Running Kubernetes in Production with NKP | Global Nutanix User Group Webinar

Welcome everyone to the global Nutanix User Group webinar running Kubernetes in production with NKP. Thank you for joining us from around the world. Today we're diving into why Kubernetes alone isn't enough for running containers in production and how the Nutanix Kubernetes Platform enables you to scale out Cloud native applications effectively anywhere. Plus we'll see live demos of NKP's day two operations and gitops in action. But before we start special call out to our friends in Ireland. The Ireland NUG is hosting a watch party today and it's great to see everybody virtually. Ronnie Hamilton the NUG leader great to see you and thanks for doing this. Our speakers today are Nimal and Kapil. Both are Technical Marketing Engineers specializing in NKP and Cloud at Nutanix and we also have Jose Gomez Director of Technical Marketing Engineering leading innovations in NKP and Cloud native. But before we start I want to pose a question in the chat and this is where you get to exercise your emoji skills. Who here has deployed a Kubernetes cluster before? Drop a rocket ship in the chat if you have and if you haven't drop a penguin for your learning and I'll give that a few seconds to go. Wow a lot of people have deployed a Kubernetes cluster. Oh looks like it's about 50/50 at the moment. Great thanks everyone for doing that. Get ready for valuable insights and practical takeaways and remember two lucky attendees will win a Nutanix Yeti Rambler so stay tuned to the end we're going to announce that on the Nutanix community Twitter handle @NutanixNation so be on the lookout for that. So with that let's get started Nimal, Kapil take it away. All right thank you Angelo thank you for the lovely introduction. I'm actually going through the comments and I see a bear so someone is doing some foul play there. All right so let me quickly share my screen all right I hope you guys can see the NKP screen. Again thank you all for joining us here today. We want to take the next one hour to talk about Nutanix Kubernetes Platform, what is NKP? This is a product that we announced roughly four months back. A very high level one line summary would be complete Fleet Management of all Kubernetes clusters running anywhere on prem data center on the public clouds at the edge with data services included in it. So a little bit about me my name is Nimal Kunnath. I work in the technical marketing engineering team. In my role I primarily focus on implementing Cloud native solutions for our customers and NKP is a core focus of my day-to-day work. I'm joined by my wonderful colleague Kapil. All right thanks everyone for taking the time out of your busy Tuesday. I'm Kapil Anandani. I'm a Tech Marketing Engineer for Nutanix Kubernetes Platform and Cloud native Technologies. I'm based out of Bangalore India. I'm excited to be here to show you guys what cool things we are doing at Nutanix in the Kubernetes space. Just some housekeeping stuff, if you have any questions you can ask directly in the chat and they'll be answered live. We will also have a Q&A in the end to answer any live questions or that are remaining. Over to you Nimal. All right thank you Kapil. So a quick look at what the plan is for today. So we want to kick this off talking about a few challenges that organizations and our customers run into when you think about running Kubernetes in production. After that we want to introduce NKP, what is NKP, why does NKP exist, what does NKP solve. Then we will dig a bit deeper into the components of NKP. The whole CNCF ecosystem, the projects that build the NKP product. And I'm very excited about the Fleet Management feature about NKP, so we'll be diving deep into that and to support that we'll have a couple of exciting demos as well. All right so to get started I wanted to talk about some big trends that we see in the Kubernetes community today. So Kubernetes has been around for about 10 years. This is the 10th year running there's a growing adoption of this technology we have a lot of analysts talking about it. There's so many conferences, events, meetups around to share information on how folks are running Kubernetes in production. By the way this slide is from an independent survey. So if you look at the first trend here most people, a vast majority, 92% of the people who are surveyed say that developers should not be running Kubernetes clusters but there's still a high number of them building their own clusters. So now if you think about it this evolved from the devops model. So earlier we had a centralized IT model where you had clear separation of roles between developers and IT. So IT had more control but the downside was that devs, they actually get to focus on building the application code the business logic part right, but they had really no flexibility and no self-service right and that's how the devops model emerged from it. So now you have your individual dev teams, your developers, your devops teams running their own Kubernetes clusters. Obviously this gave them more speed, more agility in getting their apps to market quickly but the flip side was that again they had to manage and operate their own infrastructure, their own clusters right, and this is what led to the evolution of a platform team, right a platform engineering if you're familiar with it, and I'm sure a few of you might be working in similar