 From theCUBE Studios in Palo Alto in Boston, connecting with thought leaders all around the world, this is a CUBE Conversation. The external storage array business, as we know it, has changed forever. You can see that in the survey data that we do and the financial information from the largest public storage companies. And it's not just because of COVID, although that's clearly a factor, which has accelerated the shifts that we see in the market. Specifically though, CIOs are rationalizing their infrastructure portfolios by consolidating workloads to simplify, reduce costs and minimize vendor sprawl so they can shift resources toward digital initiatives that include cloud, containers, machine intelligence, and automation, all while reducing their risks. Hello everyone, this is Dave Vellante and welcome to this CUBE Conversation, where we're going to discuss strategies related to workload consolidation at petabyte scale. And with me is Dr. Rico, who's the vice president, office of the CTO at Infinidat. Welcome back to theCUBE. Doc, always a pleasure to see you. And great to be here, always a pleasure to work with you, Dave. So, Doc, I just published a piece over the weekend and I pointed out that of the largest storage companies, only one showed revenue growth last quarter and that was on a significantly reduced compare with last year. So my first question to you is Infinidat growing its business? Oh, absolutely. It's been a very interesting year all across, as you can quite imagine. But our footprint is such that with our elastic pricing models and the fact that we've got excess capacity in almost every single system that's out there, we were really given our customers an opportunity to take advantage of that to increase their capacity levels while maintaining the same levels of performance and availability, but not have to have anybody on premises during this crazy COVID struck era. Yeah, so you're bringing that cloud model to the data center, which has obviously been a challenge. I mean, you mentioned the subscription, sort of like pricing. We're going to get into the cloud more, but I wonder if we could step back a little bit and look at some of the macro trends that you're seeing in the market and specifically as it relates to on-prem storage strategies that CIOs are taking. Yeah, you know, it's been interesting. We've seen over the course to the past five years or so, certainly a big uptick in people looking at next generation or what they believe and perceive to be next generation storage platforms, which are really just evolutions of media. They're not really taking advantage of any new innovations in storage and notwithstanding our own products, which are all software driven. We've talked about that before, but what's really happened in this past year, as you said, CIOs and CTOs, they're always looking for that next point of leverage and advantage and they're looking for more agility in application deployment. They're looking in a way to rapidly respond to business requirements. So they're looking very much at those cloud-like requirements. They're looking at those capabilities to containerize applications. They're looking at how they can shift out virtual machines if they're not directly in a container and how the storage, by the way, can have the same advantage. And in order to do so, they really need to look at storage consolidation. I think, Dave, to sum it up from the storage perspective, I love Ken Steinhart was recently on a video and he was challenged that people aren't looking at spinning rust with a derogatory way of referring to disks. And Ken so rightly and accurately responded, yeah, but people aren't really looking for QLC either. What they're looking for is performance, scale, availability, and certainly cost-effectiveness and price. Yeah, it's like I said up front, Doc. I mean, if you're a C-level executive today, you don't want to worry about your storage infrastructure. You got bigger problems to worry about. You just want it to work. And so when you talk about consolidating workloads, people often talk about the so-called blast radiation. In other words, people who run data centers, they understand that things fail. And sometimes something as simple, it might be a power supply can have a catastrophic downstream effect on application availability. So my question is, how do you think about architecting systems so as to minimize the effects of component failures on the business? Yeah, you know, it's a very interesting term, Dave. Blast radius, right? We've heard this referred to storage over the last several decades. In fact, when it really should refer to the data center and the application infrastructure. But, you know, if we're talking about just the storage footprint itself, one of the things that we really need to look at is the resilience and the reliability of the architecture. And when you look at something that is maybe dual controller, single or double power supply, there are issues and concerns to take into play. And what we've done is we've designed something that's really triple redundant, which has typically only been applied to the very high end of the market before. And we do it in a very active, active, active manner. And naturally we have suggestions for best practices for deployment within a data center as well. You know, multiple sources of power coming into the array and things of that nature. But everything needs to be this active, active, active type of architecture in order to bring those reliability levels up to the point where as long as it's a component failure within the array, it's not going to cause an outage or data unavailability event. Yeah, so imagine a heat map. I mean, people talk about the blast radius. So imagine a heat map is