Latest / Nutanix Weekly / Nutanix Weekly: Why You Should Run Nutanix Kubernetes Platform on Nutanix Cloud Clusters

Transcript

Nutanix Weekly: Why You Should Run Nutanix Kubernetes Platform on Nutanix Cloud Clusters · Nutanix Weekly

WEBVTT 1 00:00:02.670 --> 00:00:16.600 Philip Sellers: Welcome to another episode of Nutanix Weekly one of the many Podcasts here out of Zintegra, I'm, your host today, Phil Sellers, I'm the practice director here for modern data center at Zintegra. 2 00:00:16.690 --> 00:00:34.089 Philip Sellers: And you know, this is one of the many podcasts that we run, we like to call it content with context, because we take the best content from the interwebs, and we try to bring real world experience from our own customer base, from 3 00:00:34.120 --> 00:00:47.830 Philip Sellers: our past experiences in different roles, and just bring real life to the content we review here, so want to say, welcome thanks for spending a little bit of time with us 4 00:00:48.170 --> 00:01:12.119 Philip Sellers: as usual. I can't do this alone, so I've got great co-hosts riding along with me today. So 1st up we've got Mr. Andy Green, a solutions. Architect here at Zintegra. Andy, this is month number 4 for you, or we're about to start month number 4 for you. So how are things going so far as a solutions. Architect 5 00:01:12.440 --> 00:01:21.760 Andy Greene: Things are going great. It's it's been a really good ride so far. And and you know, like you said, we're about to join month 4, but can't wait for everything to come 6 00:01:22.130 --> 00:01:29.349 Philip Sellers: Yeah, absolutely. You're doing great work already. You know, I know our customers are are enjoying their conversations. 7 00:01:30.357 --> 00:01:53.239 Philip Sellers: That you're having with them about their technology. We're also joined by Chris Calhoun, also a solutions architect here on the team. Not the new guy anymore. Thanks to Andy. But, Chris, you you've been around the block now for over half a year. What's probably the funnest conversation you've had with a customer so far 8 00:01:55.060 --> 00:02:14.359 Chris Calhoun | XenTegra: I'll tell you a quick story. So one of the customers went in asking questions about Nutanix, and and he started at a high level and and tried to not show his hand. But he was definitely a a steward and a student of his craft. 9 00:02:14.360 --> 00:02:31.269 Chris Calhoun | XenTegra: He actually knew the the ins and outs of Nutanix very well, and asked real specific questions about the the product, the benefit, the features, how it would help him. And I basically called him on it. I said, hang on. 10 00:02:31.720 --> 00:02:48.450 Chris Calhoun | XenTegra: I can tell right now you've done your homework. And so then we get into hey, there's blog forums. There's reddit articles. There's podcasts even pointed him to this. Podcast so he obviously was 11 00:02:48.640 --> 00:03:11.880 Chris Calhoun | XenTegra: really appreciative of that reality check of hey, yeah, you got me, I'm definitely interested in Nutanix, but I didn't want to oversell, so it was really enjoyable just to talk with him about where he was in his journey and his interest of Nutanix, so I would definitely say that that was one of my best customer conversations 12 00:03:12.040 --> 00:03:27.200 Philip Sellers: I love that. And you know we. It goes to the heart of why we do this podcast. And so many other things. Here all the free events, free education. We love an educated customer, so it does make it fun when when we get to have those conversations together. 13 00:03:27.620 --> 00:03:45.669 Philip Sellers: We're also joined by Ben Rogers. Ben is a sales engineer, senior sales engineer over at Nutanix in the Enterprise space. Long time, friend of Zintegra and myself. Ben. What's what's the coolest kind of conversation you've had recently with the customer 14 00:03:45.670 --> 00:04:06.120 Ben Rogers: Oh, this is gonna be a good one. I did a partner event a couple of weeks ago, and I had an opportunity to talk to a customer that's in a regulated industry. And they were telling us about we were talking about AI and how they were deploying AI, and they were telling us about how they were using drones for perimeter security. They have 15 00:04:06.120 --> 00:04:06.440 Ben Rogers: so 16 00:04:06.440 --> 00:04:28.450 Ben Rogers: highly regulated environments that they need to basically every now and then send a drone around the perimeter to make sure everything's cool. And so we started asking questions, like, you know, you're sending the drone out. What is it looking for. And it was basically like anomalies. So they fly around. Once they fly around twice, see something different. They send that to an AI engine. They figure out if they got to deploy security out there or not. 17 00:04:28.450 --> 00:04:36.590 Ben Rogers: And then we started talking about you know, where is that going to go? And they were like it could be all kind of things. It could be used for 18 00:04:37.280 --> 00:05:02.339 Ben Rogers: man. ISP providers wanting to look at outages and problems, at outages, instead of sending people out send drones out. And so all kinds of wild applications that we started talking about. But this this whole AI thing a couple of months ago at Gts. I kind of had a kick in the pants moment where I realized that AI was not just about how I communicate to customers about it. 19 00:05:02.340 --> 00:05:28.479 Ben Rogers: It's about how I'm interacting with it and how I'm learning it as an individual to basically keep up with with, you know, the environment in my time so interesting stuff, probably one of the most fascinating fun conversations I've had lately. What's cool about it, though, is AI directly goes into the conversation that we're having today, which is about Nkp and containers. Man, they're kind of one in the same. You can't have one conversation without having the other one 20 00:05:28.480 --> 00:05:53.334 Philip Sellers: Absolutely. Yeah. You you teed it up perfectly for us, and and you know it's funny you're talking about your your customer there. I've also heard of applications in the insurance industry anywhere that need inspections and stuff, you know, different industrial sites that need inspections, or even just you know homeowners insurance looking at a roof and things like that. There's all sorts of new applications and 21 00:05:54.350 --> 00:06:01.079 Philip Sellers: the the ability to augment photo and video and do analysis on it is really gonna change 22 00:06:01.820 --> 00:06:07.439 Ben Rogers: The way we interact and and do business in a lot of industries over the next decade. 