roles today. So a platform team combines the best of both worlds. Your developers get that freedom of self-service for the infrastructure needs, while IT still gets to maintain and retain that control. So the second trend that we see is a majority of customers from the survey say that they do not want to run both VMs and containers on separate infrastructure, but rather they want to manage it together. Now I'm sure most of you on this call are running VMs virtualized environments today and the benefits of virtualization were very clear as you move away from bare metal servers. Now as you start adopting containers and Kubernetes you want to be able to run that on the same infrastructure. Maintaining two separate silos of infrastructure can lead to a lack of flexibility. The third stat here is that we see that a lot of containers today or most containers require storage, persistent storage. Now if you think about it a few years back in the initial days of Kubernetes, it was meant primarily for stateless applications because the tech was not really mature enough but now we see more and more workloads, stateful workloads, being onboarded to Kubernetes. In fact you have entire data on Kubernetes at popular events like KubeCon. A couple of other interesting stats there, a lot of companies want to run Kubernetes on multiple clouds, multiple locations also, almost everyone looking at running AI would want to run it in Kubernetes. All the new apps that are being written today will be run on Kubernetes. So to give you a high level summary of the challenges that I spoke about right, developers managing Kubernetes infrastructure is a waste of time. It creates a lot of complexity so this is best done by other teams, such as the platform team. Running separate silos for VMs and containers is inefficient and expensive. Ideally you would want to have a single platform that can serve both. Also, storing data for containers often involves multiplatform and multivendor solutions, so if you think about it you're buying storage from a storage vendor, your virtualization from someone else, and your Kubernetes from a different vendor right, and this is not going to scale so you have to figure out how to stitch all of them together and this is really where Nutanix comes in. So Nutanix has been in the business of delighting customers with a single operation model no matter where you're running your apps and data. We have been in the market for about 15 years and we really bring that cloud-like operational model to your onprem in your Data Center and we started out with running VMs and networking and security as part of your Nutanix Cloud Infrastructure onprem. Then after that we gave you this ability to burst your workloads to the cloud so now you're extending to the cloud and you can run, manage, and operate your apps and data with a common operation model right onprem and on the public cloud, and on top of that we added file storage, object storage, database as a service, and now what we really want to do with NKP is we want to bring that ease of administration to Kubernetes itself and first of all we want to do this from a single pane of glass. Now one thing I really wanted to highlight here is that NKP is not a 1.0 product. The product that we launched in August is actually NKP 2.12. This is a result of an acquisition. So Nutanix acquired a company called D2iQ almost a year back and integrated their product into our stack. So now this company has been in the market for more than five years. They're an early pioneer in the cloud native space and they're major performers of key technologies like CSI, the container storage interface. And one more thing I wanted to call out is that now some of you, I know we have a few existing Nutanix customers in this call, and you may be wondering about NKE or Nutanix Kubernetes Engine, so this is a Kubernetes solution that we have been offering for the past few years and the plan is that NKP will be replacing NKE so if you're using NKE today we have published a migration guide. It is actually available in the portal since yesterday, so it would really benefit you if you could take a look and see how you would want to plan your migration from NKE to NKP. All right so now to talk more about NKP specifically let me hand this off to Kapil. All right thanks Nimal. If I can share my screen. All right let me stop sharing. Just give me a heads up if you see the presentation. Yes I see it. All right so Kubernetes is often seen as the platform for modern Cloud native applications. It is incredibly powerful but brings a lot of complexity with modern application architecture and what does a CXO see? Let's think from CXO perspective. That Kubernetes is some kind of a Magic Bullet that will solve all their problems for getting out new applications, features, out in the market faster than competitors. But let's think from the developers perspective also, what does a developer need from Kubernetes? Needs an automated scaling, which allows application to adjust resources based on demand without manual intervention. A developer needs consistent environment all the way from local testing to production thereby it reduces deployment issues, shoots up the reliability, and so on. Developer wants to more focus on writing the code knowing that something is being managed or the infrastructure is being managed efficiently in the back end and moreover it would be cherry on