green. There's a big, you know, there's a yellow area and there's a red area. And what you're saying is as far as the array goes, you're essentially eliminating the red area. Now, if you take it to the broader installation, you know, that red area, you have to deal with it in different ways, remote replication, then you get it to sync and async. But essentially what I'm hearing you say, Doc, is you're squeezing that red area out so your customers could sleep at night. That, absolutely, sleep at night is so appropriate. And in fact, we've got a large portion of our customer base is they're running mission-critical businesses. You know, we have some of the most mission-critical companies in our logo portfolio in the world. We also have, by the way, some very significant service provider businesses who are providing mission-critical capabilities to their customers in turn and they need to sleep at night. And it's, you know, availability is only one factor, certainly manageability is another because, you know, not meeting a service level is just like data unavailability in some respects. So making manageability as automatic as it can be, making sure that the system is not only self-healing but can respond to variations in workload appropriately is very, very critically important as well. Yeah, so that you mentioned mission-critical workloads and those are the workloads that, let's face it, they're not moving into the cloud, certainly not in any big way. You know, why would they? Generally, you know, CIOs, CTOs, they're putting a brick wall around that saying, hey, it works. We don't want to migrate that piece, but I want to talk more about how your customers are thinking about workload consolidation and rationalizing their storage portfolios. What are those conversations like? Where do they start? And what are some of the outcomes that you're seeing with your customers? Yeah, I think the funny thing about that point, Dave, is that customers are really starting to think about cloud in an entirely different way. You know, at one point cloud meant public cloud and meant this entity outside the walls of the data center. And people were starting to use services without realizing that that was another type of cloud. And then they were starting to build their own versions of cloud. We were referring to them as private clouds, but they were really spread beyond the walls of a single data center. So now it's a very hybrid world and there's lots of different ways to look at it. Hybrid cloud, multi-cloud, whatever moniker you want to put on it, it really comes down to a consistency in how you manage that infrastructure, how you interface with that infrastructure, and then understanding what the practicality is of putting workloads in different places. And practicality means not only the latency of access of the data, but the costs associated with it. And of course, the other aspects that we talked about, like the availability metrics. And as you increase the availability and performance metrics, those costs go up. And that's one of the reasons why some of these larger mission critical data centers are really repatriating their mission critical workloads. At least the highest levels of them. And others are looking at other models. For example, AWS Outposts, which talked about quite a bit recently in AWS re-event. Yeah, I just wrote again this weekend that you guys were one of the partners that was qualified now to run on AWS Outposts. It's interesting is Amazon moves its model to the edge, which includes the data center to them. They need partners that can, that really understand how to operate in a non-premises world, how to service those customers. And so that's great to see you guys as part of that. Yeah, thank you. And it was actually a very seamless integration because of the power and capability of all of the different interface models that we have, is they all are fully and tightly integrated and work seamlessly. So if you want to use a CSI type model, to interface with your storage, you can with the infinite that. And we work with all of the different flavors. So the qualification process, the certification process, and the documentation process was actually quite easy. And now we're able to provide people who have particularly larger workloads, that capability in the AWS on-premises type environment. Yeah, now I implied upfront that cloud computing was the main factor, if not the primary factor, really driving some of the changes that we're seeing in the marketplace. Now, of course, it's not all pink roses with the cloud. We've seen numerous public cloud outages this year, certainly from Microsoft. We saw the AWS Kinesis outage in November. Google just had a major outage this month. Gmail was down, G Suite was down for an extended period of time, and that disrupted businesses. We were lying that schools, for example. So it's always caveat emptors, we know. But talk to infinite cloud strategy and mentioned hybrid. I'm particularly interested in how you're dealing with things like orchestration and containers and Kubernetes. Yeah, well, of course we have a very feature rich set of interfaces for containers, Kubernetes interfaces, downloadable through native stores. So they're very easy to integrate with. But our cloud strategy is that we are a software-centric model and all of the value and feature function that we provide is through the software. The hardware of Infinibox is really a reference architecture that we deliver to make it easier for customers to enjoy, say 100% availability model. But if you wanna run something in a traditional on-premises data center, straight Infinibox is fine, but we also give you the flexibility of cloud-like consumption through our pricing models, our elastic pricing models. So