23 00:06:07.920 --> 00:06:20.549 Ben Rogers: I mean, this is, this is kind of wild to think about. And, man, I'm really going to age myself here. I'm post 50, so I'm in the last quarter. I would just, I would assume the last quarter of my career. 24 00:06:20.680 --> 00:06:30.910 Ben Rogers: AI is like what electricity was when it was introduced, you know, whenever that was this, AI is going to change the way we live life 25 00:06:31.030 --> 00:06:55.990 Ben Rogers: and man the way the future is going to be. I mean, you know you talk about all kinds of things. I got a friend of mine that's a stockbroker that's talking about. He's looking at doing AI for all of his paper process. So they sign a contract, and the 180 documents they have to deal with and make sure eyes are totted and t's across. They're looking at AI, that so it's again. And I had a moment where I was like, I've got to be interacting into this and get myself immersed in it just 26 00:06:55.990 --> 00:07:03.439 Ben Rogers: just to stay with the times, man, so cool stuff interesting. But you know all part of day to day, and living in it 27 00:07:03.760 --> 00:07:09.085 Philip Sellers: That's it. Others on the call might say it's the best thing since sliced bread. 28 00:07:10.010 --> 00:07:21.810 Philip Sellers: I I didn't see his head pop up when I said that. But we got a bread lover on on the podcast with Chris, Calhoun. So we like to give him a hard time. He? He's a huge fan of sourdough bread 29 00:07:22.210 --> 00:07:23.500 Chris Calhoun | XenTegra: My heart bred. 30 00:07:24.235 --> 00:07:29.560 Philip Sellers: We may have had that name going back and forth in slack earlier today. So 31 00:07:29.560 --> 00:07:30.500 Ben Rogers: Never knew. 32 00:07:31.790 --> 00:07:32.630 Chris Calhoun | XenTegra: Absolutely. 33 00:07:33.010 --> 00:07:34.000 Philip Sellers: Well, yes. 34 00:07:34.000 --> 00:07:34.680 Chris Calhoun | XenTegra: Carbs. 35 00:07:35.150 --> 00:07:38.204 Philip Sellers: Hey? Yeah, I got a car upload from Mr. Calhoun. 36 00:07:39.000 --> 00:07:57.226 Philip Sellers: but you did. You? You teed it up perfectly for us today, you know, underneath a lot of these cool AI conversations is a different way of running infrastructure. And so today's blog post that we want to talk about is why you should run Nutanix Kubernetes platform on Nutanix cloud clusters so 37 00:07:57.570 --> 00:08:14.550 Philip Sellers: couple of different phraseology to type things here. Nutanix cloud customers also known as Nc, 2 which can run on azure or aws, and certain other cloud platforms. I know there's partnerships with with other service providers. 38 00:08:14.610 --> 00:08:31.080 Philip Sellers: And so it's the same nutanix that we know and love, just running inside of a hyperscale environment running on their bare metal hardware nodes connected into their network. And then, secondly, we've got new Nutanix Kubernetes platform. 39 00:08:31.340 --> 00:08:39.050 Philip Sellers: And I kind of want to just take this opportunity because we've mentioned Kubernetes a couple of times on the podcast. Here recently 40 00:08:39.320 --> 00:08:43.150 Philip Sellers: back up from that just a little bit. What is Kubernetes 41 00:08:43.900 --> 00:08:49.220 Philip Sellers: or K. 8. And and what does it actually do? And so 42 00:08:49.480 --> 00:09:04.489 Philip Sellers: for that, you know we'll we'll take kind of a 1 on one primer and say, you know, there's really 2 ways to run different applications. So we've got virtual machines which has been the typical way we run on nutanix on vmware. 43 00:09:04.680 --> 00:09:10.759 Philip Sellers: Yeah, everything lives inside that virtual machine. And it's a fully encapsulated 44 00:09:11.110 --> 00:09:14.659 Philip Sellers: operating system. So you've got the full stack there 45 00:09:14.760 --> 00:09:25.069 Philip Sellers: and then this new style of operating called containers came around where it's a very skinny 5. Think, Skinny Margaritas right? 46 00:09:25.820 --> 00:09:36.259 Philip Sellers: it. It's a skinny fide version of just what's needed to run the application. So now you don't have the full operating system. It's more of a just in time 47 00:09:36.670 --> 00:09:40.360 Philip Sellers: very reliable, very repeatable 48 00:09:40.960 --> 00:09:52.230 Philip Sellers: version of an application. Everything packaged. It needs to run. It comes online the same way every single time, so that you can deploy it out at scale. So 49 00:09:52.370 --> 00:10:05.389 Philip Sellers: with all these little containers running around, you need some way to manage those some way to connect them together some way to orchestrate them. And that's what Kubernetes is. Takes these containers. 50 00:10:05.700 --> 00:10:15.520 Philip Sellers: takes these little encapsulated applications that get deployed the same way every time, and allows you to scale those out, scale them up 51 00:10:15.740 --> 00:10:19.997 Philip Sellers: across environments and manage them in a way. 52 00:10:20.950 --> 00:10:36.239 Philip Sellers: that is highly orchestrated, so you can inject variables into those containers. But every one of those containers looks and acts the same as an old school admin, and I think, Chris, you probably will will 53 00:10:36.748 --> 00:10:48.310 Philip Sellers: confirm this, but as soon as you deployed a virtual machine it deviated from that golden master. Unless it was a Citrix machine pretty much there were differences right 54 00:10:48.310 --> 00:10:49.450 Chris Calhoun | XenTegra: Absolutely. 