or New York then thats a long distance connection 36 00:03:05,550 --> 00:03:07,820 we call it wide area network connection. 37 00:03:07,830 --> 00:03:15,480 Now some of these providers have been choosing or have gone into agreement and partnership with us in 38 00:03:15,480 --> 00:03:19,260 order to connect to the customer that their customers to us. 39 00:03:19,260 --> 00:03:20,620 How is that possible. 40 00:03:20,700 --> 00:03:27,050 There is location in between that will have provider X provider presence and its obvious presence. 41 00:03:27,180 --> 00:03:34,530 And this is called direct connect location by the way that connect or the X they are the x is a symbol 42 00:03:34,530 --> 00:03:36,870 or an acronym for data connect. 43 00:03:36,870 --> 00:03:46,320 So I have here provider partner provider rack and it track in the same data location so I will connect 44 00:03:46,380 --> 00:03:54,030 the customer router on the customer network I will connect connected to a customer out a provision by 45 00:03:54,030 --> 00:04:00,090 either by the customer or by the provider the partner provider Equinix or similar. 46 00:04:00,120 --> 00:04:05,820 So this could be a customer the customer can have a track here and then he gets the connection from 47 00:04:05,820 --> 00:04:14,580 the provider and the customer can own these two routers or they can be outsourced or least from the 48 00:04:14,580 --> 00:04:15,870 partner provider. 49 00:04:16,290 --> 00:04:18,360 So once they have this connection is that all. 50 00:04:18,360 --> 00:04:22,400 No I still need to connect to it as it is on the tracks. 51 00:04:22,500 --> 00:04:27,910 They will have what we call they connect routers like connect equipment. 52 00:04:28,110 --> 00:04:35,450 These data connect equipment are connected to back home to the different various regions all the different 53 00:04:35,460 --> 00:04:43,070 it is a very it is on in the region anyway they are connected to you of your environment from the outside. 54 00:04:43,170 --> 00:04:45,900 Now how can I connect this one with that one. 55 00:04:45,900 --> 00:04:52,540 This is AWOS responsibility this is the provide the responsibility and then either just works with the 56 00:04:52,540 --> 00:04:58,970 partner or with your daughter in order to cross-connect this route or this. 57 00:04:59,040 --> 00:05:04,520 So now the connection is complete from the customer out here all the way to the virtual gateway. 58 00:05:04,650 --> 00:05:07,330 I have this connection after I have this connection. 59 00:05:07,350 --> 00:05:08,570 What do I do. 60 00:05:08,880 --> 00:05:12,900 So this connection has to use something called sub interfaces. 61 00:05:12,900 --> 00:05:16,350 Those of you who are from MIT and from networking background probably they will understand what they 62 00:05:16,350 --> 00:05:18,340 mean but what are some interface. 63 00:05:18,350 --> 00:05:26,560 So basically if I have this connection now and then coming through the data connection and all the way 64 00:05:26,560 --> 00:05:35,080 to the VZW so I have this pipe and that pipe could be one gig could be a hundred million it could be 65 00:05:35,080 --> 00:05:36,250 10 gig. 66 00:05:36,640 --> 00:05:39,480 So basically it supports 1 gig and 10 gig. 67 00:05:39,490 --> 00:05:45,410 But on that you can have four speeds starting from 50 megabits per second hundred megabits per second 68 00:05:45,410 --> 00:05:46,070 and onwards. 69 00:05:46,120 --> 00:05:51,100 How can I use that how can I can figure that we use something called sub interfaces or basically channels 70 00:05:51,640 --> 00:05:54,090 on the pipe that you have. 71 00:05:54,250 --> 00:05:58,290 And this is supported via something called to do that one. 72 00:05:58,420 --> 00:06:07,030 Basically I split that to sub interfaces or villans what is a villain a villain is really when you send 73 00:06:07,030 --> 00:06:14,470 your packet or your frame you will have an identifier from the packet as it goes out from here and it 74 00:06:14,470 --> 00:06:20,440 passes through this hardware in order to ensure that it is received on the right channel here and this 75 00:06:20,440 --> 00:06:24,370 channel or this villain and that will and they are completely isolated from each other and they will 76 00:06:24,370 --> 00:06:26,500 have different IP addressing as well. 77 00:06:26,770 --> 00:06:32,300 And this is going all the way so that they are three IP addresses will be here and here. 78 00:06:33,290 --> 00:06:37,430 Now can I do a little to switching over this connection. 