WEBVTT 00:00.080 --> 00:00.560 Hey, Ryan. 00:00.560 --> 00:01.460 And welcome back. 00:01.460 --> 00:07.120 Now in today's video, we will look into how Nginx can be implemented as a load balancer. 00:07.130 --> 00:13.400 Now one of the great things about Nginx is that Nginx can be configured as a load balancer to improve 00:13.400 --> 00:16.610 your server resource availability as well as efficiency. 00:16.610 --> 00:17.810 So now what will happen? 00:17.810 --> 00:24.140 A user will go ahead and send a request to Nginx, which is acting as a load balancer and Nginx depending 00:24.140 --> 00:29.660 upon the load balancing algorithm that you have configured, will distribute the traffic among multiple 00:29.660 --> 00:30.260 servers. 00:30.260 --> 00:32.590 So this is the basic architecture here. 00:32.600 --> 00:38.240 Now in order to distribute the traffic among multiple back end servers, you can make use of the upstream 00:38.240 --> 00:40.640 block within the Nginx configuration file. 00:40.640 --> 00:46.430 So the upstream block can be used to specify the group of servers for which you want the load balanced 00:46.430 --> 00:47.960 traffic to be distributed. 00:47.960 --> 00:53.510 So for example, you have a upstream block and within the upstream block you specify the IP addresses 00:53.510 --> 00:55.550 where the traffic should be distributed. 00:55.550 --> 01:03.480 So let's assume the IP address of server one is 1077 0.2 and server two is 1077 0.3. 01:03.510 --> 01:08.460 So based on this upstream block, Nginx will go ahead and distribute the traffic accordingly. 01:08.460 --> 01:13.680 Now if you look into the sample configuration here, first thing here you define is the upstream block. 01:13.680 --> 01:16.560 So the name of this upstream block is called as backend. 01:16.560 --> 01:19.070 And within here you have defined two servers. 01:19.080 --> 01:27.360 First one is IP of ten 139 0.3 second has IP of ten 139 0.4 So this is one of the unique things that 01:27.360 --> 01:28.170 you need to do. 01:28.170 --> 01:33.620 And then within the server block, all you have to do is you have to specify a proxy pass. 01:33.630 --> 01:38.610 And within here you have to specify the Http followed by the name of the upstream block here, which 01:38.610 --> 01:39.140 is backend. 01:39.150 --> 01:40.650 So here you have specified backend. 01:40.680 --> 01:43.740 Now we have already discussed about the proxy pass directive. 01:43.740 --> 01:46.830 So the second part is something that we already know. 01:46.860 --> 01:52.320 The new part here is the upstream block, which is very simple to understand. 01:52.320 --> 01:55.140 So let's go ahead and look into it practically. 01:55.140 --> 02:01.350 So for today's video, I have three virtual machines available, and in all of them I have the Nginx 02:01.350 --> 02:03.750 that is installed and up and running. 02:03.780 --> 02:08.100 Among which one of the virtual machine will be acting as a load balancer. 02:08.100 --> 02:11.340 So when we discuss that, it is acting as a load balancer. 02:11.340 --> 02:17.700 It basically has this upstream block configuration file that is available and you have a server one 02:17.700 --> 02:18.810 server zero two. 02:18.810 --> 02:25.320 Both of them basically have a Nginx with a specific index.html file to identify them. 02:25.320 --> 02:29.400 So if I just copy the IP of server zero one, let me paste it here. 02:29.520 --> 02:36.270 It returns server zero one If I copy the IP address of server zero two, it basically returns the server 02:36.270 --> 02:37.080 zero two here. 02:37.080 --> 02:41.100 So similar to the architecture that we are discussing, I have three terminals up. 02:41.100 --> 02:44.460 First one is related to the load balancer virtual machine. 02:44.460 --> 02:48.450 Second associated with server zero, one, third associated with server zero two. 02:48.480 --> 02:53.220 Now let me quickly just show you if I quickly go to user share Nginx. 02:53.430 --> 03:00.870 HTML and if I do a cat on index.html it basically contains this is server zero one and same for server 03:00.870 --> 03:01.320 zero two. 03:01.350 --> 03:07.110 It basically contains this is server zero two now for the load balancer based server. 03:07.470 --> 03:14.910 If I quickly go to etcd Nginx.conf dot d, I only have one configuration available called as load balancer 03:14.910 --> 03:15.930 dot conf. 03:15.960 --> 03:20.910 Now if I quickly open this up, it has a very simple configuration. 03:20.910 --> 03:24.510 You have a upstream block here and then you have a server block. 03:24.510 --> 03:30.450 What this server block is doing, it is doing a proxy pass to backend and the backend is nothing but 03:30.450 --> 03:32.760 the reference to this upstream block here. 03:32.790 --> 03:39.240 Now within here we have specified the private IP address of the server one year and the private IP address 03:39.240 --> 03:40.080 of the server two. 03:40.080 --> 03:51.090 So if I directly make a request here, say ten 139 0.3 it basically says server two and ten 139 0.4 03:51.120 --> 03:55.770 it basically says server zero one So this is the basic configuration here. 03:55.800 --> 03:59.400 Now let's do one thing from the load balancer base Nginx instance. 03:59.400 --> 04:01.890 Let's quickly do a curl on localhost. 04:01.890 --> 04:05.310 So this will basically send the request to the load balancer itself. 04:05.310 --> 04:11.190 So when I do a curl on localhost this time it is loading the server zero two let's do a curl localhost 04:11.190 --> 04:11.910 once more. 04:12.360 --> 04:14.790 This time you see it is loading the server. 04:14.790 --> 04:21.060 Zero one again if I make a request server zero two again, if I make a request server zero one So this 04:21.060 --> 04:24.990 is one of the easiest way in which Nginx can work as a load balancer. 04:24.990 --> 04:27.300 Again, very similar to the curl client. 04:27.300 --> 04:30.390 You can even make a request from the browser. 04:30.390 --> 04:33.810 Let's load the IP address of the load balancer here. 04:33.810 --> 04:36.240 First request to server zero two. 04:36.540 --> 04:41.550 Then you have server zero one again, server zero two again server zero one. 04:41.550 --> 04:47.880 So I hope with this you understood the basics of load balancer and how Enginex can be used as a load 04:47.880 --> 04:51.240 balancer to distribute the traffic among multiple backends. 04:51.240 --> 04:53.700 So with this we'll conclude today's video.