55 00:10:49.450 --> 00:10:50.510 Philip Sellers: And 56 00:10:50.680 --> 00:11:01.570 Philip Sellers: that's the age old problem is is, as soon as you make a clone of a virtual machine it starts to become a snowflake. It starts to change in ways where it's very inconsistent. 57 00:11:01.880 --> 00:11:04.389 Philip Sellers: Containers, gives you a way of 58 00:11:04.710 --> 00:11:13.889 Philip Sellers: making it very consistent every single time, so that that consistency then drives the same outcome every single time. So 59 00:11:14.880 --> 00:11:39.961 Philip Sellers: that's that's kind of the power behind all of this. But as we talk about Kubernetes and specifically Nutanix Kubernetes platform. We're talking about an orchestration layer, right? And so it helps us manage containers at scale, and it answers all the questions of how do we get traffic to them. How do we get metrics from them? And so we'll talk about all of that through today's podcast, 60 00:11:40.970 --> 00:11:54.901 Philip Sellers: But Dwayne Lesner wrote this blog post. He's a technical marketing manager over at Nutanix, and he he poses the question, why should you run Nutanix Kubernetes platform on Nutanix cloud clusters, and so 61 00:11:56.260 --> 00:11:59.170 Philip Sellers: Let's dive in and and see 62 00:11:59.270 --> 00:12:11.890 Philip Sellers: what he has to say. So off the top of it, Andy, I'll throw it to you. What is the primary reasons that Dwayne gives us for running Kubernetes platform on cloud clusters 63 00:12:12.930 --> 00:12:30.210 Andy Greene: I think it really comes down to Nutanix being a platform, right? So it's got the combination of virtualization tools that are built into Nc. 2. And then to add to that, we now have container based tooling that are built into Nkp. 64 00:12:30.210 --> 00:12:43.030 Andy Greene: so so by running all of this in you know, in in an ecosystem of technologies. Our clients get a consistent operating environment wherever they need, whether it's in the cloud, in the data center or at the edge 65 00:12:43.720 --> 00:12:49.735 Philip Sellers: So short version on that we can run both right. We can do both. That's a great benefit. 66 00:12:50.270 --> 00:13:04.059 Philip Sellers: Ben, you and I. We we had sort of a little bit of a banter back and forth before we started recording. And and there's an unofficial reason, too, that I think, is compelling, and that we should talk about. And what was that one 67 00:13:04.460 --> 00:13:13.100 Ben Rogers: So you're talking about? We're talking about, you know. Do you hypervise your container environment? Or do you not hypervise your container environment? And there's a big debate on that. 68 00:13:13.100 --> 00:13:37.360 Ben Rogers: Us at Nutanix feel like you need to hypervise that. So you're running ahv and ahv is ahv is ahv whether running on Prem. Or an Nc. 2 and Azure or Nc. 2 and aws. And the reason we feel like you need to run the Hypervisor is because that's what the Hypervisor was built for. It was built to manage the hardware and be able to slice and dice that hardware in an optimized fashion. 69 00:13:39.080 --> 00:13:48.660 Ben Rogers: contain Kubernetes can do that, but we don't think that that's what the purpose of it was built for. It was not built to run as a 70 00:13:48.660 --> 00:14:09.220 Ben Rogers: hardware manager. It was built to run, to be able to seamlessly deploy applications and orchestrate that. And so again, just going back to best practice, having Ahv is the underlying component of Nc. 2, and running Nkp on top of the Hv. Is going to give you the best use of that Iaas hardware that you're leasing from Microsoft. 71 00:14:09.220 --> 00:14:27.649 Ben Rogers: So to us it's just a better, more efficient way of managing the hardware which is then going to come down to managing cost, and I hate to say that Nc. 2 is a cost, effective solution. But in the end, when you do Rois on it, it is a cost, effective solution for running your vms, and also now your containers 72 00:14:27.840 --> 00:14:31.769 Philip Sellers: Yeah. So not only do you get that 73 00:14:31.870 --> 00:14:53.850 Philip Sellers: operational efficiency of having it consistent across the board, no matter the destination. But, as you said, you let the Hypervisor handle the discrete different node types, and then you can work on things above that in the stack. So you know, one of the things we kind of talk about with Kubernetes is, it's sort of like dial tone, right? It's 74 00:14:53.940 --> 00:15:05.219 Philip Sellers: it's sort of like electricity. You you turn the switch on, and it works. And so Kubernetes can be that kind of equalizer across clouds. 75 00:15:05.470 --> 00:15:22.309 Philip Sellers: And again, here it it's sort of that substrate where all your applications can run. It can be that consistent thing that you're managing with the hypervisor helping you manage the hardware differences below that. So I do think that's a compelling benefit 76 00:15:22.310 --> 00:15:32.600 Ben Rogers: Man. I'll give a shout out to my manager, Tom Powell here. What he says is, we give the ability to have vms and containers, both 1st class citizens in one cluster 77 00:15:32.600 --> 00:15:33.680 Philip Sellers: Absolutely. 78 00:15:34.040 --> 00:15:42.480 Philip Sellers: And Chris. There, there's another compelling reason that we can call out as well, and it has to do with kind of the core of Nutanix. Right? 79 00:15:42.790 --> 00:16:09.919 Chris Calhoun | XenTegra: Absolutely, and and that goes to exactly what both Andy says as Nutanix as a platform, and Ben just mentioned, hey? Both virtual machines and containers are 1st class citizens because of really the innate built-in capabilities of Nutanix. The desire for Kubernetes, as what's been mentioned is consistency. 