WEBVTT 00:00.000 --> 00:16.160 So hello everyone, today we're talking about the Incas and I'm going to create a private 00:16.160 --> 00:19.280 cloud with system containers. 00:19.280 --> 00:25.280 So just very quickly, I'm a CSAT main, basically, for many years, you're going to have 00:25.280 --> 00:32.080 a league day and no Incas signed so a lot of time right now, because I work for 00:32.080 --> 00:33.080 many of it. 00:33.080 --> 00:38.320 It's not very interesting to run, just to run, to send our system containers with 00:38.320 --> 00:39.320 Incas. 00:39.320 --> 00:42.560 So what is Incas? 00:42.560 --> 00:49.080 Incas is a foc of league day, it's basically the project that is the form it's basically 00:49.080 --> 00:55.240 dead, no Incas is a new project, it's like an official website, 00:55.240 --> 01:00.760 it's a next-generation system container manager and it turns into an application manager and 01:00.760 --> 01:02.680 got your own magic. 01:02.680 --> 01:17.600 So, many of you know of Kubernetes the out-of-way, okay, some, so it's very useful to learn 01:17.600 --> 01:25.600 communities work, and for Incas, let's do Incas the out-of-way. 01:25.600 --> 01:30.800 So what we will do, we will create a cluster from scratch, we'll configure it with 01:30.800 --> 01:37.400 oven networking, create containers, inside it, we are in the container room, and I connect 01:37.400 --> 01:43.800 to cluster and I look at the communication between the different network and each cluster. 01:43.800 --> 01:49.640 So of course it's the out-of-way, you can do everything with automation, you can use 01:49.640 --> 01:55.880 a project Incas deploy it to create a cluster, basically, instantly, but this top would 01:55.880 --> 02:03.920 be very short if I use that, and for myself, I like to see all things actually work. 02:03.920 --> 02:10.160 The slide I are here, if you want to pick up them, they are schema, but I think the screen 02:10.240 --> 02:18.640 is big enough so you can get it a layer, give you some things. 02:18.640 --> 02:24.480 All right, so what do we want to do? 02:24.480 --> 02:28.040 We want to do a private code. 02:28.040 --> 02:34.640 Let's say that we have two data centers, one, we will create a right cluster, and as 02:34.720 --> 02:37.800 the other one, we will create a root cluster. 02:37.800 --> 02:45.480 Each of them have a network connection with a router, and for a router, we will see 02:45.480 --> 02:52.240 but they need to communicate in some way, that's not a story. 02:52.240 --> 03:01.000 So let's try to do that in 20 minutes, so step one, or maybe step zero, oven. 03:01.040 --> 03:08.000 We will use oven everywhere, so to speed things up a little bit, let's say that oven 03:08.000 --> 03:11.240 is already configured. 03:11.240 --> 03:16.000 By the way, when you see a right prompt, it's the right cluster, when you see a group 03:16.000 --> 03:20.280 prompt, it's a blue cluster, it'd be easier for everyone. 03:20.280 --> 03:28.640 So let's say that we already have an oven network configured, and you see three servers, 03:28.640 --> 03:34.280 I have a tagamam, and they can communicate with the wind. 03:34.280 --> 03:39.320 Right now, we need to create an increased cluster. 03:39.320 --> 03:43.960 So in blue, that will be the thing that will be creating. 03:43.960 --> 03:51.640 So for starters, very simple, we will initiate a cluster, to initiate a cluster, you 03:51.640 --> 03:58.440 just wait to have a precede, it's not that hard, we just have to give it the IP address 03:58.440 --> 04:02.880 of the host, and some default settings. 04:02.880 --> 04:08.760 In this case, there is one default storage pool, and that's pretty much it. 04:08.760 --> 04:16.880 So let's create the cluster. 04:16.880 --> 04:23.340 So at this point, we have a cluster with one server, it's still a cluster, only one 04:23.340 --> 04:29.040 server, so not very interesting, but wait, it is. 04:29.040 --> 04:33.240 Let's complex a cluster, add two more nodes inside. 04:33.240 --> 04:43.320 To do that, we go on this first server and create some join token, then we go on the 04:43.320 --> 04:49.400 second server, beta in this case, and we will create another preset file. 04:49.400 --> 04:52.920 In this case, it's a little different to join the first cluster, of course, so you need 04:52.920 --> 04:58.400 to give it the IP address and the token. 