OpenStack launched three years ago at the right time: Open source was going mainstream, and cloud was beginning to take off.
Within the next three to five years, OpenStack community member Rackspace believes, the majority of clouds will be built on OpenStack. But the Amazon Web Services constituency would likely say otherwise as the battle between AWS and OpenStack heats up.
Here, Jim Curry, senior VP and general manager of Rackspace's private cloud business, explains OpenStack's lofty plans to take over the cloud market, RackSpace's strategy, the new stack that has emerged and the revelation that not everything works better in a cloud infrastructure.
There was recently a debate about the API compatibility issue between OpenStack and Amazon. Care to comment?
Jim Curry: First of all, there's a misconception of this. People think there is someone who controls the ability to have an Amazon API. And the root of it is, there's no one to control that.
If someone wants an Amazon API compatibility built in, all they do is write the code and submit it. And that's in fact what has happened. There's actually a fairly robust Amazon API in the OpenStack project. We at Rackspace support it for customers. So if someone, including Randy Bias and his team, want an Amazon API, the answer is, put developers on it, write the code, submit it to the project and it will go in. I don't think there's a debate around it. A debate assumes someone is trying to block it; no one is trying to block it.
I think the debate is what OpenStack should become. And I think the idea that OpenStack at this point should cede its roadmap and its future to Amazon is the wrong debate. And I don't mean in terms of business strategy. When I started this project, we wanted to be great. We weren't here to emulate another company's strategy. Amazon is doing a great job; they built an incredible technology and an incredible business. We got into this because we wanted to build something better; we wanted to build something that was going to change the way compute affects all our lives in its availability. And that's exactly what we continue to do. I think if we were just trying to become an Amazon clone in order to serve peoples' business models, that would be a mistake for the project.
What are you working on now at Rackspace?
Curry: Cloud has gone to much more mainstream adoption level, as has open source. So we've seen the move from early adopters in this market to much more mainstream companies. We're working with not just the people with a high degree of technology, but also a lot of very mature folks who have decided they want to adopt a new architecture of cloud and run it in public and private [cloud instances].
We're also starting to work on rethinking the architecture of cloud -- beyond just public cloud and private cloud -- but also the concept of how to think about where and when to use bare metal. We're starting to think about the concept of specialized clouds and how those might play -- whether they are specialized by workloads or other characteristics.
Cloud is starting to morph beyond what Amazon launched and, frankly, what we launched, which was general-purpose computing for the masses, into something a little bit different. We are actively working with our customers to define that ... I think that we're starting to come up with some architecture that will be very unique and compelling.
What are some examples of that?
Curry: Three years ago, we were all rushing to move everyone to the cloud -- all multi-tenant clouds. We've actually found there are certain things that run better on bare metal. There's a desire to put stuff on bare metal and be able to tie it to cloud. And that's even evolved to 'What if we actually started to think about workloads running on clouds and optimizing those workloads specifically for cloud?'
For example, we bought a company called Object Rocket, which offers a Mongo service, and they have highly optimized the entire stack for Mongo. It's high performance and allows you to have a very cloud-like experience. For someone to be able to take a purpose-built Mongo cloud and tie it into their front end running in a public cloud, it's a very attractive value proposition. It's a different architecture. To be able to take that as well and plug it into some of the stuff you might be doing in a private cloud or bare metal also gets interesting. As we start to go through and build architectures, we have stopped being biased toward having the stuff in public cloud and are more focused on how to optimize all the pieces of an architecture -- multi-tenant public cloud, private clouds, bare metal -- and then [look to build] specialty clouds to solve a unique architectural problem for customers.
We're defining an architecture that will be unique to the world and look a little bit different than we assumed it would three years ago.
What cloud technologies will dominate in the next few years?
Curry: The progress [open source] has made in the last three years is tremendous. When you look at what people are building today and how they're building their environments, they're not using traditional operating systems, traditional databases or traditional queuing systems … they're being built from the ground up using open source components. And that sometimes gets married in with the concept of cloud, but they're very different.
About the author:
Michelle Boisvert is executive site editor for SearchCloudComputing, SearchWindowsServer and SearchDataCenter. Contact her at firstname.lastname@example.org.