Listen to Kelly Olivier, Principal Architect at Nutanix, discuss how to intelligently manage and share unstructured data to help your business make informed decisions.
Speaker 1: On behalf of Nutanix, and Carahsoft, we would like to welcome you to today's podcast focused around consolidating and controlling your unstructured data where Kelly Olivier, Principal architect at Nutanix, we'll discuss how to intelligently manage and share unstructured data to help your business make informed decisions.
Kelly Olivier: A little bit about some of the storage offerings that Nutanix offers. And a little bit about where they came from, where the where the requirements came from, and how we address some of those things a little bit about me, Principal architect at Nutanix, I've been here a little over eight years, that's been a lot of fun for me, because I've been able to see the product from fruition until now, really, as I was a customer of Nutanix before, so we'll go ahead and get started here. Unstructured data seems to be growing at just record paces. IDC estimates that pretty much 80% of all data is unstructured, the other 20% being structured in in SQL databases, Oracle databases, things like that. So when we think about data in general, especially at home, and in the enterprise, a lot of the data is similar audio, video documents, application data, things like that all kinds of unstructured data. And IDC is report that came out in 2018 suggests that by 2025, there's just these record massive numbers of data, 404, 45 zettabytes. And I think there's a, an exabyte in between terabyte exabytes data valued. So we're talking about monster data here, the old adage of big data doesn't seem very big anymore. And like I said, 80% of all that data is unstructured. I thought this was an interesting quote from this worldwide file, an object based storage forecast done by IDC, that says like the physical uniform universe, the digital universe is large, by 2020, containing nearly as many digital bits as there are stars in the universe is doubling in size every two years. That is just a staggering number to think about, I wish that the stock market would do the same thing double every couple of years, that would be wonderful. So with all this data, you know, you have a couple of problems that arise. And the more data there is, the more we have to govern that data. And with that comes more risk. Where do we put that data is that data going to get out. And of course, everyone inside the government knows about compliance. As this data grows larger and larger, it introduces some more problems. With this, we've had, you know what we could call the rise of the cloud 91% of respondents in a Nutanix survey done by IDC Nutanix sponsored survey so that 91% of respondents say that hybrid cloud is the optimal deployment model. And really, that's because public cloud, wherever win wants to go, still has some shortcomings. And those of us in the in the work in the federal government space, you know, definitely can contract some of those reasons of why we want to keep data on premises, as well as cost 18% deploying today, even with realizing that hybrid cloud is the optimal deployment model. I think the reason for that is some of the complexity that's due to the networking that's in between, there's some up and coming companies that are specializing in this space, trying to make this really easy. But networking seems to be an area of enterprises. That was hard enough on premises. So now when you're talking about connecting all these public clouds with on premises data, you know, it introduces a lot of a lot of risk as well, by 2025 49% of data will be on the public cloud. We'll see. I don't know if I believe that personally. But it's definitely an interesting stat. So reasons for public cloud, why does everybody want to go to the public cloud? I think some of the more obvious ones are the quick time to market you know, you can jump right into a console and deploy a new application, deploy a new virtual machine, deploy a containerized app, you get the simplicity of not having to manage that infrastructure that's underneath. You get to pay as you go. You get that that flexible it consumption, that a lot of enterprises are finding really attractive, not having to make us huge, huge investment upfront but able to scale out over several years, and with the continuous innovation, that that is available for things like public cloud, you know, you think of a company like Netflix, they don't have release cycles, you don't download Netflix 10.0 on your phone, and then a week later, download Netflix 10.1. They're continuously innovating with code changes really, really around the clock. So that has to be, you know, some pretty robust infrastructure. But there's obviously lots of reasons for private cloud as well. And this is really where the market has been. And really, the economics of this, you know, in some, in some data that that Nutanix has done is really that 70% of workloads are more economical on premises, those are usually those static workloads, the ones that don't change the ones that run 24/7, 365 around the clock, and provide things like authentication or a database or something like that. It's really those bursty type workloads that really make more sense, economic sense, in the public cloud. And again, like I mentioned earlier, everyone in the federal government, you know, one huge risk, or two public cloud, and one thing that we always, you know, put on the side of private cloud is that control, you have constant regulatory, regulatory concerns. Now, not a lot of the public clouds are certified for all types of data. So there's, you know, again, more reasons to keep that data on premises. There's also, you know, maybe some of the data is written in applications that weren't written for public cloud. So you now have