the cake that if it has integration with all the devops practices like ansible, terraform, that supports faster and more reliable software delivery through automated pipelines. But what does it take to build a successful Cloud native platform? Is it just about Kubernetes or is it just about deploying the microservices? Now think of Kubernetes as a powerful engine for a super car but is engine enough to drive that car? No you need nice wheels, you need a good body, you need a good suspension and same thing with Kubernetes. For Kubernetes a whole ecosystem must be tested, validated, and managed efficiently. Numerous teams have to come together to work hard to make every microservice application function seamlessly all the way from cloud administrators that are needed for designing and maintaining infrastructure supporting these application and ensuring they are scalable, resilient, cost effective, to security Engineers because microservice bring lot of footprint they have lot of moving parts each with its own vulnerabilities. So security needs to be tight and adaptable to keep up with the evolving threats while you're deploying it and all of these teams face ongoing challenges strategically selecting the best tools and infrastructure to build and operate platform that can meet the demands of developers. It's a delicate balance of performance, security, reliability all while trying to stay ahead of technological changes that's been happening in the modern world and only few companies have done that so far. When I say few companies these are multi-billion dollar companies like Google, Amazon, who have no resource shortage and just spend billions on R&D. Now to summarize you need a platform that works and scales effectively as the organization grows, ensuring that the developers have the resources to build and deploy applications without unnecessary delays or risks. Now when we talk about Kubernetes it's important to remember that it's just engine right it's just a small piece of puzzle. But when we want to run applications at scale in production, a full-fledged Cloud native platform is needed. It comprises of many moving parts, each with its own specific role and these moving parts handle everything from networking and storage, to security and orchestration and importantly every one of these component needs to be thoroughly tested, validated, maintained and upgraded to ensure that the platform is reliable and scalable. This difficulty is caused by large number of open-source projects that are there in CNCF landscape, especially those paid products. It's a vast and ever growing ecosystem of tools, platform technologies, I think more than 1,500 now, and that can be overwhelming to manage and integrate and this can become a significant barrier to success even for the most capable teams. All this difficulty you know it can turn into what might seem like hey I started this open source CNCF project by saving the cost of my company, but it can turn into a risky and resource intensive endeavor. What does NKP bring on the table? Now Nutanix Kubernetes Platform is an Enterprise ready opinionated stack of upstream Kubernetes and when I say upstream it's pure upstream Kubernetes, no forks, no proprietary and the best of CNCF projects ready for production. It is everything that a platform engineer needs to build a secure, reliable, scalable platform to run containerized application and its data at scale with true portability and no lock-in. Now this allows customer to take advantage of the state-of-the-art innovation in application development, production technologies, all the way up and including gitops and we'll see in a demo today and Nimal will show you how gitops work in NKP platform and trust that everything will work seamlessly. Now with over 22 best of the breed open source projects integrated into NKP you get a comprehensive suite of functionalities and capabilities that are right here. Now I'm not going to go each one of these just to be conscious on time, but you get this functionality designed to meet the need of platform engineers in Cloud native landscape and to give you a clear picture again, all these functionalities, let's break them down into three fundamental use cases that is aligned to what I'll show you in a few minutes that what customers most need. It's cluster provisioning. NKP uses an open source project called cluster API that handles the life cycle management of Kubernetes cluster and next is the day two operation like how does the observability stack works like if there is a microservice or application performance issue how can I troubleshoot? Where should I go? Is it just digging the logs how can I get those nice graphical representations of where the issue is happening and third is Fleet Management, like how can you onboard new clusters into application which has a lot of demand let's say during the Christmas year or it is not using much resources during the rest of the year, so how do you automate those kind of things and by focusing on these three core use cases Nutanix ensures that you have the tools and support needed to handle the complexities of cloud native environment with confidence and ease. Now let me show you a quick demo of cluster provisioning and observability stack. So I'm going to switch my screen here and if someone can confirm they see a Kubernetes dashboard. Yes. Okay thanks Nimal, so let's just assume you are a platform engineer