is designed storage of container based workloads. You can now move the compute into the storage regardless of where it's located which opens up a world of infrastructure possibilities. Learn about iXSystem's unique use of
UI enables central management. Also, we touch on their very active
community.
Mike Matchett: [00:02:13] What I like about iXsystems basically goes back to ZFS. If you're familiar at all with file systems you know about ZFS. ZFS is a great file system and it is open source and there's a community edition. You can get your hands on ZFS, particularly from iXsystems, so they deliver the open source version itself. It used to be called Free NAS because it was, you know, free. They've done some acquisitions and some branding changes over time. They've relabeled all their product line now as TrueNAS, so you're going to find everything here under under. If you go the internets, look for TrueNAS. TrueNAS Core is the free ZFS software that that lots of people know and love. I think it's the number one ZFS community download million downloads and, you know, quarter million users out there of this software, so it's pretty bulletproof. And if you are just getting into file systems and setting up your own lab, setting up your own systems and want to play with the file system, this is a great one to start to use. So talk to Chris Moore was the VP of engineering over there for quite some time about what's going on. They, of course, move from there to NAS Core into enterprise systems, which is how they make a lot of their revenue there.
Mike Matchett: [00:03:33] So if you start with the free version and want a more supported or bulletproof a set of systems that can grow with your organization, you can get into there R series of systems and go from a single controller with no HA as an entry level all the way up to enterprise systems that are very scalable, lots more, lots more capacity to them, and they've recently announced a new product. New version is called TrueNAS Scale Hyperconverged version Dave, which is really scale out ZFS, and it takes advantage of ZFS 2.0 and brings us into the realm of containers and KVM and Kubernetes. So they're playing with that, and they've got their tuna's version available for Kubernetes, which is pretty cool, and it's going to support all the great things you love about ZFS. So infusion pools and hybrid storage, tiering and things like that. What I really liked, though, is when we start talking about what do you do with software defined storage and do you take software defined storage and run it anywhere you want? Or do you just buy it on a pre-built appliance? Is most storage companies sell software defined storage these days, but they really still sell it on their pre baked appliance.
Mike Matchett: [00:04:41] So it doesn't matter here. I think you have a real opportunity to take the TrueNAS products as file systems and deploy them wherever you want. From edge to center to clouds to anything else, you've got this common mesh of ZFS that you can send a set up and manage. What's what? I like even more, though, now that they've got the scale part of it is you can run this in your Kubernetes environments. You can run it as a container. You can also. Turn around and take the TrueNAS Scale solution as a hyperconverged solution and run containers with TrueNAS, so TrueNAS is running. And now you can decide, do I want to bring my storage into my compute environment? Do I bring my storage into my container environment? Or do I want to bring my compute into my containerized storage environment? And those decisions really free you up architecturally to to really optimize where you build and operate your applications and do all your DevOps kinds of minimax things that you want to do, which is pretty cool. So without going too much more into that, I recorded a short interview with with Chris. You want to roll the tape, Dave?
Dave Littman: [00:05:46] Okay. All right. Sounds good, Mike. Let's let's roll to the clip and then we'll come back and we'll talk about it and we'll lead into our next conversation with Elisity.
iXSystems: [00:05:58] Here's a staggering fact over 90 percent of the world's data was created in the last two years. This is no longer just a big data problem. Data explosion is a challenge now faced from the enterprise business to the home and everywhere in between. Historically, businesses were forced to solve data growth by being locked into a high priced big box vendor, and home users had no choice but to buy consumer grade storage that didn't truly protect their data. That is until TrueNAS changed the face of enterprise storage by democratizing it for all to use. This means that whether you're a Fortune 500 companies Storing petabytes of valuable data or simply looking for a place to store your valuable family photos, there is a TrueNAS Open storage solution for you. In fact, TrueNAS is the only storage that can be used from the enterprise to the home with the same software and file system to keep your data safe and secure. This flexibility is what makes TrueNAS the world's most popular data storage platform.
Kris Moore: [00:06:56] So just a little quick history lesson here, so most of our users and folks who know about friends and NAS know that we've always been, you know, the ZFS guys, open source guys and those products were always based on FreeBSD. True scale brings something new to the TrueNAS ecosystem to the family, something that's been asked for many, many years now, which is the ability to run things on a Linux base instead of FreeBSD and unlock some other features. So some of the biggest things that we're bringing to the party, so to speak with this Truth in IT scale version is, of course, scale out.
Truth in IT: [00:07:27] Some features of true mass scale include Linux based versus free BSD. Scale outgrowth, leveraging cluster to run containers directly on true nets, including Linux, DR and arbitrary container images. TrueNAS Scale also supports Kubernetes pods and Helm's charts.
Mike Matchett: [00:07:47] All right, so let's let's poke into that a little bit more. You've got two NoSQL got some cluster abilities there to scale out and really make a large capacity pool, right? Do that because people now think about when they do that, think about hyperconverged. They think about big data lakes where you've got all these nodes and you going to run your compute and your storage together in some some fashion to be more optimal because you want to bring that compute closer to storage when you have a lot of storage. Let's well, let's dove in a little bit more when you say you run Docker natively on internet scale. So I've got a containerized application. Do I do actually then host that on the same nodes that you're running on? Do I give it to your operating system? Do I take up individual containers and try to make some frankenstein-ing kind of thing running my half, my up here and half the app there? What what does that really look like?