80 00:16:09.920 --> 00:16:34.659 Chris Calhoun | XenTegra: The platform well, all of that consistency really revolves around the data services that are really needed by a Kubernetes or a containerized solution. If you think about it, the the benefit. There is the persistent storage storage performance, that's what the 81 00:16:34.770 --> 00:16:47.949 Chris Calhoun | XenTegra: app developers are looking for whenever they are looking for that landing zone. Whether it is an on-prem premises, environment, or a cloud model. They've got to have that 82 00:16:48.010 --> 00:17:04.480 Chris Calhoun | XenTegra: consistent tried and true performance from both. The storage level as well as offering up an objects based storage platform, too, that Nutanix does within their unified storage platform. That's 83 00:17:04.480 --> 00:17:22.400 Chris Calhoun | XenTegra: even though it obviously has many uses. And at the time that files and all of those others came out. I don't know that it was specifically geared or focused more towards Kubernetes, but now, seeing the 2, that's a match made in heaven. That's 84 00:17:22.599 --> 00:17:33.800 Chris Calhoun | XenTegra: butter and bread, if you will, in the sense that the 2 just go together really makes sense for customers to really focus on, hey? 85 00:17:33.910 --> 00:17:54.759 Chris Calhoun | XenTegra: I know the performance and capabilities of Nutanix. And I need that exact same performance capabilities in my data services platform. Let's let's merge the 2 together, and that's where I think Nutanix is a good fit for operating containerized environment. 86 00:17:55.360 --> 00:17:58.710 Philip Sellers: Good call back on the butter and bread. There, my friend. 87 00:17:58.710 --> 00:17:59.500 Ben Rogers: Ha! Ha! 88 00:18:00.560 --> 00:18:24.019 Ben Rogers: I think another thing that's cool, man going on with what Chris says is, if you look at the paragraph there that talks about the enterprise capabilities. We're bringing compression erasure, coding all those good things that we do at a disk system level with Aos. We're bringing that to the cloud, and those are normally features that are add-ons all. I cart to your data services. So again, we're just bundling all that in into our license. 89 00:18:25.260 --> 00:18:30.196 Chris Calhoun | XenTegra: To be point. Let me say this, too, because if you think about how 90 00:18:30.760 --> 00:18:44.150 Chris Calhoun | XenTegra: public cloud models often are a shared infrastructure, and without that dedicated performance. That's what's gonna be lacking. And obviously, that's 1 of the core tenants of 91 00:18:44.370 --> 00:18:50.680 Chris Calhoun | XenTegra: Nutanix running in a public cloud is having that dedicated environment and 92 00:18:51.140 --> 00:19:06.990 Chris Calhoun | XenTegra: Utopia honestly of resources. So having those reserve dedicated instances for performance and not a shared environment is really key to foundational success. For Kubernetes 93 00:19:07.560 --> 00:19:29.690 Philip Sellers: Well, and and that that's a key point. Right? Chris. So you talk about the dedicated clusters. What does that actually translate into to an enterprise at a business level. And that's control. Right? You get to control your stack, you get to control the version of software you're deployed on and you have guaranteed performance levels 94 00:19:29.690 --> 00:19:43.580 Philip Sellers: that are different than you get in the shared computing models of the Hyperscale native environments. And so you you get to expect the same sort of things that you you have from a guarantee of service 95 00:19:43.710 --> 00:19:46.709 Philip Sellers: on-prem in your own co-location. 96 00:19:47.804 --> 00:19:56.449 Philip Sellers: Now, with the flexibility of of having someone else's regions, locations and a lot of other benefits. 97 00:19:57.063 --> 00:20:07.659 Philip Sellers: Let let's dig into this. So, as the blog post goes on, we talk about it as a cost. Effective solution. 1st thing here is is talking about available capacity 98 00:20:08.360 --> 00:20:13.849 Philip Sellers: in C 2. You know, we can scale up to 13 nodes. And 99 00:20:14.200 --> 00:20:27.700 Philip Sellers: when we talk about scale in a Kubernetes environment, this is an interesting call out that Dwayne makes but Kubernetes clusters by default only support 110 pods per node. So a little bit of terminology here a pod is 100 00:20:28.170 --> 00:20:30.029 Philip Sellers: the concept of 101 00:20:31.510 --> 00:20:59.149 Philip Sellers: a management domain inside of a Kubernetes cluster. So across multiple different machines. You create your pods. But there are some limitations here. So Dwayne calls out, as you scale you risk wasting resources with larger nodes. So this goes to Ben's point about the hypervisor. Now you mix and match vms with Kubernetes workloads. You get more efficient use of your larger nodes 102 00:20:59.200 --> 00:21:09.090 Philip Sellers: in your larger hardware. So I do think that that definitely supports kind of your theory about introducing the Hypervisor as an advantage 103 00:21:09.617 --> 00:21:13.592 Philip Sellers: to the overall system. So from a cost, effective standpoint. 104 00:21:14.600 --> 00:21:17.583 Philip Sellers: that that definitely makes sense. 105 00:21:18.510 --> 00:21:35.150 Philip Sellers: it would also make sense that we can engineer multiple instances and greater scale on a hypervisor, because just like with virtual machines, we're breaking up this Kubernetes workload into multiple different workload domains across the same host. 106 00:21:35.520 --> 00:21:37.880 Philip Sellers: So we get both of those. 107 00:21:38.558 --> 00:21:47.779 Philip Sellers: Then you called out. The next thing that Dwayne talks about, which is storage we, you know right now in Nc. 2, everything is 108 00:21:48.210 --> 00:21:55.620 Philip Sellers: Nvme. So we get not just great performance, but extremely great performance. 