a, you know, a lift and shift type of scenario where you have to rewrite an application to be able to make it what they say cloud native. So some reasons for both sides. But the obvious, you know, gravitational pole, when you go over those pros and cons of each, is that really hybrid cloud is kind of the only real option, the winning the winning model going forward. And, you know, we kind of get to have your cake and eat it too, as they say. And so even though this can produce, you know, some new challenges, like I mentioned before, like complexity, in between those two, this makes a lot of sense, you know, giving you the economic choice of where you want to run your workloads, that's really important to customers. And it's really important to Nutanix this is the model that we believe in 100% going forward. So with all of these data storage challenges, you know, we talked about the incredible amount of growth that's out there, the government, the governance and compliance control with the rise of the cloud, it makes it difficult if you're, if you choose to do this with legacy infrastructure. And the reason for that is you're, you know, kind of, like we mentioned about the applications that weren't written for cloud, you have data that was just never intended to be moved data that was never intended to grow at this kind of scale, the management of these storage systems don't scale very well and have limitations, you know, in the management side, as well. So as well as hardware that kind of just ages out, you know, if you think about a legacy storage system, as time goes by, you're not able to get that managed, easily, not able to automate, you know, that, that hardware with all the intelligence and hardware, you know, just gets old very quick. So like, you know, like I mentioned, you have those forklift upgrades that are not built for the cloud. But luckily, when you use something like Nutanix, we can solve these areas with new capabilities. Nutanix being software that was designed, or a hybrid world really to be your cloud on premises, as well as in the public cloud, through hardware in those public cloud, hyper scalars. So it's designed for simplicity, as you scale, you can start small with what you need today and be able to scale out have the option of where you want to move data, whether it's in this public cloud, or that public cloud, give you a lot of analytics and deep visibility into that, as well as all your workloads. This conversation we're having is, is geared around files, data and objects, data and things like that. But these are all things that are really just features on top of the Nutanix platform, you know, this enterprise cloud platform that's built for our public cloud as well built for a multi-cloud world. So let's talk a little bit about files and objects. So like I mentioned Nutanix, it was started company was started in 2009. We started shipping product in 2011. And the very first feature the first killer app of Nutanix was storage itself. We kind of removed the need for a Sam architecture. And we built this from the ground up and said, Hey, what are the big companies out there? You know, at the time it was companies like Google and Yahoo, you know, all these huge companies, but were not finding the scalability that they needed in legacy architecture in that three tiered architecture. So we kind of took a step back, and we said, hey, if we're building this from the ground up, what would it look like? And oh, by the way, we didn't have to reinvent the wheel. One of our founders came from Google was the chief architect of the Google file system. So they said, hey, you know, we're already doing a lot of these things. But it's not available for the public market. It's available for things like Gmail and YouTube and things like that. So this was kind of the beginnings of Nutanix. And really, that first killer app, like I mentioned, was just storage itself, taking all the complexity out of all the different aspects of a storage system, and really making them really making them invisible, as this foundation is a bulletproof foundation that has all the resiliency and business intelligence built into it for hosting virtual machines. So as you can see in the middle of this slide, it's kind of a rudimentary view of what Nutanix is, it's not the hardware underneath. It's not the hypervisor, even though we have an included hypervisor that you can use called Acropolis hypervisor, or HV. But this just becomes, you know, yet again, another value of our platform that starts with as little as three nodes, and can scale out to as big as you need. And even though, like I mentioned, the first, the reason why Nutanix came along was to make storage easier. After we had that bulletproof foundation available, it made things like file storage, and object storage deployed on top of this. It made those just really easy features and cool features. You know, a lot of companies out there, all they do is object storage, or all they do is file storage. But this, this allowed us to use our bulletproof foundation, as a hosting platform to just enable these things as features. So we're going to talk a little bit about that in the coming minutes. But this architecture allows you to start small with what you need today, just like we talked about in that public cloud portion of giving you the bite size consumption, this is a smallest three nodes, which is you know, really to you in a rack. And this could be as little as you know, 20 terabytes. Yeah, even smaller than that with some of our other models. But really just breaks it down into the inner resources of compute, networking and storage. So let's talk a little bit about Nutanix files. So Nutanix files is delivered on top of our platform, like I mentioned, in the form of some application servers, and it uses our native storage underneath to then export