Kris Moore: [00:08:41] No need to go frankenstein-ing unless you really want. Of course, a lot of ways to skin a cat, but essentially it lets you take any standard container image and then deploy it directly through the TrueNAS web interface or through the CLI or API, as well.
Truth in IT: [00:08:54] Deploy any standard container image directly through the TrueNAS web interface or through CLI or the API. Pick and choose which nodes to run containerized workloads on. The functionality is also fully integrated into the web based UI.
Kris Moore: [00:09:10] One of the coolest things we've done, I think on scale is we've really opened up the whole App Store ecosystem, if you will, if I'm allowed to say that, which is the idea that you can have multiple catalogs of apps connected in your TrueNAS. So it's possible using third party repositories to go and bring in other collections of cool applications. So if there's some series of software, it could be home use personal prosumer enterprise. There may be dedicated catalogs being hosted either by us on the site or even out in the community or businesses that want to host their own collection. Applications and scale allows customers and users to go and pull those directly into the UI and manage them all graphically and just takes all the headache and hassle out of doing this kind of container management directly on the storage, which is great because a lot of these things are all doing things on the storage side and need access to the data.
Mike Matchett: [00:10:00] All right. So so one thought that I'm having is that, you know, for a long time storage analysis, the data has gravity data sits somewhere and it's got to have a home. And people talked about a lot about containerization virtualization as these fungible concepts for compute moves around and goes into the clouds. And it's all this, all this stuff. But being stateless, right? And it's like, No, no, there's always going to be there's going to be data and just go somewhere. And so we're technologists. So of course, I mean, so where technology went for the last 10 years was to make the storage software defined so you could move the storage to where this compute was floating around in the ether. Right? And what you're telling me if I'm understanding this right is no. The data is going to be big petabytes of data. It's going to be somewhere and you want to keep your hands on it. So why don't we bring the compute to that and make it go the other way around?
Kris Moore: [00:10:48] Absolutely. That's more or less the model we've started here is just what the really rock solid storage foundation we've been providing for more than a decade on the true debt side and FreeNAS side. At this point, taking that and then marrying that up with the compute and saying, OK, you have all your data that's has got to be again somewhere. You don't get a petabyte just by snapping your fingers, you can pay for it.
Mike Matchett: [00:11:09] So, so I'm not going to I'm not going to name names here. But in the big broad brush, you've got companies like VMware who have spent a long time say, Hey, we've got this virtualization environment that moves things around something. Then we make VSAN to float around inside that store. And you guys just sort of like in a short amount of time here have just said, Look, we've got this really robust storage environment. If you're going to go to containers, bring your containers into the storage environment and you've got this, you've got, you know, even more benefits, right? So. Absolutely. It almost makes more sense to me to do what you're doing then than to think about it the other way. I don't know why I would take stuff with data heavyweight things that need replication and resiliency and move it around all the place where, you know, I want to bring the compute to where the data.
Kris Moore: [00:11:54] For us, storage integrity is number one always and being able to have that data integrity at a bit level. So you got that wonderful pedigree TrueNAS has had was the ZFS under the hood forever. And then now we're bringing saying, OK, now run your compute workloads closer to the storage, you know, manipulate that data any way you see fit. It seems like we're seeing more and more apps that do that nowadays. I mean, even apps that sit between, you know, your your business cameras, you know, ingesting data and they need to do some transformation or encode,
Mike Matchett: [00:12:22] You've got the transcoding and the exhilaration and all that stuff.
Kris Moore: [00:12:25] Yeah, yeah. Yeah. And in the old days, that meant spinning up another system somewhere to go and stand that up. And if you have a containerized version. Or even technically a VM, you could run on TrueNAS as well, which scale which has KVM integrated. You could go that route as well and still run it on the storage, and it just makes it a lot easier. It's one less thing to have to stand up.
Mike Matchett: [00:12:45] All right. So I think if people are listening to this and following along, they're like, Hey, wait, this sounds like I could build a lot of crazy stuff here. I could bring lots of computer. Are you? Are you worried at all about what people might do with this in terms of interfering with storage? We're bringing lots of compute into it. And you know, where does the compute actually go and how do I optimize all that? Or are you saying, hey, open source community have at it and give us, give us, give us some development here, use machine learning or practice?
Kris Moore: [00:13:11] I don't know if one is the right word, maybe excited. I'm really eager to see what people do with it because again freeness has always been a Swiss Army knife, if you will. It does a lot of things. A lot of flexibility were open source. We believe that all through the software, which means giving people power to use their system and software the way they want. So yes, you know, there's some risk involved there. People may put too much compute on their storage and, you know, hurt overall performance. So obviously, you know, people need to take some care and caution when spinning up these services. But you know, we just believe again in being open that people are allowed to go and set up things and run and experiment and tinker. And I think that's where the best innovation happens.