who already has an NKP management cluster and wants to deploy a new application, or a new cluster for its hybrid Cloud application. So as soon as I come into the management dashboard in NKP I see this cluster, or this management, has this NKP cluster has four clusters and it has workspaces and all these applications Grafana they come out of the box. Now workspaces you think of workspaces like a way of creating group of cluster for a specific purpose and I want to deploy my application in this hybrid Cloud workspace, so if I go to hybrid Cloud workspace, now I see that here the Clusters are zero and we're going to deploy just one right now. So we are going to actions and we see we have an option to get an "add cluster" so I can create a cluster either using easy filling a form just on the web page itself, or if I have a declarative yaml manifest that I can upload and cluster API will take care of deployment of cluster and even if you have an existing Kubernetes cluster like EKS, AKS, you can just attach it right here and manage from the NKP dashboard. So I'm going to click on create cluster as part of this demo and you can get a production ready Kubernetes cluster ready within 20 minutes with NKP. Now I have prefilled this form just for you so I'm going to click and some of the fields are pre-filled let's say the cluster name I'm going to call it NKP Dev and this is the infrastructure provider where your cluster will be deployed. Now you can have AWS, vSphere, I mean you have to pre-add that infrastructure provider and just where your deployment will basically happen. Now this is the Kubernetes version I'm going to skip SSH username and key. Now configuring node pools, every Kubernetes cluster has at least two node pools one for the Kubernetes control plane and one for the worker node where you actually run your application. Now to start with control plane node pool you see that we have an integration with Prism Central projects now think of it like you can enforce virtualization quotas in Prism and role based access control like how much resources this cluster is actually going to use and you can control those amount of resources that it will use and here it will show the cluster so if you have more than one cluster it will show up here. In case of subnets you can either Deploy on VLAN subnets or a VPC, inside a VPC and when it comes to OS images you have two options. You can use the image provided by Nutanix which is based on Rocky Linux or you can bring your own OS. In both cases the image should already exist in the backend provider in this case Prism Central so I'm using Rocky Linux at the moment that is the OS provided by Nutanix and available for all the NKP customers when they run it on Nutanix AHV, the hypervisor that Nutanix has. Then we can use categories also like category use case think of it like you can enforce the micro segmentation security policies with Nutanix Flow and automate some stuff with NCM Playbooks, the Nutanix Cloud Manager Playbooks and coming back to the IP, the Kubernetes cluster here needs two additional IP addresses which will be required, one for the control plane virtual IP, and another one for the ingress service which will be used by open source project metalLB for load balancing purpose. For control plane IP the VIP has to be in the same subnet as control plane nodes and make sure it's not already used and something called as control plane node count so you can deploy highly available control plane nodes from 3 to 5 where the downtime is not affordable and you can also deploy one in let's say testing purpose. So I have enough resources so I'm going to go with three at the moment and same thing with the worker node pool projects, cluster, subnet, OS image and categories and worker and Auto scaling. So Auto scaling think of it like in cases of you if you have an application with fluctuating demands the cluster Auto scaling can be really useful for cost optimization, handling unexpected traffic, and you can enable it and deploy the worker node count. So I'm going to leave it right here as it is and when it comes to storage with every NKP cluster on Nutanix you get persistent storage out of the box so NKP has an integration with container storage interface CSI driver for Nutanix and you get the storage out of box, and this is where your second IP will come the load balancing IP if you have more than one you can just put start and end range so in this case I just have one IP so I'm going to put it here, and here it comes the image registry where you pull the NKP images from so you can either use stalker to pull the images or you can have a private registry so in this case I have private registry so I'm going to call it registry.local. I'm going to put in username for the private registry and password for the private registry and click on create. Now I just click create on this cluster and this cluster will be ready within few minutes, but by the time it gets ready we can check view details in the cluster and we can see some of the applications that are right out of the box enabled for this cluster so we have Grafana for dashboards, we can have Grafana Logging, we have Kubecost for cost optimization, Prometheus for monitoring and Traefik for traffic purposes, load balancing purposes, and then I can just check about the cluster details while it gets deployed so it is deployed on Nutanix this is my backend cluster, it's deployed