that to your VMs or two physical workstations are things you know, outside of the virtual environment, it allows you to export those as cifs shares, or NFS shares, it can join your Microsoft DFS, the cool thing about it is it's highly available. You know, some customers are like, oh, well, we already do this today, we just deploy some Windows servers on top and use DFS. And it's like, okay, great, you can do that. But this is built in it deploys in minutes, you know, you get one single namespace across your entire cluster there and have that built in resiliency, so that if one of these application servers goes down, those connections are redirected to another file server, if you wanted to look at it that way. So when we came out with that, we came out with sip storage. And then later on, we allowed customers to do NFS, as well. And then even after that, we allowed multiprotocol, which essentially is just both its cifs and NFS, to the same parts of data, if you're familiar with Nutanix, some of the built in things like disaster recovery, and encryption at rest. And you know, these things that are at the cluster level, those hold true for these things that we're running on top here. So you don't have to worry about your data not being encrypted, don't have to worry about how am I going to back it up? Or I should say, how am I going to fail it over to another, another site? So files is a really exciting product, it's probably our best-selling product that is, you know, alongside of ALS, you know, our operating system that runs with Nutanix nodes. So inside of this is one thing I wanted to focus on later on something called files analytics. But that's pretty exciting, and provides a lot of value for folks. I'll talk now about objects very quick. Nutanix objects came around a little bit later. And just like files was, you know, the sits storage on top of Nutanix. This is object storage on top of Nutanix and it's written through the s3 API. So it's Amazon you know, compatible or any other tool Rather that you want to replicate to, you could take data that was written on Nutanix objects and replicate those to AWS or, or vice versa. So there's no kind of in compatibility there, it uses the same API. So it's all s3 compatible, you can make these deploy these very simple deployments. One thing that a lot of customers like is that these can go across clusters. So where we might have a Nutanix cluster, let's say in the data center, that's four nodes that are another one, that's four nodes. And let's say that on one of those, we have five terabytes left. And then on the other cluster, we have five terabytes left, well, we can actually make a single namespace data object store out of both of those together. So it's a way for customers to take what data is left over in their system, and be able to use it as well, as you know, obviously, we can take a whole Nutanix cluster and dedicate that for Nutanix objects as well. But yeah, there's a lot of a lot of cool aspects of this, it can do things like things like host static websites, it has support for worm, which is right once read many. And that's something that in this day and age, you know, with all of the security vulnerabilities that have made it into the news lately, worm is one of those things that really fights that it can really fight, you know, that malware war, just because as objects are written, they're immutable, they can't be changed. So if you get some kind of malware on your system, that embeds itself in an object and maybe make some changes, and then you find out about it six months later, when they're asking for Bitcoin, you know, this is definitely something that fights that battle for all of the objects, you know, you get that choice, when you build the object store, of whether or not you want to enable that. And once you write those objects, you know, they're written, they're done for so they can't be changed. All kinds of lifecycle policies built into objects, you know, as far as, Oh, if I want this aged data to now replicate somewhere else, or if I want this age data to be deleted, we can, we can handle things like that for you. And like I mentioned, data at rest encryption, built into the platform, you know, borrowing that from our existing data at rest implementation, underneath, at the platform level, https support, you know, there is a nice graphical, you know, way of interacting with buckets. So it's, you know, I remember when s3 came out, originally, and it was kind of like a command line utility. And then later on, they brought out the ability to kind of interact with it through a browser. And, and we've done that as well, you can also, I think I mentioned this host static websites in it, it's another common use case, where people don't have to deploy a web server for a quick web page or form that they're doing, you're able to do that as well. One of the biggest use cases for objects is in the backup space. And it seems like a lot of backup vendors, you know, we'll talk about some specific ones that we have integration with veem, Haiku, net backup, you know, there are a lot of backup vendors that are using object stores as their backup target. So you install the software somewhere, and you point it to this object store. And that's where you're making your backups to. And again, I think that's part of Part of the reason for that is because of things like worm, you know, when you can write your backups to a backup repository, and know that those aren't going to be changed. You know, obviously, you get all those same benefits fighting that that malware war, you can scale to billions of objects. I mean, we've tested this up into very, very large implementations. And you can get as dense as a petabyte of objects, usable in just 10 nodes and 10 nodes for Nutanix, remember, could be a two u two u node, you know, two new two u