This article is more than 1 year old

NetApp CEO Kurian: Buying SolidFire was a long-term strategic bet

Plus: Avoiding hard decisions reduces a company to mediocrity

Interview Chief executives of multi-billion dollar corporations have to have several characteristics, including steel, the ability to see things as they are, and possession of a personal value system that meshes with their corporation’s culture.

Some CEOs, when promoted to the top chair after long employment in a corporation, sometimes lack the ability to see things without corporate lens bias.

NetApp CEO George Kurian gained his seat at the top table through an internal promotion after the previous internally-promoted incumbent was found wanting. Does he have these abilities, particularly the ability to take the hard decisions?

El Reg talked to him at NetApp's HQ in Sunnyvale to get a closer look at the man tasked with resurrecting the company's growth.

Questions and answers have been lightly edited for brevity and flow.

El Reg What did it feel like as a 17-year old coming into Princeton and President Reagan's America from India?

George Kurian The world’s gotten a lot smaller since when I came. I think at the time that I came it was quite a cultural shift between the US and India. I think for example, I’d never seen an ATM, a bank teller machine before, I’d never seen a computer before, so those were new experiences. You found that in the United States there was an astonishing gap between rich and poor. We learnt that quickly when we went to New York for the first time in the mid-eighties.

El Reg And then you did your best to cross it.

George Kurian That’s right. It was astonishing to see the gap between rich and poor in the US, but you also realised that one of the things that has been an extraordinary part of the American story; the ability to, you know, move yourself up if you are willing to commit to doing the hard work, and if you get a few lucky breaks.

Early career

El Reg You went to Oracle. What did you do there? Did you write code?

George Kurian Yes, I did write code. I wrote product design specifications. I wrote code, tested it, you know, brought it to customers. And then I had an opportunity to do a little bit of product management for a while, and so learnt all of the different aspects about what it takes to build a good product, you know, team of people, understand customers’ requirements. Sometimes it’s not obvious what they’re telling you and you’ve got to take time to listen carefully.

El Reg How did the MBA come about?

George Kurian My eldest brother and my dad were both mentors to us and their feedback, which was important, you know, as part of our development, was that in the early part of your career you want to broaden your experiences, unless you’re absolutely certain that the particular thing you’re doing is your life’s passion, and so they said, hey, why don’t you try out a few different things. You’re young enough. You don’t have a family. You don’t have these obligations and commitments, you know. So my brother and I both applied to college and we got into Stanford. He went the year before I did. I stuck back for another year to make sure there was enough financial resources for the two of us to be able to go.

El Reg Were the two of you living together?

George Kurian Yes, we did. We lived together for a long time.

El Reg So you helped fund his MBA to an extent …

George Kurian That’s right

El Reg And then was the favour returned?

George Kurian Yes, that’s right. Yes. that’s exactly right.

El Reg So family is very, very important to you?

George Kurian Yes, absolutely ... My brother lives nearby here. I live in San Francisco. Thomas lives in Menlo Park/Atherton area, and we’re still extraordinarily close. He has a little boy who is eight, and I have a girl who is nine and boy who is twelve, and it's really beautiful to watch the next generation get together and be the best of friends and run around.

Community

El Reg Community is important to you. How does that mesh with things at NetApp?

George Kurian I think community, you know, is something that brings purpose to people’s lives. It makes you realise that what you’re doing is party of something bigger and makes you sensitive to the world around you. NetApp is very much a values-led, values-driven company, and one of the values that we espouse is the One Team value, which is often talked about in a lot of places but really is a core tenet of how we operate as a company, and it feels natural to be part of that.

El Reg Is that something you found at Oracle and Akamai and at Cisco?

George Kurian There were certainly teams within Oracle and Cisco ... but I don’t think the whole company operated as a team. Certainly they’re much larger institutions than NetApp,[and] the engineering teams or the people that formed a part of a business unit were really tight. But the broader company was really a group of teams.

We’ve got that idea of a community [at NetApp] and one of the things that I’m continually aware of is that, in my role, it’s essentially to help the team progress the company forward. But you’re really a steward in a line of leaders, you know ... you’re carrying forward the good things that many people have done and have built into the company, and then, yes, you have got to change some things as well.

El Reg Do you believe in putting something back into the community? Are you involved in any returning something back to the community activities?

George Kurian You know, one of the things of having grown up in India is that you are confronted by the luck and the grace that you’ve been given, and so we’re always trying to find ways to make sure we can give back to people in some form or other. I, you know, teach classes at the University of Santa Clara Business School**. I co-teach a marketing class there so that we can help, you know, bring on the next generation, give back to the next generation of talent.