04:58.400 --> 05:02.160 For the gamma cluster, it's pretty much the same. 05:02.160 --> 05:08.840 And at this point, we have a cluster, three nodes. 05:08.840 --> 05:15.440 So we just did that, created an inclusive cluster, but there is no configuration, nor we 05:15.440 --> 05:17.560 need to work in nothing. 05:17.560 --> 05:24.520 So right now, we need to create one thing, that is the fling network, it will be the link 05:24.520 --> 05:30.280 between every network inside the cluster and the sideworld, because at the server, 05:30.280 --> 05:34.200 that cannot communicate with the sideworld, it's pretty much useless. 05:34.200 --> 05:37.160 So let's begin by that. 05:37.160 --> 05:48.160 In this case, we will dedicate a physical interface, an H node to the fling, it's the same 05:48.160 --> 05:49.160 in this case. 05:49.160 --> 05:53.240 We want all servers to be able to use a fling. 05:53.240 --> 05:59.520 And we need to configure the fling in this case, it looks so scary, but it's not. 05:59.520 --> 06:07.940 It's just that the IP before gateway is actually the external router to go to the outside 06:07.940 --> 06:10.080 world. 06:10.080 --> 06:14.400 So that's it. 06:14.400 --> 06:19.880 And there, we will tell Incus to communicate with, because of course, he will need 06:19.880 --> 06:27.720 to create some of the network, and we don't want to do that with as of in NFS. 06:27.720 --> 06:34.360 So at this point, we did create a cluster and a new fling. 06:34.360 --> 06:40.800 So now, we will create the real, I mean, the internal of the network where we will put 06:40.800 --> 06:46.160 our containers, so it's important. 06:46.160 --> 06:53.880 At this point, we can use Incus for basically everything, so the first command on the top 06:53.880 --> 06:59.000 will create a network of a type of in. 06:59.000 --> 07:06.160 And it will automatically generate, use, you can see the last time, it will use randomize 07:06.160 --> 07:10.360 a network, but you can set whatever you want if you like it. 07:10.360 --> 07:16.400 And it will use a fling to go outside the cluster. 07:16.400 --> 07:18.400 So at this point, we are there. 07:18.400 --> 07:22.800 Still no container, let's create our first container. 07:22.800 --> 07:31.320 Right, the container is very easy, in this case, it's a system container, so we do that 07:31.320 --> 07:32.920 like this. 07:32.920 --> 07:40.960 And we just need to tell it where a witch network I will use and a witch storage. 07:40.960 --> 07:46.160 So in this case, we will create a container, we don't even tell it where it will be created 07:46.160 --> 07:52.000 because it's a cluster, so you'll see the location on the far, it's here and it will 07:52.000 --> 07:59.000 be created on alpha, but you can be created on any member of the cluster as a key. 07:59.000 --> 08:03.240 And then we can go inside the container, the system container, so you can run commands from 08:03.240 --> 08:11.440 it, and you can do some things air, if you remember the right people, it's the gateway, 08:11.440 --> 08:15.360 so it's outside the Incus cluster. 08:15.360 --> 08:21.360 So in this case, the container has an antenna access. 08:21.360 --> 08:28.600 So we are there, we have our first container, let's create another of in network, inside 08:28.600 --> 08:32.800 it's just a cluster with another container. 08:32.800 --> 08:41.320 So it's pretty much the same, we create a new network type of in, same common as before, 08:41.320 --> 08:47.440 and then we create another container that uses this new network. 08:47.440 --> 08:50.960 All right, we have two containers now. 08:50.960 --> 08:56.960 So we can see, by the way, it was created on beta, this time, and they have two different 08:56.960 --> 09:02.240 type of array, two different network, and they cannot communicate between them. 09:02.240 --> 09:09.800 That's a default, by default, it's as it ended, because we don't want all networks 09:09.800 --> 09:10.880 to communicate between them. 09:10.880 --> 09:17.240 We may want to have some that have more preven numbers that don't have links. 09:17.240 --> 09:24.840 If we go just a little bit in the oven hotel, they will say that the routing table, you 09:24.840 --> 09:29.000 don't have to do that, but it's interesting to see, routing table is very simple, you 09:29.000 --> 09:34.840 only have a default route at the moment. 