in hybrid Cloud workspace, and if I click okay this is active and this is how you deploy a Kubernetes cluster, a managed Kubernetes cluster for your workloads in NKP. Now coming back to the second use case, now we are going to look at the day two use cases like the observability stack in NKP. Now think that you are basically a platform engineer and you are running an application in production okay and you got reported some problem, like there's a performance issue with an application called online boutique, and where the payments are getting failed and you just got started with Nutanix NKP and you don't know how to basically troubleshoot or there must be some skill gaps. So in this demo we are going to see metrics we are going to see how to investigate some of the logs and we are also going to see one cool thing which is like AI assistant chatbot that's built into NKP so with Nutanix Kubernetes Platform we have comprehensive stack of observability, logging, governance, AI assistant, and ensuring that you have all that you need for the day two operations. Now this NKP cluster has four clusters and if I go to my application workspace here, this is my Edge application workspace, I see out of four the two clusters are assigned to this application and there are some application that are already enabled. Now if I go to applications I see that this has Grafana Logging, Loki, and all those applications that come out of the box for you that are enabled and getting you production ready. Now if I go to Cluster wide I see that these are the two clusters, one thing to note here is we have the mix of clusters also like one with the Nutanix and one with vSphere in a single workspace so it is possible to basically have the different flavors of infrastructure provider within a single workspace. Now the performance issue was reported in the Durham site but I'm going to check view details and then I'm going to check this is the Kubernetes version and these are all the application enabled but to check let's go to Grafana dashboard here so what's happening here is the Grafana dashboard itself provides you about compute networking and storage metrics right out of the box as soon as you open the default dashboard for Grafana, and it also comes with list of all these dashboards that you can just check and see where the issue is happening. So in this case I'm going to click on my Kubernetes compute workload where my application is and I'm going to change the name space to my online boutique application and there's a micro service called payment service and I can troubleshoot the log here. Similarly if the graph is not helpful enough so we can go back to the dashboard and we can see the Grafana log, or sorry the actual logs, if you want let's say Loki and we can check the actual logs for that particular service. Now one cool thing here is basically we also help the organization in a way that to get you started that we have something called an AI chatbot assistance. Now this AI Navigator has been trained with our product documentation and support base so you can ask question let's say how do I ensure my etcd is running healthy in a cluster so you can ask any Kubernetes specific question or any question specific to any cluster it will give you all the answers that are needed to perform some health checks, run some troubleshooting, and so on so that's about me, I'm in both the demos and Nimal the floor is yours to take the audience through Fleet Management. Right thank you Kapil. That is a great demo on cluster creation and what kind of platform applications get deployed along with the cluster. So I'm going to share my screen back and I guess you should see the demo slide right? Yes. All right thank you Jose. All right so now let's talk about application disaster recovery and NDK or Nutanix Data Services for Kubernetes and how it supports your stateful applications. Now NDK is packaged with NKP. It comes built in to protect, so idea is to protect your stateful applications and it offers storage replication to ensure that your data is safe in the event of a disaster. So now if you think about it currently there's a disconnect between Kubernetes and storage systems. Your traditional DR solutions fall short right as they designed for protecting VMs, protecting machines. A DR solution for Kubernetes should be container granular because now you can have multiple pods from a single app which could run on multiple VMs or multiple machines and each pod can have multiple persistent volumes for each of them so in the end what we really need is application aware data services. Now imagine you have the entire application with all its related objects like secrets and config maps all the objects in a single name space that you choose, like your definition of an application in terms of Kubernetes manifest and the associated data the persistent volumes the PVCs right all of it packaged as one unit, snapshot it and the ability to recover it on your DR site and on top of that you can set protection policies as well right so for example you can have an hourly snapshot policy to let's say your DR location A and a daily snapshot and replication policy to your location B. So this is the true power of NDK, this is what NDK offers right and all of this can be done by your developers right in native Kubernetes language using Kubernetes yaml manifests so your developers still have that flexibility of self-service