Latest / Nutanix Weekly / Nutanix Weekly: Nutanix and Canonical Partner to Simplify Kubernetes Deployments with NKP and Ubuntu Pro

Transcript

Nutanix Weekly: Nutanix and Canonical Partner to Simplify Kubernetes Deployments with NKP and Ubuntu Pro · Nutanix Weekly

WEBVTT 1 00:00:04.000 --> 00:00:19.699 Philip Sellers: Welcome to another episode of Nutanix Weekly. I'm your host, Phil Sellers. I'm also the practice director for Modern Data Center here at Zintegra and want to say thanks for spending some time with us. This is another one of those 2 00:00:20.280 --> 00:00:24.729 Philip Sellers: content with context sort of situations. We like to 3 00:00:24.910 --> 00:00:52.130 Philip Sellers: to take the content we find online from our partners and add our real world context, our experiences and anything we're hearing from our clients collectively and add that to learn together. So I want to say, Welcome and introduce my co-host for today. Jayra Cox, architect at Nutanix and all around. Good Guy. 4 00:00:52.400 --> 00:00:59.219 Jirah Cox: Phil, who's your peer? At integra? That would be the practice director for archaic Data center. 5 00:00:59.400 --> 00:01:01.130 Philip Sellers: Archaic data center. 6 00:01:01.130 --> 00:01:05.909 Jirah Cox: Your modern data center right? Then, that implies somebody else is doing the Stone Age stuff. 7 00:01:05.940 --> 00:01:14.669 Philip Sellers: Well, I was. Gonna say, I think it's that statue of Matthew we've got in the office, or something, you know, just posed over there made out of marble. 8 00:01:16.220 --> 00:01:19.140 Jirah Cox: The classic Renadon's data center. I think. 9 00:01:19.390 --> 00:01:20.160 Philip Sellers: Yeah. 10 00:01:20.560 --> 00:01:33.857 Philip Sellers: yeah. Classical data center that that's got a ring to it. Right? You know, it just seems like something you would do with a gray beard. And what do you call the little spectacle like Mr. Beanut? 11 00:01:34.190 --> 00:01:34.850 Jirah Cox: Monocle. 12 00:01:34.850 --> 00:01:36.629 Philip Sellers: A monocle. Yes, that just. 13 00:01:36.630 --> 00:01:37.209 Jirah Cox: I mean, if you. 14 00:01:37.210 --> 00:01:37.930 Philip Sellers: Yes, I mean. 15 00:01:38.320 --> 00:01:41.332 Philip Sellers: I mean, between you, me and the recording. If you launch that position, please let me know. 16 00:01:43.460 --> 00:01:56.780 Philip Sellers: Director of Classical Data center. Well, this is great. I mean, we we do talk about modern data center, modern sort of things. An awful lot. And and that's on purpose. We. 17 00:01:56.850 --> 00:02:16.100 Philip Sellers: we think that the data center has changed. And there are a lot of ways you can choose to get the job done. But we advocate for a simplistic way with lots of choice and a lot of power. So that brings our friends, like Gyra from nutanix onto the podcast to talk about 18 00:02:16.200 --> 00:02:18.020 Philip Sellers: all things going on. 19 00:02:19.720 --> 00:02:24.940 Philip Sellers: But this will be a kind of interesting conversation, because this isn't 20 00:02:25.290 --> 00:02:29.059 Philip Sellers: necessarily the traditional world of Nutanix is a gyra. 21 00:02:30.446 --> 00:02:40.840 Jirah Cox: I guess not. I mean to be to be coy and and have the audience not read into what we're talking about. Yes, it's it is, it's newer to anything, for sure. 22 00:02:41.030 --> 00:02:44.320 Philip Sellers: Yeah, but it's a good thing. I mean, it still follows 23 00:02:44.690 --> 00:03:13.480 Philip Sellers: the core tenets of choice simplicity and performance. That sort of underlie everything Nutanix does. I don't know that that's an official 3 word tag, but I think it describes what Nutanix does really? Well, so we'll pull back the suspense here and and read you into what we're talking about. We're reviewing a blog post from May 7th on nutanix.com slash blogs. 24 00:03:13.650 --> 00:03:26.130 Philip Sellers: Title of today's Post is nutanix and canonical partner to simplify Kubernetes deployments with Nkp and ubuntu. Pro wow. 25 00:03:26.290 --> 00:03:27.989 Jirah Cox: It's awesome. Yeah, it's very exciting. 