ESG Analyst Weighs in on Hitachi Virtual Storage Platform 5000 Launch

Scott Sinclair, Senior Analyst, ESG:

Hey, thanks for having me. Great to be here.

Ian Clatworthy, Global Product Marketing Manager, Hitachi Vantara:

Thank you so much. So you actually joined us at NEXT 2019 when we brought this new solution to market, the new Hitachi VSP 5000 series. And it was part of a big launch for us of lots of different new products and solutions coming to market to help companies with their digital transformation and really that move to DataOps. You were there for the whole launch, I believe. Could you give us some information on how you felt the 5000 series was delivered?

Scott:

I am a big fan of the launch — really enjoyed it. I think one of the things that Hitachi Vantara has done over the years is really focused on the message, and the story has always been very technical. You guys have always had a great technical capability. But what I liked about this launch was you took that one or two steps further and really called out the capabilities of the product. There was a ton to like with the launch.

Ian:

No, that's really good. Thank you. I think for me as well, this is for us, this is a real step where we've listened to what our customers are telling us and certainly for the real true high-end enterprise space and our big enterprise customers. They've been looking at new technologies that are in the market, like NVMe , like SCM media for flash. They are saying: Look, we want these things but maybe we're not just ready right now. We were like: Well, how can we help customers with a nonsiloed product that locks you into a technology? Yet, at the same time, because the cost involved when you've got services to migrate from, from an existing to new, is … we understand that that's a lot of work and a lot of expense and worry for some customers. So the key tendency of the 5000 series was that ability to mix NVMe and SAS technologies so you can not only scale up, which we do very well, but you can now also scale out with performance. And for me that's the real key thing. You can buy SAS today because that's what you're ready for. Actually, when you want to go NVMe over fabric … you just scale out and it's just a code change to get you those capabilities. And for me, I think that's the real value that this platform can bring … that you can consolidate more than ever before on one solution.

Scott:

I'm going to cut you off. Absolutely going to agree. I think when you see the announcement and if you just kind of glance at it briefly and you see the big numbers and all the really impressive hero numbers that the VSP 5000 is, you think, okay, this is just another huge enterprise frame. Then I work out my old frame. I've spent a bunch of money and I bring in a new frame and once you double-click into it or really look behind the covers, that's not what this is at all. I mean, it still has the amazing performance, amazing scale, but the unsung hero of the VSP 5000 is the accelerated fabric on the back end. So what that does is, because you've built it around this very high-performance, low-latency fabric, you can scale out and plug in different types of modules that can be NBME all-flash for that super low latency performance. Or, you can go for a lower cost option for lower cost capacity. So, what this does is, it allows for IT organizations to get into the VSP 5000 that's more cost effective. Something where you can build and grow, but that also gives you the headroom and scales to where you can expand and integrate new technologies as they come in. All essentially being seamless to the workloads which is really impressive because in this we've been talking about data growth forever, right? That's a big thing. Gosh, I think every slide, every company has been talking about data growth for a decade. But what people are really finding out is not only are we in this era of insane data growth, which is continuing, but now, also, people are with all these high-value workloads, analytics, artificial intelligence, all of these things. Everyone's realizing: Oh man, we've got to actually use that data. So, we're seeing these giant spikes in performance demand. It's hard to project what your performance needs are going to be, you know, six months, nine months out. It's just not feasible to lock yourself in a frame anymore. That's what's great about the VSP 5000 series: You don't lock yourself in. There's so much flexibility and you can scale quickly both in capacity and performance.

Ian:

Sure. You've touched on a very good point there because I think it shows the enterprise capability now: 30 years of us being in this industry, and the innovation was still bringing to storage, the Hitachi Vantara accelerated fabric. It's a unique to us. We're not using any kind of off-the-shelf componentry. This is something that we've designed and it gives customers that flexibility. And that goes back to what you said. We're a true engineering company and I love that about us. We're really bringing value to our customers.