Mike Matchett: [00:13:49] If you really want to optimize the system for a given application, this is what you're going to do and you're going to watch it, you're going to monitor it and you're going to feel a part and you're going to create some control management for it. And I actually actually got pretty excited when I first saw this and heard about this just a little while ago because I can picture this world where people are writing all sorts of great kinds of orchestration tools and deployment tools for their apps that now say, like, Hey, we know where the data is, let's move the computer intelligently, where the data is going to be and, you know, make best use of those resources that we do have to maybe sometimes still buy for the data center. Sometimes we substitute big storage in our data center, so let's make use of it. So very cool. You got I just want to change gears here just a little bit. You've got this. Speaking of orchestration, this TrueCommand software. And tell us a little bit about how that helps orchestrate things and even the Kubernetes part.
Kris Moore: [00:14:45] Sure. So TrueCommand is our single pane of glass software. It's available either as a standalone image container or image again or cloud service TrueCommand cloud that we offer, and you can go to the website and spin that up right now. And what's really cool about TrueCommand is it has kind of that top down external view of all your storage because what we're seeing is, you know, used to be had one storage system and then two and then it just kind of grows, you know, everything just snowballs. It seems like we don't ever delete anything or throw anything away. We just keep adding more storage as we go. So TrueCommand takes that and tries to simplify the management and monitoring of it. And so you're able to go and spin up a TrueCommand instance. Add in all your TrueNAS units. It's able to monitor them, perform updates, do alerts. All that good stuff, even hardware inventories. If you need to run an inventory report of all the disk drives and the serial numbers you have in your network, right? But because TrueCommand sits externally, that's how we also create clusters. So since TrueCommand has access to all your different true noises, you're able to in a very poignant click way, go in and say, I'd like to create a cluster. Now, this is how much space I'd like to allocate from these three nodes, for example, and TrueCommand goes and talks to each. Does all the setup does all the networking in the background. It's talking all the APIs and making that magic happen, taking away kind of those manual processes. If if you've ever deployed any kind of clustered file system by hand, you know what I mean, where it's usually like log into the first node, set up all this and all the parameters, and then go to the second node, Oh, come on, SPF SaaS is a trivial,
Mike Matchett: [00:16:12] Trivial thing to deploy, I've heard.
Kris Moore: [00:16:14] Yeah, but anyway, TrueCommand just abstracts that all the way. It's just literally how much space do you want to allocate? Ok. These nodes have this kind of space click, click click done. And the next thing you know your cluster is up and running. So TrueCommand is taking that, and that's for creating the storage. And then the next logical step for a future release of scale and TrueCommand as well will be, of course, the Kubernetes administration as well. And being able to say, OK, now I have these nodes, they're clustered. Here's a compute job. I'd like to run it on this cluster and have access to this storage and go that route. So that's stuff we're slated to start working on a little later next year. But we're really excited because the current scale that's coming out of beta actually tomorrow and going to the first release candidate is laying the groundwork for all these really exciting things to come.
Mike Matchett: [00:16:59] Awesome. Awesome. Yeah. And you know, it's sort of an open source community. I can imagine that you're going to have a lot of energy being put into making that that level of orchestration smarter and smarter and smarter very quickly. Right. That's, you know, people are just going to look at that like, I've got ideas on how to how to how to align compute with storage. I'm going to be able to implement them now using using the APIs you provide.
Kris Moore: [00:17:22] One of the most beautiful things about open source I just got to put this little plug in is that the community is very vocal and is really good about saying, Oh, you've built some really cool building blocks here, if I could do this. This and this was them that would be even cooler, so they're constantly giving us ideas and saying, build me better tools because we're going to go and take it and run with it.
Mike Matchett: [00:17:39] So CTO for you guys. Customer technology office, right? Right. You got the gun. You got the customer. Ok, so this is all pretty cool. You mentioned TrueNAS scales moving, edging towards release candidates and becoming GA pretty soon here. Where can someone maybe find out some more information about this?
Kris Moore: [00:17:58] So TrueNAS.com Just remember that one website. Go there, look at the forums. That's a great way to interact with the community. Of course, you can download Truth in IT scaled the release candidate starting tomorrow, or, of course, TrueNAS Core as well. And then also, we have a very vibrant and active discord community now. That's something we just spun up in recent recent months and that that little thing has taken off. Apparently, a Discord is a lot more popular than I thought these days.
Mike Matchett: [00:18:22] But anyway, well, you know, all of us have been working at home, have nothing to do, but, you know, play games all day long. Discord Discord is a poppin', is what it says.
Kris Moore: [00:18:30] Go figure. But we got a huge community on there that's just constantly growing, and it's a great way to interact with other users, but also developer community as well.
Mike Matchett: [00:18:38] Awesome. Awesome. Well, thank you so much, Chris, for coming here and explaining to you now. Scale. I can only see some great and interesting things coming out of this, so keep in touch and those of you who are watching. If you haven't played with FreeNAS in the past and now it's the TrueNAS Core and you're interested in that, go get that and keep an eye on TrueNAS Scale because data is only getting bigger.
