HP talks cloud delivery options, the importance of OpenStack, how it competes on price

An in-depth conversation with Bill Hilf, Senior Vice President of Product and Service Management for HP Cloud, about where Helion fits in, cloud consumption models and coming change.

1 2 Page 2
Page 2 of 2

Then even further up the stack we have multiple programming languages and frameworks people can choose from, from Python or Ruby or Java or .NET. That polyglot environment is important for us.

So we’re not only giving customers a choice of where to install and run their cloud, we also give them a lot of choice when it comes to the technology they can use because, at the end of the day, the VMware story is very vertical, the Red Hat story is very vertical, the Microsoft story, even though they talk a lot about open source, is really very vertical. Choice and a platform truly built on open source - that’s a differentiation for us.

If you’re pushing a high-end, enterprise-level story, why on the Helion website are you shouting about price so much? That kind of screams commodity.

As of 2014 less than 10% of enterprise IT is using cloud computing, so the growth opportunity is huge. And when you are trying to fight an early market battle for share, particularly for OpenStack oriented customers, you want to grab as much share as fast as possible.

One of the biggest advantages of a company like HP is we have all sorts of ways we can monetize. We don’t need to sell software at huge margins. We don’t need to sell a server for everything we do. We don’t need to sell services for everything. We have all kinds of ways we can make money through the broad HP. So that gives us a bunch of freedom, actually more freedom than I had at Microsoft because we can do things on every dimension to compete and aggressively grab market share.

And one tool we can use is price. So we can go undercut the other guy because our P&L isn’t solely based on software markets. We certainly compete with other OpenStack distributions like Red Hat. So one of the reasons we’re coming in at the price point we are is because we want to make it zero friction for our customer when they do that comparison of OpenStack distro A versus OpenStack distro B, at every level of comparison.

But, that said, almost everything we do is through a larger enterprise relationship. Typically when an enterprise is buying from HP they’re not making a singular decision for one piece of software or one server order or one set of services. So we talk about the big picture, what our cloud platform can do, how we indemnify our distribution of OpenStack, product capabilities, pricing, the whole thing.

This is really hard when you have a business model that is pegged to one thing like software because you end up between a rock and a hard place because you can’t easily discount below your margin line because it’s very difficult to make that up. Microsoft has a little bit more flexibility because they have such a breadth of software and they have such a breadth of offerings. For Red Hat and VMware it’s a little different because they are bound to their business model, so they have some very hard floors and ceilings in terms of what flexibility they have.

You recently acquired Eucalyptus which doesn’t have big OpenStack roots. They’re mostly about AWS integration. How do you see that fitting in?

Eucalyptus was really two things for us. It was a good collection of people who know how to build cloud software, and it was the AWS interoperability piece. I keep talking about choices, and we realize the design pattern of AWS is hugely relevant. So we needed the ability to tell customers, if you have or are interested in that design pattern, we have a way to support that.

So where we typically see the Eucalyptus demand is where a customer wants to have the ability to move an app out of AWS back to a private or managed cloud environment, or where someone says, I don’t know what’s going to happen yet in terms of going to the public cloud so I’m going to first build my private cloud apps with Eucalyptus and the AWS design pattern (basically meaning using the EC2 APIs, the S3 APIs, etc.), and building it in a way that gives me the flexibility to locate the work where I want.

What should we look for this coming year?

You’ll see us continue to build out our Helion distro of OpenStack and our Helion development platform, so you’ll see new services, new capabilities, that kind of thing. You’ll see us do a lot in the telco/service provider/NFV space.

And later in the year you’ll hear us talk a lot about a new model for enterprises that want to consume managed cloud services but don’t want to buy anything physical, don’t want to own anything anymore, that just want to consume, but in a way that matches their business realities today. We’ll be doing a lot in that space. I’m a believer that the cloud industry we have today is going to look very different in the future as the enterprise really starts adopting cloud technologies - and then all cloud vendors will shape their strategies to fit what enterprises want. So we’re trying to skate to where the puck will be and start to invent some of those new models.

You mentioned that analysts say only 10% of enterprise needs are supported by the cloud today. What’s the timeframe for change?

That’s the multi-trillion dollar question, isn’t it? But I see two enterprise patterns happening right now and this may inform the answer. One is the linear step. I’m going to move from virtualization to private cloud infrastructure-as-a-service, then I’ll try out some of this PaaS stuff to see how that really makes sense. Then I’ll see if I can run that across multiple data centers and then maybe see if a public cloud thing makes sense. So it’s kind of a linear mode.

The other pattern I hear, and this is the riskier one, is where the CIO says any new app inside my enterprise will be built to platform-as-a-service and can have zero knowledge of an operating system underneath it. What they’re trying to do is say, let’s start building in the new cloud-native model so we don’t have to worry about migrations and lift and shift and all of that.

But then there’s another question, and that is, which platform-as-a-service? At some point you’re binding to something, you’re making some commitment to some API somewhere. It may not be at the operating system level anymore. It may be higher up the stack in the middleware.

Then frequently we see customers say, we won’t move our existing resources to a cloud model. We’ll build the next project or the next deployment in a true cloud model. We’ll build that as a stand-alone system and then try to bridge back, usually through management tools, to the old. That is very common as well.


Copyright © 2014 IDG Communications, Inc.

1 2 Page 2
Page 2 of 2
The 10 most powerful companies in enterprise networking 2022