79 00:06:37,430 --> 00:06:40,780 The answer is no because there has to be an IP address here. 80 00:06:40,820 --> 00:06:45,320 There has to be an IP address here and you have to configure BGP routing. 81 00:06:45,470 --> 00:06:46,280 On top of them. 82 00:06:46,280 --> 00:06:46,920 Why. 83 00:06:47,060 --> 00:06:53,510 Because the user needs to talk about to send few subnets from the site across. 84 00:06:53,510 --> 00:07:00,440 And also the customer router needs to send some routes from here that they will exchange traffic with 85 00:07:00,800 --> 00:07:03,570 two instances or it services. 86 00:07:03,740 --> 00:07:05,930 So this has to be transmitted this way. 87 00:07:05,930 --> 00:07:07,910 This has to be transmitted that way. 88 00:07:07,910 --> 00:07:13,130 This is how the reader knows about the remote and this is how the customer router knows about the VPC 89 00:07:13,130 --> 00:07:13,700 site. 90 00:07:15,050 --> 00:07:19,090 They now have multiple types of these interfaces. 91 00:07:19,100 --> 00:07:26,390 These are by the way are called virtual interface or VI of yes you have two types you have public the 92 00:07:26,430 --> 00:07:32,890 highest and you have private the idea of what what they use public word or use or when doing with public 93 00:07:32,890 --> 00:07:34,130 windows private. 94 00:07:34,180 --> 00:07:43,930 If your intent is from here to try to reach to public services like what like S3 storage like glacier 95 00:07:44,020 --> 00:07:51,730 archiving if you want to get to the public services you have to use the green one here represents a 96 00:07:51,730 --> 00:08:00,250 public the IAF and this is villaine number two if you want to reach your VBC and VBC subnets then you 97 00:08:00,250 --> 00:08:06,100 go through a private VHF and this is the last one. 98 00:08:06,330 --> 00:08:09,480 What's the difference here on the public. 99 00:08:09,510 --> 00:08:16,820 You need to use public Internet through double IP addresses in this one you can use private IP addresses 100 00:08:16,830 --> 00:08:27,160 and usually it will allocate the private ones from the one 16:9 one 2:54 that X that x range to interface 101 00:08:27,270 --> 00:08:29,070 will tell you use this one. 102 00:08:29,070 --> 00:08:29,490 Why. 103 00:08:29,490 --> 00:08:33,990 Because they are also having that connect with other customers and they want to make sure that they 104 00:08:33,990 --> 00:08:40,350 don't confuse sending your traffic to another one because you have an overlapping 169 IP addresses. 105 00:08:40,350 --> 00:08:48,510 So that's why also it controls the IP address on the private lives on the public VISX using Internet 106 00:08:48,510 --> 00:08:54,360 Protocol IP addresses ensures that there will be no conflict between you and another customer because 107 00:08:54,360 --> 00:08:57,540 the Internet addresses cannot be duplicated. 108 00:08:57,570 --> 00:08:57,980 Right. 109 00:08:58,010 --> 00:09:02,280 That connect is a high performance low latency connection. 110 00:09:03,610 --> 00:09:09,950 VHF is nothing more than an 82 that one Cuvee land or basically a channel or something interference 111 00:09:09,980 --> 00:09:10,800 on the link. 112 00:09:12,640 --> 00:09:19,230 You need to have one private wires to connect to your PC and private subnet and we need the public VHF 113 00:09:19,240 --> 00:09:21,790 to connect to it as public services. 114 00:09:21,880 --> 00:09:29,420 You cannot do layer to over your data collection has to be there three and you have to use BGP. 115 00:09:29,430 --> 00:09:30,670 There is no choice there. 116 00:09:30,820 --> 00:09:37,900 And as in the VPN if I have here in that instance or that gateway in my public subnet not the private 117 00:09:37,900 --> 00:09:38,370 subnet. 118 00:09:38,410 --> 00:09:45,430 So they go through my I efforts through the internet gateway and then outside to the Internet. 119 00:09:45,430 --> 00:09:47,310 And the answer is no you cannot do that. 120 00:09:47,480 --> 00:09:51,320 High availability with directly availability of that connect. 121 00:09:51,460 --> 00:09:58,420 Basically what you can do is you can have one or more customer routers and then you can connect to multiple 122 00:09:58,420 --> 00:09:59,720 direct connections. 123 00:09:59,740 --> 00:10:06,700 They could be even through different provider partners or Equinix and AT&T and then you connect to that 124 00:10:06,720 --> 00:10:12,640 they will connect you through different instances on your PC side case. 125 00:10:12,670 --> 00:10:18,220 That's how you achieve that and through BGP you can do active active so both connections can be active 126 00:10:18,220 --> 00:10:19,280 at the same time. 127 00:10:19,300 --> 00:10:21,210 Or you can do active passive. 