you don't need to consume an entire Infinibox day one. You can grow and shrink that environment with an OPEX model or you can buy it as you consume it in a CAPEX model and you can switch from OPEX over to CAPEX if it becomes more cost-effective for you in time, which I think is what a lot of people are looking for. If you're looking for that public cloud, we have our NutriX Cloud offering, which is now being delivered more through partners. But some businesses, and especially the mid-tier, the SMB all the way through the mid-enterprise are also now looking to cloud service providers, many of which use Infinibox as their back-end. And now with AWS Outpost, of course, we can give you that on-premises experience of the public cloud. Yeah, you guys were early on obviously in that subscription-based model and now sort of everyone's doing it. I noticed in the latest Gartner Magic Quadrant on storage arrays, which you guys were named a leader. I think they had a stat in there that said, I forget what the exact timeframe was, that 50% of customers would be using that type of model. And I guarantee by whatever timeframe that was 100% of the vendor community is going to be delivering that type of model. So congratulations on being named a leader. I will say this, there's consolidation happening in the market. So to me, this bodes well to the extent that you can guarantee high availability and consistent performance at scale. That bodes well for you guys in a consolidating market. And I know IDC just released a paper. It was called, I got a copy here. It's called a checklist for storage workload consolidation at petabyte scale. It was written by Eric Bergner, who I've known for a number of years. He's the IDC's VP of infrastructure. He knows his stuff and the paper's very detailed. So I'm not going to go through the checklist items. But I think if you don't mind, Doc, I think it's worth reading and excerpt from this, if I can, as part of his conclusions. When considering workload consolidation, IT organizations should carefully consider their performance availability, functionality and affordability requirements, of course. Few storage systems on the market will be able to cost effectively consolidate different types of workloads with different IO profiles onto a single system. But that is an infinite forte. They're very good at it. So that's quite a testimonial. Why is that your thoughts on what Eric wrote? Well, first of all, thank you for the kudos on the Gartner MQ, being a leader, second year in a row for primary storage only because that documents only existed for two years. But we were also a leader in hybrid storage arrays before that. And we love Gartner. We think they're a real critical, reliable source for a lot of large companies. And IDC, Eric, of course, is a name in the industry. So we very much appreciate when he writes something that positive about us. But to answer your question, Dave, there's a lot that goes on inside Infinibox and it's the neural cache capabilities, the deep learning engine that is able to understand the different types of workloads, how they operate, how to provide predictable performance. And that, I think, is ultimately key to an application. It's not just high performance, it's predictable performance. It's making sure the application knows what to expect. And of course, it has to be in performance. It can't just be slow, but predictable. It has to be fast and predictable. Providing a multi-tenant infrastructure that is native to the architecture so that these workloads can coexist, whether they're truly just workloads from multiple applications or workloads from different business units or potentially, as we mentioned with cloud service providers, workloads from different customers. They need to be segmented in such a way so that they can be managed, operated, and provide that performance and availability at scale. Because that's where data centers go. That's where data centers are. Great, well, so we'll bring that graphic back up just to show, obviously this is available on your website. You can go download this paper from Eric from IDC, www.infinidad.com slash en slash resource. I would definitely recommend you check it out. As I say, Eric's been in the business for a long, long time, so that's great. Doc, we'll give you the last word. Anything we didn't cover? Any big takeaways you want to share with the audience? Yeah, I think I'll go back to that point. Consolidation is absolutely key for not just simplicity of management, but capability for you to respond quickly to changing business requirements and or new business requirements. And also do it in a way that is cost effective. Just buying the new shiny object is expensive and it's very limited in shelf life. You're just going to be looking for the next one the next year. You want to provide something that is going to provide you that predictable capability over time because frankly, I have never met a CXO of anything that wasn't trying to increase their profit margin. You know, that's a great point and I would add, I mean, the shiny new object thing, look, if you're in an experimental mode and playing around with artificial intelligence or automation, areas that you really don't know a lot about, you know what, check out the shiny new objects. But I would argue, you're on-prem storage, you don't want to be messing around with that. It's not a shiny new objects business. It's really about making sure that that base is stable and as you say, predictable and reliable. So doc, Tariqo, thanks so much for coming back in theCUBE. Great to see you. Great to see you, Dave, and look forward to next time. All right, and thank you for watching everybody. This is Dave Vellante and we'll see you next time on theCUBE.