109 00:21:56.380 --> 00:22:01.219 Philip Sellers: Do you want to talk a little more about storage when it comes to the cost? Effectiveness 110 00:22:02.280 --> 00:22:22.330 Ben Rogers: Sure. So you know, the other thing about the storage that I'd like to say is, you know, we also have. You know, the replication services behind it. All that that we we're gonna talk about that later in the thing. But again, it's just these are high speed performance servers that we're we're leveraging in Nc. 2 in azure and in aws. So you know. 111 00:22:22.430 --> 00:22:38.130 Ben Rogers: managing the micro waste, having the flexibility of adding nodes and depleting nodes as you need to scale up and scale out or scale in, and then also having the performance with the Nvme and SSD. In the nodes. 112 00:22:38.130 --> 00:23:01.699 Ben Rogers: it's just a great solution. And all these are things that if you looked at it from a native cloud, let's say you went to Azure, and you tried to put all these things into, you know a list you would be paying a la carte for most of these things which would make your cost very much, Skyrocket. So again, we're delivering in one license blend all of the services that you really need to run this simply 113 00:23:01.700 --> 00:23:21.160 Ben Rogers: with flexibility and with performance. And I think that's really what Dwayne's getting here with most of this, and on the performance side is that this comes with the Aos. Nc. 2. License day one. These are not services you have to add on. These are not services you have to purchase. They're going to be there for you the 1st day that you come into the environment 114 00:23:22.520 --> 00:23:30.799 Philip Sellers: Yeah, absolutely. Your data services is is a huge, huge benefit. And you know, as we talk about 115 00:23:31.290 --> 00:23:37.269 Philip Sellers: replication and other things here shortly. That really is going to come into kind of play. 116 00:23:37.490 --> 00:23:43.720 Philip Sellers: Networking is also a a huge concern, Andy. I'm gonna throw it to you from a a networking 117 00:23:44.915 --> 00:23:45.590 Philip Sellers: conversation 118 00:23:46.600 --> 00:24:10.060 Andy Greene: Yeah, well, I think Kubernetes is very chatty over the network, managing the management. The worker nodes, having everything spread out across different availability zones. Our clients end up paying for the data that's transferred between, say, the management and the worker nodes as well as all of that traffic between the cluster and the other cloud services as well. 119 00:24:10.110 --> 00:24:33.699 Andy Greene: You know, I think what we're seeing here is probably a good example of data gravity in action. We know that applications, services, and data tend to stick together, and data gravity itself can have a little bit of maybe a negative connotation for making it difficult and sometimes costly to move our applications and move our data in an efficient manner. 120 00:24:34.072 --> 00:24:57.049 Andy Greene: But what Nutanix is really demonstrating? Here is some of the value of bringing all of the data, the applications, and the services into that single platform. So you know, by doing that, we're gonna see the the benefits that we've been talking about performance cost scalability, mobility. You know all of that through having these services and and these applications on a single platform 121 00:24:57.920 --> 00:25:01.040 Philip Sellers: Yeah, I I agree with that. You know. A lot of times 122 00:25:01.480 --> 00:25:09.069 Philip Sellers: I call these cloud soft costs, and you can't see me if you're listening, but I've got my air, quotes my fingers doing air quotes. 123 00:25:09.650 --> 00:25:20.669 Philip Sellers: but these are the cloud soft costs, right. We we don't do a good job of estimating, and I'll say that even for you know us as a team working with our our clients. 124 00:25:20.970 --> 00:25:26.160 Philip Sellers: we sometimes can't quantify all of these incremental 125 00:25:26.330 --> 00:25:33.179 Philip Sellers: small costs that add up when you're running in hyperscale, either aws azure Gcp. 126 00:25:33.350 --> 00:25:40.006 Philip Sellers: and and so these egress costs around networking can definitely make a huge monetary impact. 127 00:25:41.060 --> 00:25:49.320 Philip Sellers: to your core point around data gravity. This is definitely data gravity in in full force. 128 00:25:50.750 --> 00:25:54.199 Philip Sellers: You know. The the other thing that you said, too, is. 129 00:25:54.750 --> 00:26:12.250 Philip Sellers: you've got a full package here, and so I want to call out, if you're listening to this definitely, look up the blog post on nutanix.com slash blog, because there's a great graphic here, and I call it the Nkp sunburst. 130 00:26:12.842 --> 00:26:17.000 Philip Sellers: But I really like this sort of graphic, because 131 00:26:17.260 --> 00:26:31.579 Philip Sellers: there's about 12 different layers inside of Kubernetes. And so we talk about Kubernetes as one thing. But this does a good job of kind of breaking it up. And I'm just going to recap some of these. Not all. But you know you have to make a networking 132 00:26:32.110 --> 00:26:40.250 Philip Sellers: choice. You have to make a storage choice. You have to make a security service mesh, ingress delivery, controller 133 00:26:40.360 --> 00:26:52.140 Philip Sellers: cost management, observability, automation. All of these are different choices you have to make when you're doing a Kubernetes strategy. The great thing about Nkp. And what we're getting here is 134 00:26:52.350 --> 00:27:05.030 Philip Sellers: you can use the built in because Nutanix is delivering an opinion, a choice of each one of these different 12 layers, or you can plug in your own. There's support for other 135 00:27:06.330 --> 00:27:14.390 Philip Sellers: Cncf. The Cloud native computing foundation projects that can plug in within. Kp, so 136 00:27:14.560 --> 00:27:28.939 Philip Sellers: nkp is very what we call a pure sort of product stack. It's very close to the open source. There's not a lot of customization. There's not a lot of changes to what the open source Kubernetes project is. 137 00:27:29.910 --> 00:27:38.710 Philip Sellers: There are certainly enhancements and reasons to buy Nkp, but it's a very pure sort of deployment. And so you're getting that 138 00:27:39.160 --> 00:27:46.049 Philip Sellers: close adherence to open standards with all of the necessary