26 00:03:27.990 --> 00:03:32.229 Philip Sellers: With a Linux provider, so. 27 00:03:32.230 --> 00:03:41.330 Jirah Cox: And I mean, I'll even say not. I'm not. Gonna assume. This is part of the business negotiations, but my personal favorite distro of choice for the last 15 years. 28 00:03:41.330 --> 00:03:52.400 Philip Sellers: That that it's near and dear to my heart. You know, I learned on on Red hat varieties many, many moons ago. But yeah, let's let's be honest like. 29 00:03:53.450 --> 00:03:57.329 Philip Sellers: go a little nerdy here may lose a few listeners, bear with us, but. 30 00:03:57.330 --> 00:03:58.579 Jirah Cox: On this podcast. 31 00:03:58.580 --> 00:03:59.982 Philip Sellers: I don't know. 32 00:04:00.760 --> 00:04:06.419 Philip Sellers: rpm was an amazing thing. Package management and things like that with red hat was was 33 00:04:07.050 --> 00:04:12.500 Philip Sellers: incredible to me as a young person trying to learn Linux 34 00:04:12.630 --> 00:04:15.850 Philip Sellers: and Ubuntu got all those things right, you know. 35 00:04:16.360 --> 00:04:24.888 Philip Sellers: And you know the package management is is fantastic. And so yeah, it's my preferred distribution, too. 36 00:04:25.650 --> 00:04:32.109 Philip Sellers: I think they're just also a really good community partner in general. And so I I like them for that. 37 00:04:32.450 --> 00:04:43.449 Jirah Cox: They? They definitely were, I think, in my experience, where? Yeah, early on in the community building. Well, okay. Sorry for the nerdy, pop, quiz. Phil, do you know, what do you know what ubuntu means? 38 00:04:44.386 --> 00:04:52.100 Philip Sellers: I don't specifically know what it means. I do know that it is a a term from Africa. 39 00:04:52.100 --> 00:04:52.840 Jirah Cox: True. 40 00:04:53.330 --> 00:04:56.940 Philip Sellers: Not mistaken. Remember what it means. 41 00:04:56.940 --> 00:05:00.660 Jirah Cox: It means. I am what I am, because of what we all are. 42 00:05:00.810 --> 00:05:01.720 Philip Sellers: Yeah. 43 00:05:01.720 --> 00:05:27.640 Jirah Cox: And it so the community aspect of it is, even in the naming of the distro. I used to actually send off for the free the free Dvds right? Every single new, you know, rodent named release rodent, or I guess animal animal kingdom named release and for a while I, I, you know, had them on my bookshelf. And then I realized that was really quite silly. But but yeah, they were great. So the community is right there. 44 00:05:27.996 --> 00:05:35.480 Jirah Cox: baked into even the naming of of of ubuntu, and not to get terribly inflammatory here. But yeah, it's like 45 00:05:35.520 --> 00:05:54.800 Jirah Cox: usable, friendly, Debian. Sorry. Sorry, Debian. Diehards. But what do you mean? Debian is not friendly and usable. You know. Check out our forums. Okay, we'll stop that. But but yeah, and you know the the other funny 46 00:05:54.800 --> 00:06:13.320 Jirah Cox: high in here for the this. This theming of okay, of course, the the headline announcement is, you can now use use ubuntu and really ubuntu pro right. The actual commercially supported variant, not the community supported variant as the underlying substrate OS for running the Nkp container platform. 47 00:06:13.470 --> 00:06:18.113 Jirah Cox: which is great and fantastic for all of our joint customers. 48 00:06:18.990 --> 00:06:33.470 Jirah Cox: I scroll past a meme earlier today in in other nerd circles. Right? That was like, Oh, it's serverless. But then you look on the covers and it runs on servers right? And so the the tie in here would be okay. It's containers, so I don't care about the OS. 49 00:06:33.950 --> 00:06:38.370 Jirah Cox: And underneath that is, the OS. Is an OS. 50 00:06:39.100 --> 00:06:48.379 Philip Sellers: Yeah. And and here here I'm gonna ask the stupid question, why would you want to run Nkp in an ubuntu vm on Nutanix? 51 00:06:49.680 --> 00:07:02.899 Jirah Cox: But or anywhere, right in a Vm or or anywhere. Yeah, for sure. So and I guess here's my simplistic answer back is because so the containers have to run on Kubernetes. Kubernetes has to run on an OS. Somewhere. 