You try to mentor young people who are in the early stages of their careers and try to give them advice and help and check in with them when they are going through hard times, or are trying to make tough decisions. I was just on a phone call with somebody this morning who was looking for my advice on something. And then, you know, we’re certainly involved with a whole range of community activities as part of my children's school. They’re at a Quaker school in San Francisco.

El Reg I did want to touch on that, just as little bit. I wondered how important Christianity as a religion was to you.

George Kurian My faith is a personal topic. My faith is part of the value systems I have and I prefer to live out my faith in the actions I take rather than talking about it. I think it’s a really important part of the value system my parents ingrained in us. I think the Quaker school my kids go to; the humanitarian tradition of the Quakers is something of great importance to me.

SolidFire and hyper-converged systems

El Reg How come NetApp was late to the all-flash array market?

George Kurian There were a couple of reasons for ... us being late to the all-flash market. I think certainly the most important one was that we were really focussed on completing our clustered storage systems, right. That was ... an all-in, all-consuming exercise. And we’ve been very successful with that. I think that what it allowed us to do subsequently is to catch up to the market very quickly because we could add solid-state to it.

El Reg SolidFire won’t contribute to increasing NetApp revenues for some years. Why was it bought?

George Kurian SolidFire allows us to take on a very important part of our customers’ technology roadmap. It enables them to build data centre environments that resemble what Google or Amazon want to build, and many of our customers are increasingly going to spend a higher percentage of their budgets on those types of environments rather than the traditional kind of environments.

We think that there’s enormous opportunity there, and SolidFire was clearly the best architecture for the job. … it’s really about portfolio management.

Without taking these long-term, strategic bets you don’t make progress as a company.

El Reg What’s your view on hyper-converged arrays?

George Kurian I don’t think hyper-converged replaces converged at all. ... They are two different design paradigms and there are benefits to both models, but they play in different parts of the landscape.

I think that in the hyper-converged model, the core value proposition of what it offers customers is the ability to radically simplify their environment and to get to rapid time to value .... I think it doesn't have the flexibility to do multi-workload consolidation, and some of the things that some of the hyper-converged vendors talk about I think will cause them to lose the unique benefits of simplicity.

SolidFire gives us a chance to offer the same benefits to customers because, ... it is essentially a server-less SAN product. It doesn't have a classic Fibre Channel architecture and it's built on on a scale-out paradigm. You'll see us continue to integrate that with solutions like VMware management systems to allow customers to have a different way to accomplish the same benefits that hyper-converged vendors offer.

El Reg So compute and SolidFire will come together in some way?

George Kurian I'll tell you honestly, the physical integration of compute and storage on the same processor I think has limited value. It’s essentially a new form of direct-attached storage, and has some of the same challenges that direct-attached storage has. It's great for a really small office potentially, but becomes really complex as you're trying to scale ... people want to scale storage at different rates from compute; they want to upgrade compute at different life cycles than they want to upgrade the storage and so on.

Fabric matters and the cloud

El Reg Does NetApp have thoughts about implementing NVMe as a fabric?

George Kurian I think latency is an important consideration as people think about systems and we’ll look at all the different ways to ensure that our systems have competitive latency graphs. I think NVMe has certainly made progress as a technology; it’s getting to a place where lots of the elements of what an enterprise-ready fabric solution is due for … it’s certainly been part of our systems thinking roadmap.

Having come from the networking world you’re always dealing with the two sides of the architectural conversation; the compute and storage guys that hang off the network keep making advances and then the networking folks make advances.

You get these step function improvements in either memory technology or latency and, in particular, fabrics, and our long-term view of that is that you need a consistent data management model that spans all of these landscapes and extends into the cloud because the range of environments where people put data, and share data from, continues to expand, and that concept is what we refer to as data fabric.

El Reg You would assert that there’s a very, very low likelihood that on-premises IT will move entirely to the public cloud?

George Kurian I think it’s a long time before an enterprise will move all of their applications to an external environment. I think that our view is it’s not just about on-premises to the public cloud. It’s also what we call multi-cloud.

El Reg So you use Data Fabric to manage a data pool [Correct] across Amazon, Azure, possibly Google, and on-premises as well?

George Kurian That’s exactly right.

El Reg You have Direct Connect with Amazon. Will you do that for Azure and possibly for Google?

George Kurian Yes. We do that with Azure. We that with some cloud providers in Asia:Pacific. We do that with telecommunications carriers. We do that with a lot of different folks... NetApp’s ONTAP platform runs on the Amazon cloud and you’ll see us making it available on other clouds in the future, for sure.