layers 139 00:27:46.230 --> 00:28:01.350 Philip Sellers: to be able to fully bake your Kubernetes strategy. And so that's huge. Right, I I would say in less words. They're delivering the staples easy button for Kubernetes. You remember that one that was easy 140 00:28:01.510 --> 00:28:05.519 Ben Rogers: And that's pretty powerful because we go into a lot of 141 00:28:05.700 --> 00:28:17.999 Ben Rogers: you know, K. 8 shops that they're struggling to make these decisions. You know. I mean, you've got the K 8 environment, and then you've got you know what you're calling the sunburst around it. And to your listeners I would go and look at this graphic, because 142 00:28:18.100 --> 00:28:39.039 Ben Rogers: basically what this graphic is showing us is all the things you got to have to really have a full blown Kubernetes environment, and with it being open source. That's a great thing. I love open source. But it's also a complex thing, because then you have to make decisions on what's best for your environment. And we take that complexity out of you by giving you a platform 143 00:28:39.140 --> 00:28:51.059 Ben Rogers: making these decisions for you. Now, you're not bound to these decisions. These are decisions that we make. You can change them if you want. But if you decide to stay under our umbrella, not only are you getting a complete platform day one. 144 00:28:51.060 --> 00:29:14.090 Ben Rogers: But you're getting the lifecycle management behind that platform. So if you stay in this sunburst that Phillips called it, you then have management ability from then on, because, as these things, you know, enhance and grow and have changes made, you're going to want those to be brought into your environment. And you're not going to want to deal with these as individual pieces. So again, we're bringing that under a sunburst umbrella. 145 00:29:14.090 --> 00:29:28.769 Ben Rogers: we're giving lifecycle management. We're giving you the best of practice that we know of. And Nkp, the guys that came along with that I'm going to give a little history lesson here. That was a merger with a company called D 2 IQ. Which was mesosphere 146 00:29:28.770 --> 00:29:55.379 Ben Rogers: and man. These are guys, you know, Toby, Dan and Bruce. These are guys that were ground floor of the Cncf foundation. I mean, these are guys that know Kubernetes inside and out. It is their lifestyle. And so I feel very good and very opinionated when we bring this product in, because I know it is the collection of services that somebody needs to operate a Kubernetes environment day one, and it is a platform, and it can be managed. 147 00:29:55.380 --> 00:30:14.039 Ben Rogers: Post day, one in the day, 2 day, 30 day, 60 day, 3, 65. So again, a powerful platform that is built to run on our Hypervisor day one. So, man, I am very proud that we're we're we have the ability to supply this to our customers. 148 00:30:14.750 --> 00:30:23.680 Philip Sellers: Yeah, you know it it is, and not to steal your thunder there, I mean it. It is made to do amazing things on the Nutanix platform. 149 00:30:23.800 --> 00:30:31.030 Philip Sellers: but it's also still open. This runs on Aws on azure, on Gcp. 150 00:30:31.190 --> 00:30:32.080 Philip Sellers: On bare metal. 151 00:30:32.500 --> 00:30:57.639 Philip Sellers: There's lots of choice on vmware. There's lots of choice for Nkp. Now you may not receive all of the goodness. And so I think that's the core of what we're talking about here in this blog post. Is, it really is a better together story, right? Nkp, plus the nutanix cloud infrastructure provides you with extra benefits. And so a lot of that does kind of come back to. 152 00:30:57.640 --> 00:31:13.600 Philip Sellers: You know the things we talked about data services, the hypervisor, the portability, the orchestration, the easy button, if you will, across the platform to make it easier to to operate the same way 153 00:31:13.600 --> 00:31:19.600 Ben Rogers: I'm gonna throw one more in there, 90 plus Nps score on your support. You get in trouble. You can call us. We'll help you 154 00:31:20.410 --> 00:31:29.370 Philip Sellers: Yeah, best best in class support that that has to go a long way when so many of us are frustrated with 155 00:31:30.250 --> 00:31:35.360 Philip Sellers: how poor support is in in a lot of places inside the it industry. 156 00:31:35.930 --> 00:31:50.170 Philip Sellers: Let's let's talk a little bit about this. When Dwayne goes on to talk about mobility, the freedom to move across environments. Chris, I'm gonna kind of send this one over to you and and let's talk mobility a little bit, because this is a huge 157 00:31:50.500 --> 00:31:58.300 Philip Sellers: benefit to the business. The stakeholders that you're working with as a technologist 158 00:31:59.040 --> 00:32:08.420 Chris Calhoun | XenTegra: Absolutely. And this ties in, too, for some of our past few conversations in our podcast around, the 159 00:32:08.600 --> 00:32:28.636 Chris Calhoun | XenTegra: capabilities from a business perspective of being able to have the freedom to not get locked into a long term. Technical and financial decision based off of where's my data located? Going back to Andy's statement of data gravity. 160 00:32:29.230 --> 00:32:48.759 Chris Calhoun | XenTegra: with Nutanix, you do have that freedom and that flexibility. And that's really an overarching longer discussion and relates back to like we said past podcasts. But I really do believe that with a containerized operating model 161 00:32:48.850 --> 00:33:04.019 Chris Calhoun | XenTegra: that allows and lends itself to both both business agility and technical abilities too. So think about this. What if it's a global based company that has 162 00:33:04.050 --> 00:33:23.149 Chris Calhoun | XenTegra: specific needs for their industry, their business to keep regionalized data, but through mergers and acquisitions may have brought on other companies. With those other companies they may need to run independent. And for 163 00:33:23.150 --> 00:33:34.368 Chris Calhoun | XenTegra: that global company one cloud vendor meaning aws may be better suited for European based workloads for this customer, let's say. But if they have any 164 00:33:34.820 --> 00:33:54.029 Chris Calhoun | XenTegra: any part of business dealings within the States, maybe it's Microsoft azure that they need to leverage just because of Elas and those contracts that we've talked about in the past. Well, with Nutanix as a whole, just staying high level and not giving away the goodness of 165 00:33:54.040 --> 00:34:06.999 Chris Calhoun | XenTegra: the the next couple lines in the in the blog. That's the freedom and flexibility that you're looking for from the business perspective mindset is, do we have more than one choice? 