 Hi my name is Don Wake. I'm a technical marketing engineer with Hewlett Packard Enterprise and today I want to show you how Hewlett Packard Enterprise with the Esmeral container platform is helping you to utilize the the power of the Kubernetes container orchestration software in ways that haven't been available before. So to get us started to give you an idea of how Hewlett Packard thinks about things like data science or or any kind of organization that has to utilize a lot of people to to do these data heavy types of applications. For example let's just take data science for example and and maybe machine learning application you know artificial intelligence. It's a team sport. You have data scientists that are building models and training models. You have data analysts that then are taking those those models maybe monitoring them or you know building applications using that data. You also have a group of software engineers that are doing the continuous integration and continuous development and they need access to tools like Jenkins and GitHub or BitBucket that type of a thing. And then you also have the operations team which maybe could be called a systems administrator or you know somebody that's required to make sure the systems stay online whether those systems are you know in a given data center or online maybe in a favorite public cloud. So when you have all of these people and you have all of these different pieces of hardware and software that you need to organize you need a central control plane. And Kubernetes is great at organizing the containerized software but what about all the other things like different role-based access like for data engineers and scientists and app developers and DevOps. All of these different individuals do not need the same access to the same tools. For example a data scientist probably doesn't want to worry about how to import a given cluster whether it was created on a cloud or on the on premises in your local data center. That person just wants to create a machine learning model. So with the container platform that we're showing here we allow you to use multiple different clusters of and even different versions of Kubernetes orchestrated CNCF certified containerized clusters. So you can have a multi-tenant environment so multiple people could be using the same hardware but not have the same access to for example the Kubernetes namespaces. The Esmeral data fabric is also part of the the Esmeral software solution and it offers a enterprise grade storage for doing things like replication a global namespace and mirrors and encryption at rest and you can have all of these things that you can have the separation of the compute and storage managing the data on premises or in your favorite public cloud. So there's a lot of things that having a central control plane wraps around Kubernetes orchestration. So let me just show you some quick ways that you can integrate your standard Kubernetes administration with the container platform and use it in any way you see fit. So what I'm showing here I'm on let's pretend I'm a systems engineer or perhaps a systems administrator and I'm connected to my Kubernetes platform and so I'll use a simple command called the kube control command or kube ctl and this is connected over the network because I've downloaded a plugin and I've authenticated myself and taken care of all of the security things so that I can you know communicate with my Kubernetes cluster wherever it is. So just get pods commands one of the simplest commands every Kubernetes expert wants to do you know at the atomic level that's how Kubernetes handles applications container app rise applications is at the pod level. So this is the first thing I'm just kind of looking around at my at my namespace and what I have access to. There's another command here to just give you an idea and highlight multi-tenancy when you set up your context this is the particular Kubernetes namespace that me as let's just say I'm the MLOp system administrator and I have access to this and I can do things to help data scientists utilize this cluster to do their job. So this just gives you an idea you know if you're familiar with Kubernetes these are the kind of things you want to be able to do as an administrator and perhaps you you do things that are even much more complicated like using rest API commands and writing Ansible scripts but let's say you now want to abstract things away a little bit and put on your system admin hat and log into our web UI. So this is another way to interact with the platform logging in with Active Directory credentials the Active Directory credentials that were set up by my organization before I even installed a container platform for example. What you see here is this very rich set of tools that you can use to import clusters create new clusters manage clusters and then also create the namespaces the Kubernetes you know tenants is what we like to call them because it's really multi-tenancy environment. I have a lot of tenants here and I logged in I happen to have site admin privileges you can see all of the different resources and you know you can quickly see oh I've got some clusters that might be pushing the envelope on resource utilization maybe I need to go in and add some more servers but what we're going to look at is what we we showed on that command line is the ML ops tenant so this is a tenant or Kubernetes namespace that utilizes the Esmeral software that puts together a whole bunch of things in context related to a full ML pipeline. So as a Kubernetes administrator I may want to come in here and just make sure my pod utilization is looking good okay this pod is using this much of the cores and CPU limits but you know to kind of to bring it back to um to what we were looking at there on the command line I looked at at what pods I had access to and if you go here to one of the contexts of machine learning is a Kubernetes I'm sorry a Jupyter notebook and I have a service endpoints here I can get all kinds of details here and if you remember we did our Git pods there's our pod name this is a Jupyter notebook so you get an idea that if you wish to be a systems administrator you can abstract a lot of the details about how you instantiate the pods or what servers are where so it's a really nice feature rich interface but also supports full REST API as I mentioned and with something like ML Ops you know you're really taking a lot of the abstraction I'm making it very abstract that you don't have to worry about which pod did you know what's the association with the pods in terms of my machine learning or my training engine here I just have access to endpoints so as a as a data scientist I don't have to worry about a lot of the details I can go straight to my notebook and in fact I could I could send this link to a data scientist and say hey in your cluster you asked me to spin up your Jupyter notebook and here I connected it to your GitHub so here's the URL they just click on that they now have access to a cluster a very powerful cluster and they can run their code in this Jupyter notebook on servers that are much more powerful and potentially have GPUs etc so that's just a quick tour to kind of show you some things you could do using the container platform to organize the Kubernetes orchestration there's lots of more things I'd love to talk about and I think if you found this interesting I'm going to point you to some of our resources online we have a demo portal where you can have actually get access to you know an environment that you can play with and of course you could go to our online documentations and there's there's always lots more to learn there's I didn't give a reference here to give you an example of how easy it is to you know get your kube CTL plugin so that if you wanted to use your laptop like I was doing I not have to be remoted in through a server or something you could do that so that's uh that's all I wanted to show you today and it's kind of like a lightning amount of information and lots more to learn but hopefully I waited your appetite and please feel free to contact me or anyone at hpe with the with that container platform and hpe.com and we'll get you some more information thank you for watching