Scott:

I've had that, but on that point, that's the thing with Hitachi. You know what's funny? We're talking about the flexibility, but at the end of the day, it almost goes without saying that it's going to have that high-level, super-mission-critical resiliency, right? It's got, you know, support for main framework and environments. It has everything that you would expect and that people have come to know about what Hitachi can do at that high end, but really just delivers it in this package that is …. it's not a frame. It's this almost this evolutionary data platform that can evolve and change based on your needs. And that right there just delivers so much simplification right now. One of the things we've seen in our research is the No. 2 biggest skill shortage right now for IT organizations is IT architecture and planning, right? It's very difficult to figure out how to get all these technologies together and architect it for your workloads. The No. 1 skill shortage is in cybersecurity. That should be no surprise to anybody. But, No. 2 is: How do I get all these technologies to work together? What makes that harder is when you have to lock yourself into some architecture or some frame: that is, where you're trying to project what your needs are going to be three, four years out. And so with something like this, what's nice is you can deploy an architecture or this framework or platform that is able to adjust and evolve as your needs change. And because of that you just get far more flexibility. But at the same time consider your hero numbers: 21 million IOPS, 16 and grow to 69PB. You know, there are very few work organizations that are going to ever exceed those limits. You can go into VSP 5000 knowing no matter where our business goes with data, we're confident the VSP 5000 series will cover it.

Ian:

Yeah. And I think you've touched on a very important point there, which is simplicity. And one of the things that was a big announcement at Hitachi NEXT was our new Hitachi Ops Center software suite that is absolutely part of our VSP portfolio and really enables the 5000 series to become that simple cloud consumable platform. I mean, there are few people in the market here that can offer one single holistic management interface and replication from entry all the way to enterprise. And again, that's our strategy and depending on next-generation platforms, too … it's really a single holistic environment, which I think is super powerful.

But, to add to that, the simplicity has to be around, as you said, with these new workloads coming on with analytics and insights and machine learning. How do we make this simple for our customers? And it's all down to automation. And that's exactly what Hitachi Ops Center brings to the party. We know that sometimes these large solutions are very difficult to implement and manage. And, to that extent, people don't have teams sitting around looking after these platforms anymore. The world's moved on. So we have to think, okay, well, do we go and reinvent our software stack and look at new management products, et cetera, and design new management systems? And we came to this conclusion, which is: No. We really want to make this a very cloud consumable style solution, with lots of different terms in our industry PPU sets, et cetera.

Actually, in this terminology I'm talking about, if I go into AWS as any user, I can spin this up with a credit card and I can create a workload. It's exactly that capability we wanted to bring into the 5000 series. And we've done that with our integration with our open APIs and with automation and analytics in the box. We can integrate with IT ticketing systems. You go into your IT ticketing system and say: I need a new workload. And it's literally automated through the process. You can set your parameters, you can set the level of security and replication you need on that particular workload, and off it will go. It will go and deploy that on the right media it needs to be on, depending on the performance requirements on the box, looking at live analytics. And that I think is really powerful.

Let's make those simple things that actually take a great deal of time in someone's work week and just make it automated, give it directly to the customer. You know, our customer's customers.

Scott:

I absolutely agree. It's funny, the term simple gets used so often, but the definition of what simple needs to be, has evolved tremendously. It was only a few years ago when people said, we're simple because it's only seven clicks now instead of 12 clicks, or something silly like that. Those still exist, but what we're seeing now is just this need to wear IT; it's almost all the digital business demands for many companies now. IT can't deliver fast enough.

We did some research earlier this year. We looked and asked line of business executives: How do you — what do you think of IT? One in four, 25% of these organizations' line-of-business executives considered IT a business inhibitor, the preventer of digital services. I think there's a comic strip that has the IT guy listed as that. That's one in four. To put it in perspective, only 6% of organizations thought IT was competitive differentiator. When we asked these companies, these line-of-business execs, okay, what is it about it that makes them an inhibitor? The No. 1 answer, the most common answer, was IT services are just not getting delivered fast enough. If you think about that, take a step back. IT that you know, the role and responsibility of an IT organization, of a CIO, has changed. The goalposts have moved, as it were. It used to just be okay, uptime. We just got to make sure everything's up and running and everything's fine. That's not enough anymore for line-of-business executives. You have to keep everything up. You've got to keep it resilient, you've got to protect all the data, but now you have to do things quickly. You have to speed it up. You have to expedite services. Part of that is because of the perceptions of cloud, but also part of it is just because now more and more businesses have tied IT services in the use of data to revenue projects. The last thing anyone wants to see is, well, this revenue opportunity for the company has been delayed a year because IT can't spin up services fast enough, for example. So, because of that, any ways that you can automate, deliver automation, and expedite IT services are insanely valuable.