52 00:07:03.150 --> 00:07:08.360 Jirah Cox: You could run that on like a free, open source. Community supported Linux distro 53 00:07:08.830 --> 00:07:25.459 Jirah Cox: that option does not go away. Just because there's now this also strong partnership with canonical for vin 2 pro but folks like security teams. Right? Folks like it. Leadership. Love to be able to say there's support at every level of the stack 54 00:07:26.006 --> 00:07:52.269 Jirah Cox: or we get better than say, free, open source software. Sla's around things like Cv patching right? Vulnerability management. You know, rapid mitigation. And so that's that's to to skip a little bit ahead to some of the the F's. And B's, as they say. You know of what this brings to the to the stack. That's something that you can do pro has in spades over over something that's, you know, would be a community based distro. 55 00:07:53.080 --> 00:08:12.590 Philip Sellers: Yeah. And and 2. I I think you hit the nail on the head. 1st of all is that long term commercial support that's necessary for the enterprise, and for really any business, right security has to be at the forefront. So getting those vulnerabilities patched, not having, you know, as aggressive 56 00:08:12.710 --> 00:08:28.080 Philip Sellers: of an upgrade cycle as open source sometimes forces upon people. These are problems that ubuntu pro solves their long term releases not unlike the model that Nutanix has had with their long term leases 57 00:08:28.581 --> 00:08:43.860 Philip Sellers: releases have have always helped with that from a support, ability standpoint. But you mentioned naming to, I think briefly, there, I love the naming. 25, 0, 4. Right now is plucky puffin. I mean. 58 00:08:43.860 --> 00:08:45.340 Philip Sellers: come on like. 59 00:08:45.980 --> 00:08:53.440 Philip Sellers: if you're listening to this, podcast. Do you not need a plucky puffin in your night life, I mean like, come on like. 60 00:08:53.440 --> 00:08:56.969 Jirah Cox: My life is, my life is plucky, puff, and deficient for sure. 61 00:08:58.350 --> 00:09:00.818 Jirah Cox: which is also a tongue twister. 62 00:09:01.230 --> 00:09:02.978 Philip Sellers: It is, they all have been 63 00:09:03.460 --> 00:09:04.386 Philip Sellers: But yeah, 64 00:09:04.990 --> 00:09:15.209 Philip Sellers: you know this. This is a real thing for the enterprise, you know, and obtaining support and and keeping things up to date. That's that's a huge thing. But 65 00:09:15.630 --> 00:09:21.029 Philip Sellers: this is also an embrace of a nciless 66 00:09:21.300 --> 00:09:27.759 Philip Sellers: way to leverage in Kp, because this Linux can run on bare metal. 67 00:09:28.300 --> 00:09:28.690 Jirah Cox: Yeah, sure. 68 00:09:28.690 --> 00:09:30.110 Philip Sellers: Support. Nkp. 69 00:09:30.410 --> 00:09:52.239 Jirah Cox: Very true. In fact, I mean with all respect to our friends that have moved to not even a new capability introduced here right like Nkp, since since it's since well, since before it joins the new 10 X family, right? Has been able to be part of your fleet management running containers anywhere, including on bare metal, for sure. Now, of course, the same. I would submit. The same value. Prop exists for 70 00:09:52.240 --> 00:10:05.018 Jirah Cox: using like an ubuntu pro on bare metal as it does exist, for in a Vm. It's, you know, security doesn't really care what the substrate is. It's an OS that needs some kind of a patch or mitigation. So you know, chop, chop, get to it, 71 00:10:05.670 --> 00:10:09.831 Jirah Cox: And I like the way that you felt called out like 72 00:10:10.540 --> 00:10:18.224 Jirah Cox: aligning more closely with enterprise patching right? Because so one of the earliest things that the article calls out here that 73 00:10:19.428 --> 00:10:43.729 Jirah Cox: one of the promises right of this platform from canonical is 10 plus years of security maintenance. So it actually kind of separates out what in most like consumer OS's are are to get are together intertwined mitigation of Cbes and security vulnerabilities and patching. And now with this you get those are intentionally broken out. I can mitigate findings from my security scanners. 