 Welcome to this CUBE Conversation. I'm John Furrier, host of theCUBE here in Palo Alto, California. We've got two great remote guests to talk about some big news hitting with Scalely and Yulia Packard Enterprise, Jerome LeCat, CEO of Scalely and Chris Tinker, distinguished technologist from HPE, Yulia Packard Enterprise, Jerome Chris. Great to see you. Both CUBE alumni is from an original gangster days, as we say back when we started almost 11 years ago. Great to see you both. Wow, great to be back. So let's see you. So really compelling news around kind of this next generation storage, cloud native solution. It's really kind of an impact on the next gen, I call next gen DevOps meets application, modern application world. And some we've been covering heavily. There's some big news here around Scalely and HPE offering a pretty amazing product you guys introduce. Essentially the next gen piece of it, Artesca we'll get into in a second, but this is a game-changing announcement you guys announced. This is an evolution continuing, I think it's more of a revolution, but I think storage is kind of abstractionally of evolution to this app-centric world. So talk about this environment we're in and we'll get to the announcement which is an object store for modern workloads, but this whole shift is happening, Jerome. This is a game changer to storage how customers are going to be deploying workloads. Yeah, Scalely really, I personally really started working on Scalely more than 10 years ago, close to 15 now. And if we think about it, I mean the cloud has really revolutionized IT. And within the cloud, we really see layers and layers of technology. I mean, it all started around 2006 with Amazon and Google and Facebook finding ways to do, initially what was consumer IT at very large scale, very low credible reliability. And then slowly it creeped into the enterprise. And at the very beginning, I would say that everyone was kind of wizards trying things and really coupling technologies together. And to some degree, we were some of the first wizard doing this, but we're now close to 15 years later and there's a lot of knowledge and a lot of experience, a lot of tools. And this is really a new generation, I'll call it cloud native, you can call it next gen, whatever. But there is now enough experience in the world, both at the development level and at the infrastructure level to deliver truly distributed automated systems that run on industry standard servers, obviously good quality server deliver better service than service. But there is now enough knowledge for this to truly go at scale and call this cloud or call this cloud native, really the core concept here is to deliver scalable IT at very low cost, very high level of reliability, all based on software. And we've been participated in this solution, but we feel that now the breadth of what's coming is at the new level. And it was time for us to think, develop and launch a new product that's specifically adapted to that. And Chris, I will let you comment on this because the customers or some of them, you can add the customer to that. Well, you're right, I've been in the, I've been like you, I've been in this industry for a, well, a little long time, a little longer too, 21 years as an HPE in engineering. And look at the actual landscape has changed with how we're doing scale out software defined storage for particular workloads. And where a catalyst has evolved here is in analytics. Normally what was only done in the three letter acronyms and massively scale out POSIX namespace file systems, parallel file systems. The application space has encroached into the enterprise world where the enterprise world needed a way to actually take a look at how to simplify the operations. How do I actually be able to bring about an application that can run in the public cloud or on premise or hybrid, be able to actually look at a workload off my step that aligns the actual cost to the actual analytics that I'm gonna be doing, the workload that I'm gonna be doing and be able to bridge those gaps and be able to spin this up and simplify operations. And if you are familiar with the parallel file system, which by the way, we actually have on our truck, I do engineer those, but they are, they have their own unique challenges. But in the world of enterprise, where customers are looking to simplify operations then take advantage of new application analytic workloads, whether it be Spard, Mesa, whatever it might be, right? I mean, if I wanna spit up a MongoDB or maybe in a last search capability, how do I actually take those technologies and brace a modern scale out storage stack that without breaking the bank, but also provide a simple operations? And that's why we look for object storage capabilities because it brings us this massive parallelization. Thank you guys. Well, before we get into the product, I wanna just touch on one thing, Jerome, you mentioned, and Chris, you brought up the DevOps piece next gen, next level, whatever term you use, it is cloud-native. Cloud-native has proven that DevOps, infrastructure as code is not only legit, it's being operationalized in all enterprises. And security in there, you have DevSecOps, this is the reality. And hybrid cloud in particular has been pretty much the consensus is that