166 00:34:07.110 --> 00:34:23.110 Chris Calhoun | XenTegra: And with that does that lend us freedom over the next year or 2 to have expansive capabilities in other markets. Even so, I think that that's really a key about this technology as as a whole. 167 00:34:23.620 --> 00:34:28.200 Philip Sellers: Well, and and the fact that we're talking hyperscale here. But 168 00:34:28.300 --> 00:34:37.490 Philip Sellers: what if this strategy needs to be your edge? What if it needs to be close to manufacturing floor. We're talking about the same consistent stack that can go 169 00:34:37.790 --> 00:34:43.439 Philip Sellers: cloud to edge and give you really the maximum amount of flexibility 170 00:34:44.570 --> 00:34:47.049 Philip Sellers: where you run and choose to run things. 171 00:34:47.590 --> 00:35:05.939 Philip Sellers: Dwayne calls this out at the heart of data. Mobility is the nutanix data services for Kubernetes or Ndk Ndk is that thing that helps connect the core storage system to kubernetes. And that helps specifically with containerized workloads 172 00:35:06.623 --> 00:35:16.539 Philip Sellers: so that we get things like copy data management data protection. Dr. You know, these are all things that we take as table stakes. 173 00:35:16.660 --> 00:35:29.169 Philip Sellers: but they change in a Kubernetes strategy versus what we're used to doing in terms of Dr. For virtual machines, Andy, I'm curious, you know. How does that look with Ndk and Kubernetes? 174 00:35:30.110 --> 00:35:57.939 Andy Greene: Great question. So you know, within Dk, we can take a snapshot that's going to include all of the application configurations, things like the Kubernetes resources, stateful sets, services, config maps, and secrets. And then through that snapshot we're able to restore the application to a different site if we needed to. So for Dr. Purposes, for data mobility purposes, we've got that ability to pick up our application and move it into another site. 175 00:35:58.490 --> 00:36:13.629 Andy Greene: You know, this is contrary to some of the cloud native block storage that's going to require 3rd party tools, higher Rtos, for replication. You know, Ndk really simplifies the process without adding any additional overhead 176 00:36:14.220 --> 00:36:38.279 Philip Sellers: Well, you know it is. You know I've I've heard it said like the perfect network is a network without users. Right? You know it. It is the perfect application and application without data. The same sort of thing like to me, data is the crown jewels. We've got to be able to protect that and move that around. And that's essentially what you're saying here is that Ndk 177 00:36:38.570 --> 00:36:46.330 Philip Sellers: gives us that ability with persistent volumes inside of Kubernetes workloads 178 00:36:46.810 --> 00:36:47.810 Andy Greene: Absolutely. 179 00:36:48.200 --> 00:37:06.760 Ben Rogers: Now you got to remember Kubernetes is not just used for applications. They're starting to containerize databases as well. So when you look at, you know. Now you've got a mongo database that's running inside a container. Well, guess what you got to protect that container, because now you got active data that's being written to it, and being read from it. So 180 00:37:06.760 --> 00:37:23.140 Philip Sellers: Absolutely. And you're not just talking the new style databases oracle. SQL. Server. There's a SQL. Server that runs in a Linux that can run in a container. There, there's all sorts of things. But you're right, right. That persistent data is the key 181 00:37:23.140 --> 00:37:31.709 Ben Rogers: And a lot of people, a lot of a lot of people kind of forget that. That's kind of an after fact of oh, we've gotten there. Now, what? And so 182 00:37:31.710 --> 00:37:55.300 Ben Rogers: we try to bring that point early on is, how are you not only deploying your container environment, but how are you protecting your container environment. And if you did have a disaster, how are you getting that back into production as quickly as you can? And that's where Nutanix is really starting to shine, because, knowing where the zeros and ones are and how to handle those and how to get those protected and restore. 183 00:37:55.300 --> 00:37:59.960 Ben Rogers: That's where we started from, and we continue to build on that goodness 184 00:38:00.230 --> 00:38:14.439 Philip Sellers: Yeah 100%. And you know, it's it's cool that you started talking AI at the beginning of the podcast. Because, we're gonna come full circle again. Here, as Chris was talking about the benefits of the platform you mentioned object store. 185 00:38:14.690 --> 00:38:27.713 Philip Sellers: But when it comes to AI workloads files is a huge, huge benefit right and a lot of AI. Most AI is written on Kubernetes it. They're written in containers. And so 186 00:38:28.370 --> 00:38:41.989 Philip Sellers: we did a whole podcast episode, Ben Gyro and myself talking about Ml. Perf storage and the amazing amount of throughput and performance you can get on nutanix files. 187 00:38:42.120 --> 00:38:50.549 Philip Sellers: training and using files with the AI workload. So if you haven't listened to that, podcast it's a few episodes back. 188 00:38:51.091 --> 00:38:53.740 Philip Sellers: But look at the episode about Ml perf. 189 00:38:53.930 --> 00:38:59.339 Philip Sellers: But that just goes to kind of bring us full circle. Right? You've got a suite of product 190 00:38:59.490 --> 00:39:01.440 Philip Sellers: around Kubernetes 191 00:39:01.680 --> 00:39:10.850 Philip Sellers: that enables the business to provide cloud services on your terms the same way, no matter what destination you choose. 