74 00:10:44.044 --> 00:10:58.489 Jirah Cox: Very rapidly, but not actually, Rev. The version of the software that runs my stuff which might need a new requalification or go back to the lab, or whatever. Go back through engineering. I can securely patch, but not have to not have to upgrade. 75 00:10:58.960 --> 00:11:18.929 Philip Sellers: Well and and canonical, has a feature called live patch, also helping where doesn't actually require a reboot. So you're able to do kernel vulnerability patching so very low level in the operating system while keeping the system online. And so that's a huge differentiator for 76 00:11:19.429 --> 00:11:38.399 Philip Sellers: running Nkp on a ubuntu pro system. So you know that, and a host of other security features, you know, as with any pro or enterprise release, they're really talking about security here under the open source. 77 00:11:39.070 --> 00:11:41.540 Philip Sellers: Canonical umbrella. 78 00:11:42.010 --> 00:12:05.990 Jirah Cox: Yeah, and that matters so much to to practitioners, right? Because containers can't live migrate right? So to to drain to drain the host of the containers deposits running right. They have to get restarted somewhere else, which could be a failure in one of them, or if it was like a monolithic app in my container ecosystem. Would be minimally disruptive to my user base. If I can patch the host 79 00:12:05.990 --> 00:12:14.250 Jirah Cox: with the pod staying right where they are. That's of course, way way better. I can now Patch, honestly kind of like the rest of the Nci stack non-destructively 80 00:12:14.320 --> 00:12:16.168 Jirah Cox: during the business day. 81 00:12:17.217 --> 00:12:26.609 Jirah Cox: at the at the time of my choosing, and get those nights and weekends back, and not need to be on my laptop on a Friday night on the VPN. To to to push out a patch. 82 00:12:26.950 --> 00:12:42.449 Philip Sellers: Yeah. And I think that's that's the other part of this, too, is beyond the patching. You've got choice again in ways to deploy cost mitigation things like that, you know. If you're running nkp on bare metal 83 00:12:42.720 --> 00:12:59.009 Philip Sellers: ubuntu pro, you're not gonna have some of the same features. You get on Nci, that are storage specific and and things like that. But you know, for certain workloads that may be okay. If you've got ephemeral workloads that don't necessarily need data. 84 00:12:59.010 --> 00:13:15.090 Philip Sellers: But then you've got databases. This gives you a way to to unify Nkp across all of your different operating environments while also benefiting from the unique value that you get from an nci deployed Nkp stack. So 85 00:13:15.120 --> 00:13:25.890 Philip Sellers: again, getting a high level of choice in the way that you do things while being able to benefit from common standard operational practices. 86 00:13:27.210 --> 00:13:28.809 Jirah Cox: Totally. I mean, if I picture, like. 87 00:13:28.870 --> 00:13:57.490 Jirah Cox: you know, containers running underneath like a cash register that probably doesn't really need an nci stack right. But but there's still going to be a container host, right? And that's going to need some kind of patching. So now I can. You know my my security team or my container team can now patch whether it is in the data center in the cloud under the cash register, all with one mechanism right? And and one level of service to the business of now our OS patching doesn't need to take anything down or ever be noticed. 88 00:13:57.800 --> 00:14:02.009 Philip Sellers: But and I love that you went there with your example of cash registers, because 89 00:14:02.930 --> 00:14:20.610 Philip Sellers: I hate to use fast food. But it's an approachable thing we all interact with. And every one of the restaurants now has an app, which means they're running some level of infrastructure in every one of those stores, and everyone who's doing some level of door dashing and stuff like that. 90 00:14:21.010 --> 00:14:46.629 Philip Sellers: All of our restaurants, especially Post Covid. And and during Covid they became incredibly more connected, and a lot of them chose to create and architect these mobile applications and these user experiences on containers and on Kubernetes. That's something that we see over and over and over again with our enterprise customers. They're choosing to go this route. 