128 00:10:21,220 --> 00:10:24,000 So basically one will be active the other one will not be active. 129 00:10:24,010 --> 00:10:27,350 And here's one example for that. 130 00:10:27,360 --> 00:10:32,910 So here I have two customer routers to dial connect connections. 131 00:10:32,910 --> 00:10:38,820 I can have them to two different direct connect locations and then they connect all their back all to 132 00:10:38,820 --> 00:10:40,740 that virtual private gateway. 133 00:10:40,820 --> 00:10:46,560 Of course this is again as I said I can do active active or active backup active standby. 134 00:10:46,800 --> 00:10:50,470 But this is an expensive redundancy. 135 00:10:50,610 --> 00:10:57,360 You can use it in case if you cannot afford if one connection fails you cannot afford to downgrade the 136 00:10:57,360 --> 00:11:00,200 performance or increase the latency on the connection. 137 00:11:00,420 --> 00:11:02,950 And that's why you go there to connect both ways. 138 00:11:02,960 --> 00:11:09,220 So to customers out there so that Internet connections and to connect locations and maybe we should 139 00:11:09,220 --> 00:11:09,890 be routing. 140 00:11:10,020 --> 00:11:10,440 OK. 141 00:11:10,470 --> 00:11:15,420 So here we have another example of redundancy or fault tolerance. 142 00:11:15,450 --> 00:11:20,310 And in this case I have a high speed connection that connects the alternative or the backup. 143 00:11:20,300 --> 00:11:22,180 The connection will be VPN. 144 00:11:22,230 --> 00:11:24,990 So that's a lower performance more latency. 145 00:11:25,050 --> 00:11:25,770 In this case. 146 00:11:25,790 --> 00:11:28,680 So I can still have two customer routers here. 147 00:11:29,010 --> 00:11:29,790 I have two connections. 148 00:11:29,790 --> 00:11:32,010 One is I connect the other one is VPN. 149 00:11:32,130 --> 00:11:35,030 But of course the performance will not be the same. 150 00:11:35,160 --> 00:11:38,090 And the VPN definitely will not be used as active active. 151 00:11:38,100 --> 00:11:40,200 This will be the backup connection in my case. 152 00:11:40,230 --> 00:11:42,860 And of course VPN is a cheaper backup connection. 153 00:11:42,870 --> 00:11:50,970 One thing to note about that connect is once you have that connection to the region to have peace in 154 00:11:50,970 --> 00:11:54,810 the region do you have access to one of these on to the citizens. 155 00:11:54,820 --> 00:11:59,510 Are all of them you have access to all the Easy's in the region. 156 00:12:00,840 --> 00:12:06,870 And you can establish if you want to connect to remote regions you can establish IP SEC since this one 157 00:12:06,870 --> 00:12:12,980 has access to all the ICT that it has public IP addresses and public ranges. 158 00:12:13,140 --> 00:12:20,430 You can establish an IP connection to the VZW in remote regions over the public VHF and that's how we 159 00:12:20,430 --> 00:12:25,850 can even connect to your remote regions through the that connect. 160 00:12:25,870 --> 00:12:31,240 This is not related to that specifically but this is generally speaking in one routing table I can have 161 00:12:31,240 --> 00:12:33,110 only one default route. 162 00:12:33,280 --> 00:12:40,600 So they can have only one entry that points to 0 0 0 0 and this has to go either due to AGW to your 163 00:12:40,610 --> 00:12:45,040 VZW or to your peering connection very busy peering connection. 164 00:12:45,040 --> 00:12:46,850 But you cannot have more than one. 165 00:12:46,870 --> 00:12:53,330 Remember that propagation can be the alternative for the VZW to exchange routes with your routing table 166 00:12:53,920 --> 00:12:55,590 and that's it. 167 00:12:56,980 --> 00:13:03,440 So in the next lecture we'll start looking at the scenary questions and we'll have more fun applying. 168 00:13:03,450 --> 00:13:08,050 But we have learned through the theoretical lectures I know it might have been lengthy ones but I hope 169 00:13:08,050 --> 00:13:16,540 also that it was detailed enough and it made the point and the purpose by ensuring that the PC is very 170 00:13:16,540 --> 00:13:17,110 very clear. 171 00:13:17,110 --> 00:13:23,200 Now security groups network ACLC routing tables VZW is AGW VPN. 172 00:13:23,200 --> 00:13:29,680 All that is very clear and you don't have any ambiguity or concerns answering any question or any scenario 173 00:13:29,710 --> 00:13:35,310 and then we'll apply that also to the questions and hopefully by the end you have mastered and you can 174 00:13:35,320 --> 00:13:38,860 ask any related question in the exam. 175 00:13:39,040 --> 00:13:40,090 So I'll see you in the next lecture.