Interview B&F has been talking to Anthony Lye, NetApp EVP and GM of its cloud business, about making the Data Fabric vision a reality with ONTAP services available in the three main public clouds.
But it left customers running their virtual datacenters in the public clouds in a manual and time-consuming way, he tells us.
Lye says he wanted to have NetApp supply CloudOps services to customers so that they could manage their virtual datacenters as automated app infrastructure platforms.
And, of course, the company isn’t the only one in the game. “It’d be nice if I was the only person who had figured it out. But I’m not. I think CloudOps is becoming a really, really big space. I think you look at VMware, you’ve got to look at ServiceNow. I think you have got to look at IBM – it acquired Turbonomic. ServiceNow acquired Lightstep, VMware built Tanzu. You’ve obviously got to look at the public clouds themselves; look at the whole FinOps movement.”
Tanzu is a CloudOps thing? B&F thought Tanzu was just a way to get cloud-native apps to run inside a virtual machine.
Lye retorts: “That’s not a bad description of what it really is. But VMware, I think, wants to be the platform for modern applications as it was the platform for legacy applications.”
And a platform is more than a VM containing microservices.
NetApp appears to have chosen to buy in CloudOps expertise rather than building it in-house and has acquired eight companies.
How do they know which cloud ops companies to acquire? “A lot of our data comes from customers. We talk to the analysts a lot, to Gartner, Forrester and GigaOm,” says Lye.
He mentions SREs (Site Reliability Engineers) and says: “I obsess about that person. I’ve done time and motion studies, like watching them work… I’m trying to build like a cockpit for that person really.
He goes on to mention high-tech marketing guru Geoffrey Moore’s analogy from Crossing the Chasm. “We’re in the early adopter phase right now. And in the early adopter phase, people see CloudOPs – CICD – as competitive advantage. And so they’re prepared to buy point solutions and stitch them all together.
“But what happens with successful things is they cross this chasm. And now they get adopted by the early and late majority.”
But point products are not enough. “When you look back at every product in the technology industry that’s crossed the chasm, Larry Ellison would always say, ‘Suites always win’.”
People don’t want to manage the integration cost and “that’s what I’m building for. I’m building an integrated suite that’s platforming a set of independent point solutions.”
NetApp has a healthy lead of three to five years on storage competitors in the CloudOps market and powered ahead while Dell, Hitachi Vantara, HPE, and IBM have started their operations.
Lye thinks the total addressable CloudOps market is about $11 billion and growing at a mid-twenties CAGR. He said: “I joined five years ago [and] the business has gone from less than a million to last quarter’s $469 million. We disclosed at the financial analyst day that almost half of that business was CloudOps now.
“And we didn’t even start in CloudOps until two and a half years ago.”
Why? He discussed it with NetApp CEO George Kurian.
“I kept saying to George, ‘There’s another side of the cloud that’s really big, that we are not serving’.”
Lye says he “went to talk to all of my apps friends. And they would literally go, storage? No, not interested in storage.”
He asked them what they were interested in.
They said they were deploying applications on infrastructure. “They would basically walk me through a set of steps. And those steps always started with compute. And then compute, the wizard, would pick storage partners.”
Storage must be a feature of compute, but it’s also secondary.
Lye says: “I’m sitting behind someone else in a decision, and I don’t want to be doing that. The primary things are that you define the price, performance envelope, more compute, and storage, and network. And if you can see all of those things together… you can do things to applications that nobody else can.”
A key starting point was buying Spot, the cloud broker that optimized (lowered) effective cloud prices, in June 2020.
Lye’s app contacts also said: “I spend too much money. Every application person in the world basically says this cloud freedom is awesome [but] I get sticker shock. When my application people pick, this is the stuff they need to use, they pick something and then double it. Because they don’t want to have any issues. So they’re always over-provisioning.”
Here’s the killer point: “And the customers would say, isn’t there a tool that can actually listen to my application and pick all the infrastructure that it needs? And we went out and looked around and found Spot. And I bought Spot.”
Does Lye think the CloudOps opportunity on its own could become a billion-dollar-a-year business?
He said: “Yes, without doubt. Everybody’s doing it.” But that doesn’t mean the on-premises world is going away, far from it. Lye says there are a few things the public cloud can’t do, such as overcoming data gravity and providing low-latency processing for on-premises machinery.
“If you’re running a car manufacturing plant, and you’ve got these robots doing all these processes, they’re not making web service calls to Amazon. Nothing would ever get built. There are tonnes and tonnes of low latency requirements. And there are tonnes and tonnes of data center requirements that don’t make sense for public cloud.
“What you’re seeing is the data that’s being generated in these on-premises environments is growing very, very quickly. And so actually the on premise business for us, is a growth opportunity. It’s not the high double digits that cloud is, but it’s growth.”
NetApp’s public cloud business buttresses its on-premises business. “The more I can show customers how committed we are to the public cloud, the more confidence they have in buying an app on-premise. Because I’ve given them symmetry, I’ve given them flexibility. Yes. Nobody else has done that.”
What Lye has also done is helped NetApp become the pre-eminent hybrid multi-cloud data storage services business and then gone a step further and developed the $235 million run rate CloudOps business. It is now building a CloudOps suite, under the Spot brand, and this gives it an opportunity to sell to new, non-NetApp storage customers. That’s extra revenue for NetApp and it can also lead to extra on-premises sales of NetApp’s storage products. Other mainstream storage company with hybrid cloud strategies must be taking notes.