91 00:14:47.327 --> 00:15:01.779 Philip Sellers: And it makes for that infinite scale. When you've got hundreds or thousands of different points, presence. It's just the way that the industry has shifted. Do you think we'll see that 92 00:15:02.300 --> 00:15:08.220 Philip Sellers: experience shift? Move down to smaller businesses over the next few years. 93 00:15:09.250 --> 00:15:16.880 Jirah Cox: Absolutely. I mean the easy one to always reach. For when we're, you know, doing a thinking deep thoughts about the feature, of course, is AI right like. 94 00:15:16.880 --> 00:15:17.300 Jirah Cox: sure. 95 00:15:17.300 --> 00:15:29.889 Jirah Cox: and even, and even with even even staying on fast food for one more second, like we've all seen the screenshots right of places, ex- experimenting with voice to text for ordering right? And so do I want that to run. 96 00:15:29.890 --> 00:15:30.700 Philip Sellers: Angles. 97 00:15:30.700 --> 00:15:31.350 Jirah Cox: There you go! 98 00:15:31.970 --> 00:15:52.179 Jirah Cox: Do I want that to run? Don't want that to run the cloud and pay by the pay, by the token, or don't want that to be effectively a container sitting on a Gpu or adjacent to a Gpu that can run through. You know the latest. Whatever whisper model it is that can do voice to text reliably and generate the order in my structured data format and push that over to cash register. 99 00:15:52.505 --> 00:16:09.769 Jirah Cox: And yeah, it's easy to imagine that kind of use case not needing a whole lot of virtualization, but needing an awful lot of governance around lifecycle of the the containers and the things that run them for sure. And and at that point retail, broadly right, I think, is is. 100 00:16:09.920 --> 00:16:20.930 Jirah Cox: I think, I mean this very respectively. It's like an AI playground, right like a AI applied to retail broadly is is I think, a a very much booming already, and only going to be growing industry. 101 00:16:21.370 --> 00:16:25.669 Philip Sellers: Yeah, I can only imagine what things I was seeing 102 00:16:25.820 --> 00:16:39.569 Philip Sellers: in the retail space around video, not just surveillance, but analysis of what a traffic flows through my stores look like what attracts more attention. You know, if I redesign this end cap 103 00:16:39.720 --> 00:16:49.000 Philip Sellers: in my store, which in cap gets more attention, what brings people to that, and makes them more 104 00:16:49.810 --> 00:16:54.389 Philip Sellers: intent to buy all those kinds of things being run. 105 00:16:54.850 --> 00:17:15.950 Jirah Cox: Or even like I mean, it's it's easy to imagine a B testing across stores right with data driven, you know. Let's take the let's take the subject of analysis out of it. Right? Here's the data that this in cap in that store makes people linger more than that one in that store, therefore, next week that one becomes our new standard, you know, on our on our layout models. 106 00:17:16.319 --> 00:17:24.369 Philip Sellers: Yeah, I mean you. You look at the Amazon app. You know they A B test a lot on this in live. You know that 107 00:17:24.999 --> 00:17:38.249 Philip Sellers: transaction where the checkout looks a little different. That's not a mistake. Y'all, that's them doing live a B testing with with their audience. And so you know that that's the kind of thing that 108 00:17:38.724 --> 00:17:57.099 Philip Sellers: it's happening more and more in the physical world. Now, just like you're talking about. And we didn't have access to to some of that. I thought it was cool that a video could be used to count the number of of people foot traffic, you know. But that was 10 years ago, and 109 00:17:57.209 --> 00:18:01.076 Philip Sellers: we've come so much further beyond that today. 110 00:18:01.779 --> 00:18:06.659 Philip Sellers: sentiment, analysis, you know, of voice and of text. You know. 111 00:18:06.660 --> 00:18:08.100 Jirah Cox: Real Time inventory. 