standard. So, or de facto standard, whatever you wanna call it, that's happening, multi-clouder on the horizon. So these new workloads have these new architectural changes, cloud, on-premises, and edge. This is the number one story and the number one challenge all enterprises are now working on. How do I build the architecture for the cloud, on-premises, and edge? This is forcing the DevOps team to flex and build new apps. Can you guys talk about that particular trend and is that relevant here? Yeah, I now talk about really storage anywhere and cloud anywhere. And really the key concept is edge to go to cloud. I mean, we all understand now that the edge will host a lot of data. And the edge is many different things. I mean, it's obviously a smartphone, whatever that is, but it's also factories, it's also production, it's also moving machinery, trains, planes, satellites. That's all the edge cars, obviously. And a lot of data will be both produced and processed there. But from the edge, you will want to be able to send the data for analysis, for backup, for logging to a core. And that core could be regional, maybe not one core for the whole planet, there may be one corporate region, a state in the US. And then from there, you will also want to push some of the data to probably cloud. One of the things that we see more and more is that the DR data center, the disaster recovery, is not another physical data center, it's actually the cloud. And that's a very efficient infrastructure, very cost efficient especially. So really it's changing the paradigm on how you think about storage because you really need to integrate these three layers in a consistent approach, especially around the topic of security because you want the data to be secure all along the way. And data is not just data, it's data and who can access the data, who can modify the data. What are the conditions that allow modification or automatic erasure of the data? In some cases, it's super important that the data be automatically erased after 10 years. And all this needs to be transported from edge to core to cloud. So that's one of the aspects. Another aspect that resonates for me with what you said is a word you didn't say, but is actually crucial to this whole revolution, it's Kubernetes. I mean, Kubernetes is now a mature technology and it's just the next level of automatized operation for distributed system, which we didn't have five or 10 years ago. And that is so powerful that it's going to allow application developers to develop much faster system that can be distributed again, edge to core to cloud because it's going to be an underlying technology that spans the three layers. Chris, your thoughts hybrid cloud. I've been having conscious with the HPE folks for, got years and years on hybrid clouds now here. Right. Well, you know, and it's an exciting layout, right? So if you look at like whether it be enterprise virtualization, that is the scale out to the purpose of virtualization workload, whether it be analytic workloads, whether we know data protection is a paramount to all of this orchestration is paramount. If you look at that DevSecOps, absolutely. I mean, securing the actual data, the digital last set is absolutely paramount. And if you look at how we do this, look at the investments we're making, we're making, I know if you look at the collaborative platform development, which goes to our partnership with scale, it is we're providing them an integral aspect of everything we do, whether we're bringing Esmeral, which is our software BU's orchestration. Look at the veneer of its control plane, controlling Kubernetes, being able to actually control the actual community clusters in the actual backing store for all the analytics that we just talked about. Whether it be a web skill app that is traditionally using a positive thing space has now been modernized to take advantage of newer technologies running in NVMe burst buffers or 100 gig networks with flingshot network at 200 and 400 gigabit. Looking at how do we actually get the actual analytics, the workload to the CPU and have it attached to the data at rest? Where's the data? How do we land the data? And how do we actually align essentially locality, locality of the actual asset to the compute? This is where we can leverage, whether it be Azure or Google or name your favorite hyper scale, leverage those technologies leveraging the actual persistent store. And this is where scale with this object store capability has been an industry trendsetter setting the actual landscape of how to provide an object store on premise and hybrid cloud running in the public cloud to be able to facilitate data mobility and tie it back to an application. And this is where a lot of things have changed in the world of analytics because the applications, the newer technologies that are coming on the market have taken advantage of this particular protocol as three so they can do web scale massively parallel and concurrent workload. You know what, let's get into the announcement. I love cool and relevant products. And I think this hits the mark. Scalely, you guys have Artesca, which is just announced. And I think it, you know, we, obviously we reported on it. You guys have a lightweight true enterprise grade object store software for Kubernetes. This