box. So we're really talking about one rack less than a rack actually, for a petabyte of usable objects. And one thing I didn't mention about objects is that you get two terabytes with this with every Nutanix cluster. So you can you can take it for a spin, you can deploy it, play with it, use it, see how it works for you. And you can do that totally free of charge. And by the way, I didn't mention that about files, you get one terabyte of file storage, you know, on our Nutanix cluster, where you can deploy, you know, same type of thing where you can deploy it. It's as simple as creating a name for your server and then your shares and given it some IP addresses and it will do all the rest, you know, 20 minutes later. So you have that, that working file server and storage where you can jump into a browser whack, server, whack, share and open that guy up and start putting some files there. And again, it could be stuff completely off of Nutanix. That's using that. And the same with objects. So I alluded to this a little bit earlier, as far as ways that backup vendors are integrating with objects. And as you can see, on this slide, here, we have something called mime, where you can purchase veem, or Haiku, you know, software that we have done integration with, you know, everything, everything works, it's all validated, and you're actually running a Nutanix cluster with that software. And you can buy that all under one skews. So you can talk to, you know, your reseller that works with Carahsoft, and is able to get you a quote for that. Or you can also, you know, if you want to use a different underlying storage, maybe don't want to have a dedicated objects cluster, you know, you could use these even in things like, like AWS, for example, you can actually deploy a Nutanix cluster up in AWS, deploy objects on it and use that as your backup target. It really just depends on what your goal is. But these software vendors to the right Veritas comm ball arcserve, Haiku can back up directly to objects. And like I said, you get a lot of a lot of value with using the worm underneath, being able to write data directly to those know that it's safe, know that it's secure. And you can, you can move on. Again, even though you can use this for backup, you can still use Nutanix underneath, you know, for disaster recovery, like I mentioned, of course, managing objects, and managing files is as simple as just going into prism. And for those of you that aren't Nutanix customers today, maybe not familiar, this is the prism UI, it's the same place that you're managing that that underlying platform, when I was talking about the bulletproof infrastructure underneath, this is what it looks like, you can kind of see at the top left here that this one, the hypervisor is HV as six terabytes free of you know, this is a pretty smaller cluster, 90 VMs running on it, it shows you some performance data. But this is the dashboard, this is the top level view into Nutanix. And so if you're already using Nutanix, for virtual machines, or block storage, or you know, or just exploiting the storage off to physical servers, any of that you're using the same platform as you are for files and objects. So again, the cool thing about these, if you're already a Nutanix customer, it's a no brainer, because files and objects just become features, you know, you kind of just check the box, and you can deploy them very quickly. There's not a lot of setup, not a lot of planning, it's just you know, which IP addresses Do you want, and how much storage Do you need. So that's a huge benefit of being able to manage inside of one UI, all of your virtual machines, all of your file storage, all of your object storage, all of your block storage, you know, in the form of volumes, or I scuzzy lands to export it, export and servers, all of that from one UI. So I alluded to this a little bit earlier about data. And file analytics is one of those cool features inside of the platform that really gives it you know, a deep look and allows you as the administrator to jump in here from a top down view, and say, okay, these are the biggest shares, these are the ones being used the most. But the cool thing is, it allows you to really figure some things out that maybe would, it would take you a lot of a lot of time otherwise anomaly alerts here up at the top, right, that's a huge benefit. You know, from a data loss prevention standpoint. You know, if you had the next node and God forbid, in your environment or something, you would notice that, hey, off hours files are being copied, this is kind of unusual, you know, the system would alert you of that. So, there's a lot of, you know, a lot of intelligence built in there. Also full audit ability again, oh, a lot of strange things are being copied, who's doing it, okay, I'm able to see exactly what they did. Now, let me go figure this out, that type of deal. So you can also export this data into reports, you know, that that you could give to your upper management if you wanted to. Also, you know, like I mentioned, just lots of lots of views, lots of quick information to define those anomalies. So one thing I want to do encourage you guys to do is check out a test drive of Nutanix so if you go to Nutanix comm forward slash test drive, there is different ones that you can do have all kinds of parts of the products, not just objects and files, but there are test drives for a lot of different aspects. Another one is run HV comm and it will have a bunch of videos of things that you can do very quickly to help you out on your journey. So I wanted to end with this. Great, thanks, everybody.
Speaker 1: Thanks for listening. If you'd like more information on how Carahsoft or Nutanix can assist your organization, please visit www.carahsoft.com or email us at nutanix@carahsoft.com. Thanks again for listening and have a great day.