At times of industry change, like when the cloud was created, it’s important for us to embrace those transitions and put our customer’s interests first ... There’ll be a transitional period where customers try to figure it out with us, but I think over the medium to long-term we’ll be phenomenally successful.

Acquisitions

El Reg How do you view acquisitions?

George Kurian I’m not a person that believes that every technology that we sell to customers has to be developed internally at NetApp. Not at all. Not for one second. The vast majority of intellectual property is outside our company, and I actually subscribe to the fact that, if we find something that is of value to our customers and can generate an economic return for our shareholders, we would acquire it.

Object Storage

El Reg Object storage and StorageGRID; will that become an integral part of the Data Fabric as well?

George Kurian It is a part of the Data Fabric. It allows us to do two things. Certainly for customers that want to build their own object environment, we are able to enable them to build a highly scalable, highly differentiated object store.

It also allows us to enable customers that have born-in-the-cloud apps that want to move on-premises to be able to seamlessly do that. And it gets connected in to the back end of our AltaVault environment, ONTAP, SolidFire, all of them as an archival platform. So we’ve got lots of interesting ways we can offer solutions to customers with that.

El Reg Object storage has a future? It’s going to be pushed?

George Kurian We’re certainly optimistic about it. I think there are changes that need to happen to the application model that enables object storage to be a bigger market than it us today. We’re being disciplined about our investment around it, but we think that, by connecting it to these other technology platforms you can get the benefits of object storage, which is the ability to manage enormous amounts of data very efficiently without having to rewrite all the applications.

El Reg So an S3 interface is probably mandatory for that sort of thing?

George Kurian I think there’s an S3-class interface. I also think there is the connectivity to traditional POSIX or NFS front-ends and we’ve done both of them with our object platforms.

El Reg Is it regrettable that object storage can’t be presented as part of ONTAP?

George Kurian You’ll see us continue to use object interfaces on ONTAP for the right reasons. I don’t think we’re going to make it the primary object repository. But, for example, there are use cases where customers want to keep their hot data in a repository like ONTAP, but keep the large archive in an object environment, and be able to seamlessly integrate namespace and things like that across the two.

El Reg So ONTAP could pump cooler data to StorageGRID?

George Kurian Using S3, over a period of time you’ll see us do that.

Making difficult decisions

El Reg How do you approach taking hard decisions where people will be upset?

George Kurian First if all I think that you can’t shy away from those decisions, because I think if you shy away from them you end up, you know, reducing the company into mediocrity.

I always tell our leadership team it’s the role of leadership to make the toughest decisions, because if you can’t make those decisions then you essentially dis-empower everybody below you from making those decisions. So that’s the first. You’ve got to lean in on the tough decisions.

The way I take them on is ... to put yourself in the person impacted's shoes as you go through those decisions. You’ve got to make sure that in the conversation and the dialogue that you have around it that you are empathetic, that you can listen to their concerns but also hold your ground.

I always say that if everybody is happy about a decision then that decision was made too late. Because it was self-evident the decision needed to be made. And so what I am usually testing is are we pushing fast enough to make the tough decisions so that somebody or multiple people are left unhappy or at least saying ‘I’m not happy with the decision but I agree with it.’ I mean that’s the sentiment you’ve got to have around it.

Comment

There's no doubt Kurian has steel and no doubt that he can run a large organisation and make the tricky decisions needed. His personal values and sense of community integrate well with NetApp's culture. The company is making strategic bets with Clustered ONTAP and SolidFire the on-premises pillars of the company's Data Fabric strategy, the meshing together of private clouds (on-premises IT) and public clouds to provide a single IT resource pool.

Converged systems, via Cisco, are strongly supported by Kurian; note the Cisco background in his CV. Hyper-convergence is not supported with simply managed and operated SolidFire providing, Kurian says, equivalent benefits. Object storage has a bulk capacity role to play and NVMe over fabrics may feature in NetApp's product future.

Our feeling is that NetApp wants its customers to adopt the Data Fabric concept and will look to bring developing IT ideas into it, such as, for example, containerisation.

He's needing to arrest NetApp's revenue decline and betting on the product triumvirate of cDOT, the Data Fabric and SolidFire to take the business forward. The bets behind these bets are that customers retain a significant on-premises IT presence, with shared external storage playing a big role, and that they don't go all-in to a single public cloud provider. Seems reasonable and, if things start going the other way, Kurian will decide what to do to avoid NetApp fading into obscurity. He didn't take on the CEO spot to see that happen. ®

** Santa Clara University is a Jesuit University.

More about

More about

More about

TIP US OFF

Send us news


Other stories you might like