09:34.840 --> 09:39.920 So we are there, we have two of in network, separated, both have an antenna access, but 09:39.920 --> 09:42.800 that's it. 09:42.800 --> 09:51.680 All right, now, we have a location between two of in network, inside the same cluster. 09:51.680 --> 09:55.200 That's basically appearing. 09:55.200 --> 10:02.280 So we will peer the two networks, do you have plus common two in terms of networks? 10:02.280 --> 10:08.360 You need to create a peer in bus ways, first way to the network one, to network two, 10:08.360 --> 10:10.800 and to network two to network one. 10:10.800 --> 10:15.160 But it's end the inquisterals to do that. 10:15.160 --> 10:23.200 So when you did that, you can see the pairing right there, and you can see the peer in bus 10:23.200 --> 10:24.200 direction. 10:24.200 --> 10:26.600 All right? 10:26.600 --> 10:33.680 And now, we can think containers from one network to another. 10:33.680 --> 10:40.520 And if we look again as a routing table, and of in, we will see that there is a new route 10:40.520 --> 10:50.240 in bus, in bus, hotel, virtual hotel, to give access to each network. 10:50.240 --> 10:52.680 We are there at the moment. 10:52.680 --> 11:02.440 So right now, if we want to continue and create our big cluster, basically it's an 11:02.440 --> 11:04.520 ability to do that at this point. 11:04.520 --> 11:09.120 If we want to create a region, we need our second cluster. 11:09.120 --> 11:16.880 So I won't do the same again, but let's consider that we did the same with a blue cluster. 11:16.880 --> 11:18.800 To a different point, we are there. 11:18.800 --> 11:25.760 We have two clusters with a open network, virtual machine, whatever, but it cannot communicate 11:25.760 --> 11:26.760 between them. 11:26.760 --> 11:32.000 Each cluster doesn't know zero there, particularly. 11:32.000 --> 11:35.400 So we will create a communication between them. 11:35.400 --> 11:45.920 In a open terminology, it's a open interconnect, so we will connect two of in cluster, 11:45.920 --> 11:49.760 basically. 11:49.760 --> 11:58.840 So we need to do some of in configuration, and the command is a little scary, but in 11:58.880 --> 12:05.760 fact, it's not that hard, it's just that we need to configure all the interconnect, to 12:05.760 --> 12:12.080 know each of the members of the blue and red cluster, so that they can communicate between 12:12.080 --> 12:13.080 them. 12:13.080 --> 12:23.360 I like it in blue and red, it's IP, you don't have to do that by end again, this time. 12:23.360 --> 12:31.940 And then you have some of in connection to do to configuration, it's here is to tell the 12:31.940 --> 12:40.800 of in network that it must do at our connection, and I don't think you can see it's not 12:40.800 --> 12:44.760 very important anyway. 12:44.760 --> 12:53.200 So at this point, you need some more configuration that we will tell of in to automatically 12:53.200 --> 12:57.960 learn the hood between the cluster, as you will have to add them manually in the routing 12:57.960 --> 13:03.600 table, which can be in the long. 13:03.600 --> 13:11.960 And here you can go back to in-cast, and you will need to create an iteration network. 13:11.960 --> 13:20.000 This integration, it's purple, it's a purple, it's a name of our region, as a claim. 13:20.000 --> 13:27.920 So here we, at the moment, there is only open integration, but in the future we can do some more, 13:27.920 --> 13:36.880 and here you can, you have to tell in-cast where to communicate with the open interconnect 13:36.880 --> 13:39.160 elements. 13:39.160 --> 13:48.720 So it's the same configuration, as we did right there, in fact. 13:48.720 --> 13:57.260 But this time, it allows in-cast to control, so if you end the elements to create a link 13:57.260 --> 14:00.460 between networks. 14:00.460 --> 14:06.900 And this, okay, it's a, it's a template to create, you don't have to name all the, all 14:06.900 --> 14:09.420 the integrations. 14:09.420 --> 14:21.020 Okay, so we need to do that on both cluster, blue and red, and at this time we did create 14:21.020 --> 14:25.780 an interconnect connection between the two cluster. 