is the announcement. Jerome, tell us about it. What's the big deal? Cool and relevant. Come on, this is cool, right? Tell us. I'm super excited. I'm not sure that you can see it as we're on screen, but I'm super, super excited. You know, we introduced the ring 11 years ago and this is our biggest announcements for the past 11 years. So yes, do pay attention. And you know, after looking at all these trends and understanding where we see the future going, we decided that it was time to embark on the product. So there's not one line of code. That's the same as our previous generation product. They will both coexist. They both have a space in the market and Artesca was specifically designed for this cloud native era. And what we see is that people want something that's lightweight, especially because it had to go to the edge. They still want the enterprise grade that Scalely is known for. And it has to be modern. What we really mean by modern is we see object storage now being the primary storage for many applications, more and more applications. And so we had to be able to deliver the performance that primary storage expects. This idea of Scalely serving primary storage is actually not completely new. When we launched Scalely 10 years ago, the first application that we were supporting was consumer email for which we were and we are still today, the primary storage. So we know what it is to be the primary store. We know what's the level of reliability you need to hit. We know what latency means and latency is different from throughput. And you really need to optimize both. And I think that still today we're the only object storage company that protects that after both replication and error coding because we understand that replication is faster, but the error coding is better and more current and larger file. We are fast in terms of latency. It doesn't matter so much. So we've been bringing all that experience but really rethinking a product to that new generation that really is here now. And so we're truly excited against people a little bit more about the product. It's a software, Scalely is a software company and that's why we love to partner with HPE who's producing amazing servers for the record and history. The very first deployment of Scalely in 2010 was on the HPE servers. So this is a long love story here. And so to come back to Artexca, it's lightweight in the sense that it's easy to use. We can start small. We can start from just one server or one VM instance. I mean, it starts really small, but it can grow infinitely. The fact that we start small, we didn't limit the technology because of that. So you can start from one to many. And it's cloud native in the sense that it's completely Kubernetes compatible. It's Kubernetes orchestrated. It will deploy on many Kubernetes distributions. We're talking obviously with Esmeral. We're also talking with Tanzu and with the others in terms of Kubernetes distribution. It will also be able to be run in the cloud. Now I'm not sure that there will be many true production deployment of Artexca in the cloud because you already have really good object storage by the cloud providers. But when you are developing something and you want to test that, just doing it in the cloud is very practical. So you'll be able to deploy Artexca on a Kubernetes cloud distribution. And it's modern object storage in the sense that it's application-centric. A lot of our work is actually validating that our storage is fit for a single-purpose application and making sure that we understand the requirement of these applications that we can guide our customers on how to deploy. And it's really designed to be the primary storage for these new workloads. The big part of the news is your relationship with Yulia Packard Enterprise. It's some exclusivity here. As part of this announcement, you mentioned the relationship goes back many, many years. We've covered your relationship in the past. Chris, also, we cover HP like a blanket. This is big news for HPE as well. It is big news. What is the relationship? Talk about this exclusivity. Could you share about the partnership and the exclusivity piece? Well, the partnership expands into the Pan-HPE portfolio. We made a math investment in Edge IoT devices. So we actually have... How do we align the cost to the demand? For our customers to come to us, wanting to look at... Think about what we're doing with Marines like consumption-based modeling. They want to be able to consume the asset without having to do a capital outlay out of the gate. Number two, look at how do you deploy technology really depends on the scale, right? So in a lot of your web scale out technologies, putting them on a diet is challenging, meaning how skinny can you get it? Getting it down into the 50 terabyte range. And then the complexities of those technologies as you take a day one implementation and scale it out over multiple iterations over quarters, the growth becomes a challenge. So working with scalability, we believe we've actually cracked this nut. We figured out how to A, number one, how to start small, but not limit a customer's ability to scale it out incrementally or grotesquely. Grotesquely, depending on the quarters, the months, whatever the workload is, how do you actually align and be able to consume