to self-service their DR needs without IT intervention and at the same time your IT teams are creating these guard rails within which your devs operate right so your IT can still set quotas limits right on how these resources are being consumed from an infrastructure side. So and now for application backup NKP provides Velero which is an open source project that has been designed to simplify backup and migration for Kubernetes environments. Velero can back up both your Kubernetes resources and your persistent volumes to S3 storage right and restore it to a DR cluster right so many of you might know that Nutanix storage portfolio includes object storage which is S3 compatible, so Nutanix Objects serves as an ideal back end to store these backups right and again all of this is available from the same Prism Central console. All right so now earlier a few minutes back Kapil showed how to get started with the Kubernetes cluster right and deploy all your platform applications on it quickly to get to day two production readiness right, but once you have all of that sorted right maybe you're thinking about hybrid Cloud about multicloud right so these kind of multicloud strategies introduce a lot of additional complexity because the primary reason here is that each of the public clouds have its own different Kubernetes distribution, a separate tech stack that comes with each of the cloud vendors which means that all of them behave a little differently right. So the challenge here really is in providing a single unified experience across all these different environments right wherever your clusters are being deployed, so quite often what happens is that once you start adopting the tech stack from a particular vendor or particular Cloud, you're kind of locked into it. Now certainly your application is not so portable anymore even though Kubernetes and containers have this promise of being able to run anywhere right, so you would want to keep this in mind as you get deeper into a particular cloud or a vendor's ecosystem and tools and this is where NKP helps right. This is why NKP is common set of platform applications all based on graduated CNCF projects and widely adopted by the community as well right these can really help as they help you with application portability and what I'm really excited to talk about today is the Fleet Management capabilities that are built in to NKP. Now you can think about a fleet as essentially a group of clusters it's a grouping of Kubernetes clusters. Now with workspaces here you can allocate Kubernetes clusters to different departments. Now in this example you would see there's a finance workspace, there's an HR workspace which has got a few clusters in it, there's an NKP cluster, there's an EKS cluster from AWS, there's an AKS cluster from Azure right, so Fleet Management allows you to centrally manage and operate this group of clusters so now we have features like security, visibility, automation and governance, can all be consistently managed and applied at a fleet level right at a group of cluster level. Now this massively simplifies operations and ensures consistency at scale, so now if you think about it now your application updates are being rolled out at the same time to let's say the 10 clusters in your Finance workspace or let's say you make an update to Grafana and you want to make an update in general and this can be rolled out to your NKP, your EKS, and AKS cluster right in the HR workspace, so it means that managing Kubernetes clusters and the applications within it across different environments can be done in a consistent manner right so whether these clusters are running on Prem, in the cloud, or at the edge. And I want to show this in a quick demo so I'm going to stop sharing and try to switch to my demo. All right so here we are back in the NKP console here, so we looking at this from a platform engineer's eyes right from his perspective, let's imagine this platform engineer is responsible for an edge application that he wants to roll out using gitops methodology. So in this workspace that we have called my Edge applications we see there are two clusters there's a Nutanix cluster in Durham and there's an Oakland cluster that is running on vSphere so now let me click on projects on the left hand side right so there's a single project here called online boutique now you can think of project as a high level construct within NKP. The project is used to manage and group resources, policies, and users across multiple Kubernetes clusters. So when you create a project it creates a namespace in this example there's a namespace called online boutique that has been created and Federated and pushed out to all the Clusters that are associated with it. So in this example this namespace is created in these two clusters Durham and Oakland. Let me open the project and I see there is a continuous deployment page right so this is the gitops integration that earlier Kapil also mentioned means that when your development team rolls out a new version of your app to your git repo which is a source of truth for your application, the code will be automatically rolled out to all your clusters right, and let's see what else we have here so I'm clicking on Role bindings here. Role based access control, RBAC, is built in to NKP so there's predefined roles and permissions that come out of the box. You can also create your own custom roles which aligns