And what's really fascinating out of this is every now and then I run into someone who says: Wait a minute, automation? That's going to kind of eliminate my job. Well, that argument has just become laughable. I've already talked about how IT architecture and planning is the No. 2 skill shortage in IT right now. But as some people that follow this note: If you look at college graduation rates and if you look at the way data demands are growing, we're going to have a huge shortage of people to manage all the IT services that we need to manage. So the only way to get there is through automation. Also, companies really want their IT people to start leveraging and figuring out how to use data more, how to help the business leverage data to create new opportunities. The only way to do that is to do more automation around storage provisioning and more basic blocking and tackling within IT.

Ian:

That is just fascinating for you to say that because I think for us one of the key questions was: How do we bring together for our customers, not just a point solution? So, we're not looking at a large frame for open workloads. It has to be, and again, not a siloed solution for mainframe, or even for the new sort of container workloads that are really becoming mainstream now for a lot of customers. For us it was bringing all of those together. So, it's not only if you're going to do large-scale consolidation. Let's not just say it does large-scale consolidation. What does that mean? Well, Mr. Customer, it means that you yourselves can run traditional mainframe workloads alongside your mission-critical open workloads, your LLTP databases, and so forth. Actually, you can also use this for your development teams and for those container-style workloads that are driving new business for you.

And I think part of that is, and again, you touched on it: Well then hang on, doesn't that mean I've got to make a huge investment up front because I'm buying a solution that kind of does everything? And for us it was like: Okay, how do we make this large-frame-style idea start off small? So this is where the VSP 5100 comes in, which is our entry-level system. It includes that fabric acceleration and allows you to upgrade and scale out to upgrade to a VSP 5500 and then scale out up to three blocks in that solution. So, and again, as you said, we can mix SAS, intimate SAS and MVME. We can put in a spinning media, we can tear from MVME to spinning media. You know, we have this. Obviously, we support SAS flash as well. So for those people who want to buy flash today … that works for them.

But I think the key thing with all of that is we don't want customers to then throw away solutions that they've already invested in. You know, if they've got a tier two solution, it works fine, but it's sitting out there as a silo. You've got this great idea from a touchiness, it's all together. We're migrating workloads. But actually we do understand that there are still going to be use cases where you're not going to want to consolidate everything on one platform. But you still have that challenge of automation, analytics and also management. So one of the key tenants from us was what we have done for many, many years now: offered a storage virtualization. And, really, that comes again to fruition even more powerfully with the VSP 5000 … that ability to virtualize a third-party solution, allowing it to basically appear as a VSP. We can tier to a third-party solution and even offer data reduction services on that third-party solution. It becomes part of the Hitachi VSP family as a virtual instance, with all the same persistent data access and everything that you get from the underlying technologies of all the VSP platforms. And I think that really comes back to the forefront here with this solution.

Scott:

Oh yeah. Absolutely. You know, I was trying to figure out the right time to break in, but I wanted you to finish your thought. I mean that's insanely valuable. When I talk to CEOs, one of the big challenges is, whenever you come in with a new solution, we have the next great thing. One of the big problems is: Wait a minute, I just bought the next great thing two years ago. What did my throwing all that stuff away … what am I doing? And what's great about your platform is, not only does it give you the flexibility to scale, start small and add in new technologies, but also, as future technologies come in it rolls right in within the Hitachi family. But, essentially, if I've invested in a third-party array and it is starting to … if it's getting late in life or maybe I'm looking at some way to retire that … what I can do is I can put that in as a, I think you called it a virtual instance. I can bring that underneath the platform and still use that maybe as a secondary, maybe for extra capacity … some other way to still utilize that asset because at the end of the day, many of these systems still last longer than functionally what their useful life is in terms of being a performance leader. Being a capability leader is relative to other innovations, right? So, what you're able to do is you're able to give people the latest technology, the MVME, flash technology in there, but still use some of this older storage that they already have and say: Hey, you can still take advantage of that. So, those assets still have value and still have use. Not only is there tremendous savings for IT organizations in terms of management because it's all within one management framework, but also tremendous savings from a capex standpoint because now I don't have to throw all this stuff away. I don't have to retire. I can still use some of that, some of my existing storage systems as part of the overall ecosystem.