it? So now whether it be on our Edge line products, our DL products, go right there. Now what Jerome was talking about earlier, I mean, we ship a server every few seconds. That won't be a problem. But then of course into our density optimized compute with the Apollo products. And this where our two companies have worked in an exclusivity, where the scaly software bonds on the HPE ecosystem. And then we can of course provide you, our customers the ability to consume that through our Greenlink financial models or through a CapEx purchase. Awesome. So Jerome and Chris, who's the customer here? Obviously there's an exclusive period. Talk about the target customer and how do customers get the product? How do we get the software? And how does this exclusivity with HPE fit into it? Yeah. So there's really three types of customers. And we've worked a lot with a company called Use Design to optimize the user interface for each of the three types of customers. So we really thought about each customer role and providing with each of them the best product. So the first type of customer are application owners who are deploying an application that requires an object storage in the backend. They typically want a simple object store for one application. They want it to be simple and work. I mean, honestly, they want no frill. They just want an object store that works and they want to be able to start as small as they start with their application. Often it's the first deployment, maybe a small deployment. Applications like backup, like VM or rubric or analytics, like stock or vertical file system that now available as a software, like CTR, there's a really great departmental NAS that works very well, that needs an object store in the backend of high performance computing, WCAF file system, and there's an amazing file system. We also have vertical application, like Roadpeak, for example, who provides Origin and VOD software for broadcasters. So all these applications, they request an object store in the backend and you just need a simple high performance working well object store and Artisca is perfect for that. The second type of people that we think will be interested by Artisca are essentially developers who are currently developing some Kubernetes or cloud native application, your next gen. And as part of the development stack, it's getting better and better when you're developing a cloud native application to really target an object storage rather than NFS as you're persistently. Just think about generations of technologies and NFS and file system were great 25 years ago. I mean, it's an amazing technology. Now when you want to develop a distributed scalable application, object store is a better fit because it's the same generation. And so same thing, I mean, they're developing something, they need an object store that they can develop on so they want it very lightweight, but they also want a product that their enterprise or their customers will be able to rely on for years and years on. And this guy's really great fit for that. The third type of customer are more architects, I would say IT architects that are designing distributed system where they're going to have 50 factories, a thousand planes, a million cars, they're going to have some local storage which they want to replicate to the core and possibly also to the cloud. And as they design these really new generation workloads that are incredibly distributed, but with local storage, I'll discuss a really great fit for that. And talk about the HP exclusive, Chris, what's the, how does that fit in? Do they buy through Scalely? Can they get it for the HP? Are you guys working together on how customers can procure? Both ways. Yeah, both ways. They can procure it through Scalely. They can procure it through HPE and it's the software stack running on our density optimized compute platforms which you would choose to align those and to provide an enterprise quality because if it comes back to it in all of these use cases, it's how do we align up to a true enterprise stack? Bringing about multi-tenancy, bringing about the fact that if you look at like a local racer coding, one of the things that they're bringing to it so that we can get down into the DL325. So with the exclusivity, you actually get choice and that choice comes into our entire portfolio whether it be the edge line platform, the DL325 AMD processing stack or the Intel DL380 or whether it be the Apollo's or like I said, there's so many ample choices there that facilitate this. And it just allows us to align those two strategies. Awesome. And I think the Kubernetes piece is really relevant because I've been interviewing folks, practitioners and Kubernetes is very much maturing fast. It's definitely the centerpiece of the cloud native both below the line, if you will, below under the hood for the infrastructure. And then for apps, they want to program on top of it. That's critical. I mean, Jeremy, this is like, this is the future. Yeah. And if you don't mind, I'd like to come back for a minute on the exclusive with HP. So we did a six month exclusive. And the very reason we could do this is because HP has such a breadth of server portfolio. And so we can go from, you know, really simple, very cheap, you know, HDD on DL380, I mean, it's a machine