with your organizational needs and further with NKP you can integrate your NKP platform with your existing identity providers in your environment right through LDAP or SAML. I'm going to click on Quotas here so again with quotas and limits you can specify the resources that you want to allocate both from a compute and from a storage perspective, from namespace perspective, for like how much resources do you want to allocate for your developers within this name space. I'm going to go back to clusters and again I confirm that we have two clusters here so I want to see this app now right so I'm clicking on the Durham cluster and I see all the platform apps listed here let me click on the Kubernetes dashboard that comes built in and it opens up the default namespace but I know that my application is running in the online boutique namespace so I'm going to drill down to that namespace and sure enough I see a bunch of pods from all the microservices that run within that app and let me click on services and I know my application is being exposed as a load balancer service. So let me click on that IP address and sure enough this is the boutique application that's running here. Great, this looks great. This is running on both the Durham and Oakland clusters. Now what I want to do is maybe I want to onboard a new cluster or maybe decommission one of the existing clusters right so it's very easy right I can just quickly go here and click on edit project and if you look at it right now there's a selection here is to manually select clusters, which means we're explicitly setting that these two clusters should contain this application right but maybe I want to switch it to dynamically select clusters, which means that the Clusters will be added dynamically depending on a label that I add to the project, now a label is kind of like a key value pair similar to categories in Prism Central so I have added a specific label here of infra ID and the value is AHV MKT 236 and you can instantly see that the vSphere cluster got deselected right because the label is only assigned to the Nutanix cluster and the application once I hit save this application is going to get removed from the vSphere cluster right so let me hit save here and once I hit save and come back to the project page I can see that there's, in the Clusters page, the vSphere cluster has been removed, right the application only exists in my Durham cluster. Now what I could do is I could come back to clusters and quickly create a new cluster let me add a new cluster in this case this cluster is going to be in EMEA in Amsterdam and all I have to do is ensure that as I create a cluster ensure that that label is tacked to this cluster as well, and while the cluster is provisioning I can quickly come back to projects and see that hey now this EMEA Amsterdam cluster is being added to the project right and I can drill down to the project and see that the vSphere cluster got removed and instead I have the new EMEA cluster running on Nutanix, now this is really powerful right, so which means that as long as the cluster has got that tag that label the application gets rolled out to that cluster, it doesn't matter if it's one cluster right because I could add one cluster, 10 clusters, 50 clusters whether it's an NKP cluster running onprem or an EKS cluster on AWS the application gets rolled out and that is the true power of Fleet Management, which means that if you're a company if you have multiple development teams working on different microservices, each team can be assigned a project within NKP right so this ensures that each team gets access to its own set of name spaces, resources and policies, now you could have a workspace admin or a tenant admin that retains control over all the Clusters in that workspace but then you also have a global admin who can retain control over your cluster wide governance. All right so I'm going to switch back to my deck and yes, that was about Fleet Management right. All right, so we have seen a little bit of projects and workspaces and we saw how their key components and building a true MSP grade multi-tenant stack. Now if there are service providers in this you would find this particularly useful right because there are a lot of challenges that MSPs would face when you build a platform to support a tenant right how do you secure user access with the right side of permissions, how do you manage the infrastructure, is it going to be shared or dedicated to specific tenants how do you ensure that one tenant is not consuming or eating up a lot of resources right, so this is where NKP has been built from the ground keeping MSPs in mind. You have hard multi-tenancy by allocating dedicated Kubernetes clusters to specific tenants or you have the concept of soft tenancy as well, where a cluster gets shared across projects. Now user access become seamless by integrating into existing identity providers too. So in this example you can think of tenant one workspace. All right, so we are almost towards the end of this webinar so as we wrap up today I wanted to leave you with the key reasons why Nutanix NKP is an excellent choice for your platform engineers to run containerized applications right, now first we have is freedom of choice now NKP integrates seamlessly with Nutanix and it's fully supported across multiple clouds, multiple public clouds, multiple hypervisors no matter where your infrastructure