192 00:39:11.381 --> 00:39:26.929 Philip Sellers: And that takes us right into our next point, which is operational consistency across on premises and the cloud. So, Chris, I'm gonna throw it to you for organizations with Nutanix and Nkp. What? What's the huge advantage here from an operational standpoint? 193 00:39:27.800 --> 00:39:43.629 Chris Calhoun | XenTegra: To me, it goes back to a similar talk track from some of these other podcasts around, having separate teams manage separate environments. It's basically one where you can now focus on 194 00:39:43.760 --> 00:39:49.420 Chris Calhoun | XenTegra: having a operational consistent approach to 195 00:39:49.520 --> 00:40:03.600 Chris Calhoun | XenTegra: a business workload. And that's really the idea of having a seamless environment of support and management, whether it's cloud or on Prem or at the edge, as we've all mentioned. 196 00:40:03.620 --> 00:40:19.989 Chris Calhoun | XenTegra: and that's the model that you're really looking for is not to have those siloed teams where one group manages one thing, another group manages a different location of cloud operating model or to something to that effect. It's 197 00:40:20.040 --> 00:40:31.799 Chris Calhoun | XenTegra: it's staying in an agile environment where everything seamlessly works. And it's managed the same way. So that's the consistency from 198 00:40:31.890 --> 00:40:40.119 Chris Calhoun | XenTegra: an infrastructure perspective, a support. Model. That's really what you're looking for. That consistency across the board 199 00:40:40.500 --> 00:40:41.110 Philip Sellers: Yeah. 200 00:40:42.140 --> 00:40:57.449 Philip Sellers: yeah, not having to retrain your team. And I think the fact that in general on an Nci club or a new tanks cluster. In general, we we do what about 15 min of training and most administrators can pick it up from there 201 00:40:59.380 --> 00:41:27.159 Ben Rogers: Man. I just had my 14 year old go through move, and he was, you know, outside of some terms he didn't understand. He didn't know what a Vlan was, but he was able to sit down and kind of go. Okay, here's my source. Here's my target. Here are the objects I want to move from one to the other, and it was. It was fascinating for me to sit down and go. How easy is this, you know, and he picked it right up. I would like to add on to what Chris was saying. You know, as far as consistency across on premise and cloud 202 00:41:27.180 --> 00:41:45.570 Ben Rogers: that can be. Hv, so you know, if you're running Hv. On, prem, you're running Hv. And Nc. 2. That's going to give you a nice seamless landscape, but it don't stop there. If customers are running Eks Aks in the Google, the container platform in Google. 203 00:41:45.570 --> 00:42:03.589 Ben Rogers: we can bring that in Nkp, and we now can manage all of those silos for you under one umbrella, so it don't just have to be on hv, it just doesn't have to be on nutanix infrastructure it can be in native azure. Infrastructure aws infrastructure. 204 00:42:03.590 --> 00:42:20.340 Ben Rogers: That brings the point that Chris made with you. Now take silo teams and bring them together as one unit, and you don't have an azure specialist or an aws specialist. You have a nutanix specialist that can work in any silo that you choose to deploy 205 00:42:21.360 --> 00:42:24.540 Philip Sellers: Yeah, I mean that that's the heart of Nutanix, right choice 206 00:42:26.000 --> 00:42:39.390 Philip Sellers: choice at all different levels and layers. And this is another great example of that. But you know as Andy's I mean, as Chris said, silo busting from the beginning right breaking down the barriers between 207 00:42:39.820 --> 00:42:48.220 Philip Sellers: compute and storage teams, and that that's been going on since day one since you introduced your 1st product. So it it's 208 00:42:49.010 --> 00:43:17.810 Philip Sellers: good to see that consistent strategy really play out right at the end of the day. What you're buying, and and Dwayne wraps it up this way. I'm just gonna read what he says by running Nkp on Nc. 2, you gain a future proof environment for all your applications. Vms and containers. I I couldn't sum it up better than that. So I'm just gonna read the words from Dwayne. And I think it's it's the perfect way, right? You're giving yourself a future proof 209 00:43:17.950 --> 00:43:25.299 Philip Sellers: solution because you're giving yourself choice at the end of the day, and I think that's the ultimate thing. That kind of 210 00:43:25.420 --> 00:43:32.030 Philip Sellers: has been the thread throughout all of today's conversation is you're giving yourself choice. 211 00:43:33.430 --> 00:43:55.990 Philip Sellers: Oh, guys, I I think this has been a really great conversation. I'm super happy to really get to dive in on Nkp, a little bit more. If Kubernetes is new for you, and and that's somewhere that you want to talk more reach out my team. We're all here to have that conversation to help you and your organization 212 00:43:56.160 --> 00:44:18.060 Philip Sellers: with strategy around Kubernetes around containers. This is a journey that that I think all of us have been on either on the client side or working with our clients over the last 5 years. As this strategy as this technology is matured, so we'd love to talk to you more, one on one, and 213 00:44:18.857 --> 00:44:29.972 Philip Sellers: you know, if there's enough interest, maybe we'll we'll throw out a a webinar or some other group event. So do reach out. If you've got questions, we're always available and 214 00:44:30.640 --> 00:44:40.490 Philip Sellers: for everyone listening. Thanks for spending a little bit of time with us hope you have a great day, and we'll catch you on the next episode. Thanks for listening