112 00:18:08.100 --> 00:18:13.719 Philip Sellers: Yeah. Oh, the inventory stuff, I mean, you know all the little robotics and and applications 113 00:18:14.150 --> 00:18:19.730 Philip Sellers: inside of warehouses. You know, we we have a large retail customer. 114 00:18:20.170 --> 00:18:28.539 Philip Sellers: The systems in their warehouses cannot go down, that is death to them, and it is an exponentially expensive 115 00:18:28.840 --> 00:18:43.609 Philip Sellers: per minute cost, and so that uptime becomes paramount. For the infrastructure that's behind the scenes. But you know, to bring that closer and full circle. Now we're talking about some of that same stuff running in a 116 00:18:43.900 --> 00:18:46.350 Philip Sellers: a chicken shack down the road. 117 00:18:47.390 --> 00:18:50.004 Jirah Cox: Yeah, well, or or even 118 00:18:50.921 --> 00:19:19.029 Jirah Cox: you know, most of these retail environments also, you know, maintain like a store lab, right? Or a test, a test store nearby. Hq. For rapid prototyping, right and development. And so how do I ensure that the way I deploy to a a science project lab also is the way I roll out at scale in prod across the actual retail edge. So that kind of standardization leads to simplicity leads to uptime as well. 119 00:19:19.370 --> 00:19:25.359 Philip Sellers: Well, and and that drives the need for consistent, reliable. 120 00:19:26.180 --> 00:19:34.130 Philip Sellers: flexible infrastructure as a basis. Then you can iterate, and then you can try all this crazy stuff on top of it, because 121 00:19:34.890 --> 00:19:38.680 Philip Sellers: the difficult complex stuff has been made simpler. 122 00:19:39.130 --> 00:19:41.839 Philip Sellers: The actual okay, fun stuff. 123 00:19:42.070 --> 00:19:51.019 Jirah Cox: With with all the respect in the world for my my container colleagues. Now who are genuinely smart people, I often 124 00:19:51.330 --> 00:19:58.629 Jirah Cox: oversimplify containerization. To this is the fashionable way to install an application right now, is it? 125 00:19:59.190 --> 00:20:20.489 Jirah Cox: It actually does have a bit more complexity and scaffolding underneath of it. Sure! But the complexity is worth it when I want to install it 100 times a day. Right is then to your point when I can automate that with guaranteed outcomes. Now I get to go to the fun part right, and spend less time automating infrastructure and more time actually driving business value on the app layer on top of that. 126 00:20:20.490 --> 00:20:27.990 Philip Sellers: Absolutely. You know. I I moonlighted as a app dev many, many years ago, so 127 00:20:28.160 --> 00:20:36.312 Philip Sellers: I feel for them, but like I never did at dev to the level that that people do today professionally. 128 00:20:36.970 --> 00:20:49.069 Philip Sellers: But even then there's something rewarding about that cycle of iterating and improving and showing your work in real time. And so that drives 129 00:20:49.260 --> 00:20:52.970 Philip Sellers: the behavior you're talking about of deploying, you know. 130 00:20:53.250 --> 00:21:00.160 Philip Sellers: 100 250 times a day as you're iterating through and making incremental progress on the 131 00:21:00.290 --> 00:21:02.809 Philip Sellers: bigger problem that you're solving. 132 00:21:05.050 --> 00:21:30.320 Philip Sellers: It's fun stuff. We may have gotten a little off topic. But, you know, underlining kind of why we're here today. Nutanix and canonical. Their partnership is strong. Nkp on ubuntu pro opens and solves some real core enterprise problems in terms of security flexibility and longevity 133 00:21:30.865 --> 00:21:44.379 Philip Sellers: but also choice, and how you choose to go to market and meet your business needs. So Jaya, I appreciate you talking with me again today and just kinda 134 00:21:45.070 --> 00:21:50.690 Philip Sellers: going back and forth around this topic of kubernetes and our favorite distro. 135 00:21:51.700 --> 00:21:54.149 Jirah Cox: Thank you for wrapping it up. I need to go find some fried chicken. 136 00:21:54.530 --> 00:22:00.999 Philip Sellers: See planet alright. Everyone have a great day, and we will see you on the next episode.