14:25.780 --> 14:34.540 So right now, we can come, we can peer to, to open network, we can make a link between 14:34.540 --> 14:40.700 two of the network inside the clusters, for instance red net one, and blue net two, whatever 14:40.700 --> 14:41.700 we want. 14:41.700 --> 14:55.220 To do that, so it's a green thing, there, we will create a peer between two network 14:55.220 --> 15:00.180 and side between the red and blue. 15:00.180 --> 15:05.660 So the first thing, it's to tell the first cluster that it will create a connection between 15:05.660 --> 15:12.420 one of its network, in this case, had net one, and that, it will, basically, export it, 15:12.420 --> 15:17.740 that's what it's remotes, and that's it. 15:17.740 --> 15:22.780 Now, if we look at, I don't, I don't want, we can look at the peer, and we will see 15:22.780 --> 15:27.580 two peers, the first is a local peer, because at the beginning we did create a connection 15:27.580 --> 15:34.860 where it had net one and net two, it's a local one, and the other one, the purple, is a 15:34.860 --> 15:35.860 remotes one. 15:35.860 --> 15:41.820 So it's ready to be connected to the other cluster. 15:41.820 --> 15:50.740 So we do with the other same thing on the blue cluster, create a network situation, and 15:50.780 --> 15:58.420 we, this time, we export, basically, the blue net one network, right? 15:58.420 --> 16:04.060 So we did a, in bus side, we did exports for the red net one, and blue net one with 16:04.060 --> 16:07.420 a notification name. 16:07.420 --> 16:16.300 So if we look in, again, we will see the time that there is a port that was been created 16:16.300 --> 16:22.860 to, to do the link, the actual link in, between the availability zone. 16:22.860 --> 16:30.740 You can see availability zone blue in the top, and the command is pretty long, but availability 16:30.740 --> 16:32.140 zone red, the bottom. 16:32.140 --> 16:40.460 So there is a link in the vein between the two, two zone. 16:40.460 --> 16:47.100 If we look at the ring table, we will see, and so, everything that's, that's oven as 16:47.100 --> 16:48.100 added. 16:48.100 --> 16:55.740 So this time you can, you will see the road to reach the other, and it's worth in the other 16:55.740 --> 16:59.100 cluster. 16:59.100 --> 17:03.660 And right now, we can try to ping things between the network. 17:03.660 --> 17:15.980 So in this case, we did a link between the red net one network, and blue net one network. 17:15.980 --> 17:23.820 So we go from the red net one to the oven, and connect, then to the other, and connect 17:23.820 --> 17:32.700 on the blue one, and also way back to the oven network. 17:32.700 --> 17:48.140 So right now, we can ping from the blue container one, red container one, and there, 17:48.140 --> 17:59.140 and this is, and this is, and this is it, I guess I'm in pretty fast. 17:59.140 --> 18:07.020 So at this point we did all this, so it's a world relation that was created, and we have 18:07.020 --> 18:15.180 a cluster, right cluster, that is, availability zone red, and availability zone blue, and 18:15.180 --> 18:21.980 we can add some more added green cluster, and whatever we want, and we can do a lot of things 18:21.980 --> 18:30.420 actually, we can add firewalls in the routers, whatever we want, we can do pureing between 18:30.420 --> 18:40.500 all of the, of the, of the in zone, and we can also add remotes between clusters, so 18:40.500 --> 18:49.140 we can copy containers between one cluster to another one, and we'll just restart into 18:49.140 --> 19:13.260 a cluster, do we have any questions, now, any questions, now? 19:13.260 --> 19:34.060 So, yeah, I'm just going to just expand on that, it's like, you know, basically, so it's 19:34.060 --> 19:41.060 the hard way, yeah, but we think as the play, you can do it the easy way, where you don't 19:41.140 --> 19:48.780 just need to define an inventory of number of servers that you have, in this case, you 19:48.780 --> 19:54.220 are doing clusters of three, so like, two clusters of three, and once you do it out, you 19:54.220 --> 19:57.780 have to get the answer, but against that, it's going to figure out the IP addresses and put 19:57.780 --> 20:02.460 the open services on that stuff, wherever it's needed, and when you do have two clusters, 20:02.460 --> 20:07.820 then you can pick how to do the deal of an interconnect, the other thing here is the weight 20:07.860 --> 20:16.820 was set up, is that every server was participating as a potential interconnect chassis, 20:16.820 --> 20:21.180 I think we call it, so that's one way to do it, but in some environments, you're going 20:21.180 --> 20:26.460 to have servers that have, that are going to be heavier on the network side, that have 20:26.460 --> 20:33.300 like faster next and like, in which case you don't actually activate the, the interconnect 20:33.300 --> 20:37.020 bit in open V-switch from every one of them, but just on the, on the ones that you want 20:37.020 --> 20:48.780 to act as effectively the transit to go from one cluster to another, yeah, we're just showing 20:48.780 --> 20:55.060 it, they probably because that's the simple way to do that, and it gets the play, so if anyone 20:55.060 --> 20:59.460 uses it in case already, and you want to learn this stuff, in case the play, lets you 20:59.460 --> 21:04.620 kind of do it nested effectively, so we've got Terraform files that will create a cluster 21:04.620 --> 21:12.140 of five machines using VMs, so five VMs with shared storage, extra disk and everything, 21:12.140 --> 21:17.500 and then when you run the Ansible bit, it's going to deploy other, save cluster that 21:17.500 --> 21:22.100 VM basically all of those cluster features together for you and those machines, and then 21:22.100 --> 21:26.900 you get to play with a cluster that's got basically all of those features, all enabled. 21:29.460 --> 21:45.860 Yeah, thank you for the demo, and the Residation and the whole, can you put some light 21:45.860 --> 21:55.460 on, you typically use cases, you can see for this kind of deal length, I mean, sorry, 21:55.460 --> 22:03.940 for Ingels it, yeah, I mean, with comparison like with some lightweight Kubernetes installations 22:03.940 --> 22:14.740 or something like this, yeah, it's a little different, I think I run Kubernetes and 22:14.740 --> 22:22.340 side in case, if you want, I think, but for Ingels, you can run system containers that's 22:22.340 --> 22:29.180 a little different from default application containers, you can have the run application 22:29.180 --> 22:38.980 containers right now, but in this case, it's ready to create a private code, for us, it's 22:38.980 --> 22:48.100 the basic infrastructure, it's below, it's on the hardware, the network, everything, if you 22:48.100 --> 22:53.780 want to run Kubernetes, you can run Kubernetes inside the SID in case, basically. 22:53.780 --> 22:58.740 Yeah, it's going to say like, pretty often for those deployments, so you want InCurse as the 22:58.740 --> 23:04.820 bare metal solution, it runs directly on metal, it provides stability to do content, well, 23:04.820 --> 23:08.900 because you can just input in our virtual machines at that level, and often for Kubernetes, 23:08.900 --> 23:14.500 you're actually consuming a full physical machine, cannot stretch these Kubernetes in many ways, 23:14.500 --> 23:18.580 where it's actually a better fit to create virtual machines on top of something like InCurse, 23:18.580 --> 23:22.740 and then run your case cluster within that, because you're going to want to run many 23:22.740 --> 23:26.900 different clusters, kind of different sizes, draw them, like, draw them, have some staging, 23:26.900 --> 23:31.940 production, all that kind of stuff, so it actually makes more sense, and what we see are not 23:31.940 --> 23:36.500 a fit rest in right now, with InCurse in those kind of deployments, like, proper that 23:36.500 --> 23:39.940 that sent out your deployments with multiple sites, and all that kind of stuff, are people who 23:39.940 --> 23:47.220 are moving away from VMware. So there's a lot of current VMware environments that are very 23:47.220 --> 23:52.020 difficult, very complex, very expensive to run with the new licensing model, and they're 23:52.020 --> 23:56.340 moving over to InCurse to a very private replacement for that, because we can run their 23:56.340 --> 24:02.420 metal, we are multi-tenant, we support doing cross cluster network and supporting most of 24:02.420 --> 24:08.340 the storage solutions, and those folks usually, they do run case, but they already run those 24:08.340 --> 24:11.060 case clusters within VMs, so it works the exact same way. 24:11.060 --> 24:19.140 I'm doing some case, it's nice to have system containers, because they are set full by default, 24:19.220 --> 24:27.220 and for instance, for databases, it's easier to, in my, it's easier to run them in system containers. 24:31.300 --> 24:38.180 What's there anyway, and if you can, it's okay, thank you very good, thanks a lot!