Ian:

You're right. I want to just circle back to one thing that you picked up at the very beginning of our conversation. You said that the No. 1 issue for customers on resources was around security and data protection. And are you seeing customers looking now more holistically at solutions? Where is security baked into a platform rather than necessarily being a third party or an overlay on top of a solution?

Scott:

Absolutely. One of the jokes I have is: At ESG we're an analyst firm. We do a bunch of research. You could ask any question to IT organizations … doesn't matter what it is. And if one of the options is security, it will be the most common selected response. That is how front of mind security is for everyone in IT. And it should be. I've talked to many organizations, many storage administrators, that have said our business, our IT organizations, made security, data security, job one. And so what that means is there are challenges not only within hiring (How do we get the right data security people?), but also in ensuring (How do we investigate every single type of storage technology, infrastructure technology and see, are we doing enough to secure our data?). And that should be and is, in most organizations, part of everyone's job.

Ian:

Yeah. You know, and for us, we were hearing this as well. When the design team really took the 5000 series off the drawing board and started to re-engineer the solution, we could hear that from our customers, too. They said: Look, it's all well and good saying, you've got data protection built in and you can do that on a workload basis, but actually, more fundamentally, we're looking at the platforms themselves and making sure they're hardened enough for local attacks as well. And we've tried really hard with the 5000 series to really beef up that security. So, not only have we got full encryption that's baked in, data protection with Hitachi Data Instance Director, which can give you live recovery on your application layer, but right away, we're FIPS 140-2 compliant on this platform and these are pretty industry-standard things.

But, what we want to do is go further and we wanted to push this even more. So, we took a look at our own solutions and we said: Well, okay, the current systems … where do we see we could improve them even more? So we've really hardened that capability for hacking. You know, if someone's on that network and they want to hack into the interface on the 5000 series, they're going to have a very, very hard time because we've decided to hide as much information as we can and really make it exceedingly difficult for that to happen. It's super, it's the best in the market for security. But actually we want more. And we wanted to really take a lead in this area to just make our solution really a volt for our customer's data, you know?

Scott:

Oh yeah. It's so great to hear and it's a difficult message, I'm sure, to give you your customers because at a high level it comes off as: "Hey, we're more secure." But at the end of the day, these are essential things that you guys are doing, which are insanely valuable.

So for example, I want to say it was earlier this year, some small storage start-up, very, very tiny storage start-up, had an exploit and it caused some problems. I was talking to some people in the press and I said: What do you think this means? And I said, the problem with these types of exploits is, it's not like you can say: "Oh, we offer this feature," which is going to stop all exploits because the "feature" (people that are listening to this won't see me do the air quotes when I said feature), the feature as it were, is just really good software development and following excellent security practices. And so what that means is that if I'm an IT buyer and this starts to become more and more of a problem, then I sit down and I say: You know what, I'm going to go with a company like Hitachi that I know has been building mission-critical enterprise systems for a long time and that I know does these high-quality security practices within their development procedures. Because at the end of the day, those are the things that don't show up in the datasheet very well. It's hard to put good security development practices, for example, right next to 21 million IOPS and really have it shine in age or in a marketing blurb. But at the end of the day, that is where it honestly may be the most valuable portion because it's saying: Hey, you guys are going that extra step to make sure that your customer's data is protected. At ESG, our founder Steve Duplessie has a saying where he says, the great thing about cyber security is it's that or data security. It's a problem that can never be solved because there are always people out there that are actively trying to break it. That it is something that is a problem that requires 100% diligence from everyone all the time. And so companies like yourself that are going this extra mile … it is insanely valuable.

Ian:

I think you summed up really nicely there and you know what, I'm conscious. We've taken a lot of time with these good listeners here, too.

Scott:

Well the time's just gone by.

Ian:

Well it's been an absolute pleasure to have you on this podcast, Scott. Hopefully, we can do a few more of these and hopefully for the audience out there, you get some value from having someone like Scott who has the most incredible industry knowledge. So Scott, firstly, thank you for you. Appreciate it.

Scott:

Well, thank you for having me. And this was fun.

Ian:

And look, can I count on you being at NEXT 2020?

Scott:

Oh yeah, absolutely.

Ian:

And there's a message to everyone out there. You know, you've got to come to this. This was starting to become more than just an industry event … a vendor event. This is more of an industry event now where we really are looking at the future of data, everything from edge to cloud. So, it's really worthwhile. And Scott, again, thank you so much for your time.

Scott:

Thank you.