is running you have the flexibility to run your applications where it makes the most sense for your business. Next is one unified platform, so with Nutanix you have a consistent platform that spans all your environments whether applications are traditional or containerized whether they're running onprem on the cloud with Nutanix you get a seamless consistent experience. Power of open, because NKP is built on pure upstream Kubernetes there are no vendor lock-ins, no proprietary forks, which means that you get all the benefits of the open source community of all the Innovation that is happening in the community, with the security and the support that Nutanix provides. Now most of you might be familiar that Nutanix has been consistently delivering an exceptionally high NPS of 90 plus over the last many years. And finally enterprise class data services. We combine world class leading storage and data services with the Best in Class Kubernetes platform right, so this really ensures that your deployments are reliable and scalable. So I know we covered a lot of ground. We have an NKP Test Drive that is available from that URL, Nutanix.com/one-platform where you can take, you can see most of these demos that we showed and you can try it out for yourself. It's free and it's not just NKP, but you can try out all the products that we have in our portfolio today. Yeah so that's what I have today. Great thank you! Jose I wasn't sure if you wanted to take some questions live or you just answering them? Yeah we can take a few of them on live Angelo and the first one will go for Nimal and this is related to a question around when they hear about Kubernetes a lot of things are designed as microservices and on their existing application many of them they are monolithic running on bare metal or VMs, does NKP help to migrate those applications from monolithic on bare metal or VMs into microservices into Kubernetes? Yeah so Kubernetes is a container orchestrator right so which means your application has to run in containers, now usually microservices are running containers but it doesn't have to be microservices but generally that is a way, so in this case it would not be applicable because NKP and Kubernetes needs containerized applications and the second part of that is does Nutanix help in migrating from monolithic to microservices, no so currently we do not help our customers in the migration process, we help you deploying, we deploy a production ready Kubernetes but only after your applications are already containerized. Thank you, Nimal. Could you explain Kapil, about if it's mandatory NDK and Velero, when you deploy NKP or they could be optional? They can be optional, I mean either you can have open source Velero or you can use Nutanix data services, but again it's inhouse or built for NKP so data service I would say is a preferable option here. Thank you, Kapil. Nimal, could you explain a little bit around licensing for NKP, how does it work when it comes to virtualize environment, bare metal, as well as full stack. So with NKP there are a few different licensing models, so the first licensing model is tied to NCI, so if you are an NCI Pro or an NCI ultimate customer, you would get unlimited NKP starter clusters. Now NKP starter is already a huge upgrade from NKE because you're getting a cluster API based infrastructure underneath and on top of that you can add NKP Pro or NKP ultimate licenses for the true Fleet Management capabilities and the AI chat bot and all that. Now but this is again only if you're an NCI customer, now there's a different licensing model where you have the NKP full stack so this you can think about it as the entire stack right from your infrastructure NCI NKP you get some storage through NUS, you get NDK and you actually get some databases through NDB and we have different starter Pro editions as well for it and all of this information is there in our licensing page on the Nutanix website. Thank you, Nimal. Another one for you, when it comes to multi-tenancy specifically on role based access control, can we specify the role of admin to a user for a specific project? For a specific project? So in other words, could we have a project admin in a specific project? Yes you could have a project admin who is typically it could be a developer but it doesn't have to be, but this user would have admin rights to that project which is a shared resource across multiple clusters which mean they only have RBAC access to that specific space, and this is the role that comes built in with NKP. Thank you, Nimal. Another one for you because it's related to the demo that you did, you deployed an application that was landing in two clusters within the same work space, or in this case in the project, could you use a global load balancer over the clusters? Yes, yes you could, yes. Thank you, Nimal. If there are any questions please don't hesitate to send them by email. This will get, I believe Angelo, posted on the Nutanix User Group website I guess that comment will be enabled, so they could leave there the question right cool, over to you Angelo. Thank you, all right thank you everyone for joining us. Don't forget to check out the @NutanixNation Twitter handle to see if you've won the Yeti Rambler and also check out the Nutanix Community and post your questions about NKP there at next.nutanix.com and with that thank you everyone have a great rest of your day. Thank you.