that retails for a few $40. I mean, it's really like simple system 50 terabyte. We can have the DL325 that Chris mentioned, there is really a powerhouse or NVMe flash, all the storage is NVMe, very fast processors or, you know, dense large large system like the Apollo 4500. So it's a very large breadth of portfolio. We support the whole portfolio and we work together on this. So I want to say that, you know, one of the reason I want to send kudos to HP for the breadth of the server line, really. As mentioned, Autesca can be ordered from either company, hand in hand together. So anyway, you'll see both of us and our field is working incredibly well together. Well, just on that point, I think just for clarification, was this co-designed by-scaled in HPE because Chris, you mentioned, you know, the configuration of your systems. Can you guys, Chris, quickly talk about the design of the co-design? From a code base, the software is entirely designed and developed by-scaled from a testing and performance server. This really was a joint work with HP providing both hardware and manpower so that we could accelerate the testing phase. You know, Chris, HPE has just been doing such a great job. I've really focused on this. I know I've been covering it for years before it was factorable. The idea of apps working no matter where it lives, public cloud data center or edge, and you mentioned edge lines, been around for a while. You know, apps centric, developer friendly, cloud first has been an HPE kind of guiding first principle for many, many years. Well, it has, and you know, as our CEO and tenant here has stated by 2022, everything will be able to be consumed as a service in our portfolio. And then this stack allows us the simplicity and the consumability of the technology and the granulation of it, allows us to simplify the installation, simplify the actual deployment, bring it into a cloud ecosystem, but more importantly for the end customer, they simply get an enterprise quality product running on a density optimized stack that they can consume through a orchestrated simplistic interface. Customers, that's what they're wanting for today is for they come to me and ask, hey, I've got this new app, new project, and you know, it goes back to who's actually coming. It's no longer the IT people who are actually coming to us. It's the lines of business. It's that entire dimension of business owners coming to us going, this is my challenge and how can you HPE help us? And we rely on our breadth of technology but also our breadth of partners to come together and of course, scale these as hand in hand in our collaborative business unit, our collaborative storage product engineering group that actually brought this to market. So we're very excited about this solution. Chris, thanks for that input and great insight. Jerome, congratulations on a great partnership with HPE, obviously great joint customer base. Congratulations on the product release here, a big moving the ball down the field as they say, new functionality, clouds, cloud native, object store, phenomenal. So wrap up the interview. Tell us your vision for scalability and the future of storage. Yeah, I start, I mean, scalability is going to be an amazing leader for me. But yeah, so you know, I have three themes that I think will govern how storage is going. Obviously, Marc-Andris had said that software is everywhere and software is eating the world. So definitely that's going to be true in the data center and storage in particular. But the three trends that are more specific are first of all, I think that security, performance and agility is now basic expectation. It's not like an additional feature, it's just the basic table. Security, performance and agility. The second thing is, and we've talked about it during this conversation is edge to core. You need to think your platform with edge core and cloud. You know, you don't want to have separate systems, separate design, interface point for edge and then think about core and then think about cloud and then think about the diverse cloud. All this needs to be integrated in a design. And the third thing that I see as a major trend for the next 10 years is data sovereignty. More and more, you need to think about where is the data residing? What are the legal challenges? What is the level of protection against who are you protected? What is your independence strategy? How do you keep as a company being independent from the people you need to be independent? And I mean, I say companies, but this is also true for public services. So these for me are the three big trends. And I do believe that software to find distributed architecture are necessary for these trends, but you also need to think about being truly enterprise-grade and that has been one of our focus with the design of Afezca. How do we combine a lightweight product with all of the security requirements and data sovereignty requirements that we expect to have in the next 10 years? That's awesome. Congratulations to the news, Scaldi, Artexca, the big release with HPE Exclusive. For six months, Chris Tinker, Distinguished Engineer at HPE, great to see you. Jerome McCat, CEO of Scaldi. Great to see you as well. Congratulations on the big news. I'm John Furrier with theCUBE. Thanks for watching. Thank you.