Paul Krill
Editor at Large

OpenStack execs: Red Hat, Yahoo, Comcast are our adopters — and contributors

analysis
Jul 25, 20148 mins

In an interview at OSCON, OpenStack officials said upcoming releases will focus on high-level services

OpenStack, the open source cloud computing platform for both public and private clouds, has been generating a lot of buzz in the IT world, garnering supporters ranging from AT&T, Rackspace, and the Linux Foundation to IBM, Red Hat, Oracle, and Yahoo. But it has been viewed as lacking critical capabilities in networking and usability.

InfoWorld Editor-at-Large Paul Krill met up with Jonathan Bryce, executive director of the foundation, and Mark Collier, the foundation’s COO, at the recent O’Reilly Open Source Convention in Portland, Ore., to talk about where OpenStack stands and its future.

InfoWorld: What’s the progress of OpenStack? Where is it headed?

Collier: We’re hitting this point where there’s a critical mass of users. Although of course we had a lot of active developers and that continues to grow, we’re seeing the users getting very engaged. Companies like Disney, Bloomberg, Comcast, eBay, and PayPal, they’re all getting very engaged in defining the software that they’re running, so they’re all utilizing it to move their businesses faster, and it’s an overall trend in every company [that] software is really becoming more strategic, and they’re turning to OpenStack as the tool. At the same time that they’re deploying it, they’re helping influence the future. We have a new version every six months, and users are helping us define what features they need, and in some cases they’re even writing some of the code.

InfoWorld: Are you pleased with the support you’ve gotten so far from companies like Red Hat?

Collier: Absolutely. We’ve not only seen the biggest, the most innovative companies backing OpenStack, they’re really putting their money where their mouth is. You asked about Red Hat. They employ dozens and dozens of developers whose full-time job it is to contribute to OpenStack. The way we’re organized as a foundation, we help coordinate the management of the project, but we don’t employ the majority of the developers. Most of them work for companies like Red Hat, IBM, HP, and Rackspace.

Bryce: Every six months, we do a release, and we see who contributed to the release. In the last release, a number of the top 20 contributors were users like Yahoo and Comcast. They’re not selling OpenStack services, they’re not a vendor that’s trying to make money selling something about OpenStack. They are [users who are] benefitting from the software and then turning around and investing back into it. That’s the ultimate power of an open source model.

InfoWorld: One of our writers wrote: “The core issue of OpenStack is the lack of some fundamental features that an IaaS cloud requires, including better networking, which was the buzz at OpenStack Summit this year. However, that’s not all that’s needed. Although less-discussed, OpenStack needs some core cloud infrastructure features around stability and usability.” How would you respond to that?

Bryce: There are different ways to consume OpenStack. There’s the upstream, and there’s the downstream. The upstream is what the community works on and releases every six months. Then a variety of companies take that code and package it up into different products and services. Some companies that are large and have a lot of technical staff, they will go straight to the upstream code and build their cloud using that, and then they do some extra work to tie it into whatever systems they have. For things like manageability, installers, different monitoring capabilities, those kinds of things, the hooks are all in the upstream code, but the downstream vendors like Red Hat or Mirantis, they will build products that go after a specific feature set or a specific vertical to fill in those gaps. It’s pretty similar to other open source models.

[The comments] mentioned networking specifically. The networking in OpenStack is actually extremely strong. [What] he’s probably referring to is that there was an initial networking model that OpenStack had in the beginning, and there’s a next-generation model that’s been worked on for a couple of years. Those are both capabilities of OpenStack, and users get to choose which one they want. The Neutron project [provides] the latest and greatest of software-defined networking and network function virtualization, and those are two very advanced trends. Not everybody is ready for that yet. If you need a more traditional networking model, that’s built into the system as well. But I think that’s confusing sometimes to people because they go, “Well, why are there two ways of doing it?” It’s because there’s a massive shift going on in networking right now.

InfoWorld: What gaps still need to be filled in OpenStack?

Bryce: It’s not about the core infrastructure services. If you look at enterprises and if you look at companies in every industry, they are relying more than ever on software and data analysis to drive business decisions. What people are really starting to work on in the OpenStack ecosystem are higher-level services. Savanna is a big data analytics processing service. It’s like Elastic MapReduce as a service on top of OpenStack infrastructure. There’s a relational database as a service, Trove. There are these higher-level services that are built on top of the core compute storage and networking, and that’s where I think you see it going and where there are gaps in enterprises where they need to go fill these capabilities to make their business decisions.

InfoWorld: What are going to be the highlights of the upcoming Juno and Kilo releases of OpenStack?

Bryce: One of the things that is a pretty massive update on the storage side — it’s actually available now from the object storage upstream, but it will be in the Juno release — is storage policies. That’s a way to set rules and service levels inside of your storage environment so that you can meet performance requirements, whether it’s speed or price or redundancy. It’s a pretty powerful feature that takes the reality of what most data centers look like, where they have solid-state drives, they have SAN, they have spinning disks, they have a lot of different kinds of storage. And they are having to manage those independently, and it now gives you a system to manage all of this holistically and at a software level.

InfoWorld: What about the Kilo release?

Bryce: One the things that’s been an incubated project inside of the OpenStack ecosystem for a while is something called Ironic. And a lot of people think about cloud in terms the next step of virtualization. Hypervisors have been one of the key components of the well-known clouds, but one of the bigger business reasons why we’ve seen people move off of the big public cloud providers and into internal infrastructure is for performance, where they need guaranteed performance. The best way to get that is by going directly to the hardware, to the bare metal. Ironic allows you to manage bare metal in the way that you manage virtual machines and, again, to do it through OpenStack. You can have an environment where you’re provisioning virtual machines for some workloads and you’re talking directly to hardware servers for other workloads as the need dictates.

InfoWorld: Do you have numbers on how many user sites, whatever, are deploying OpenStack technology?

Bryce: It’s really hard to get a number because it’s freely available software, it gets distributed through dozens of different commercial products and services. We have known deployments in about 150 cities around the world and they range in size from very small environments with a few dozen servers to large environments, like the one that CERN (European Organization of Nuclear Research) has where they are studying the fundamental nature of matter.

InfoWorld: These different users — Disney, Wells Fargo — what really drives them to OpenStack?

Collier: It’s really the need to automate their infrastructure. Every company is increasingly building different applications, so they have developers that want to quickly get resources so they can build an application, basically create value for their company for their customers, and they want to do it faster and cheaper. OpenStack is a really appealing way to do that because it’s all about automation. Taking something that would have taken weeks before to provision some servers for their developers to get going with their project and they can do it in minutes and they can use a self-service dashboard and push that out all the way to the developers inside the company.

This story, “OpenStack execs: Red Hat, Yahoo, Comcast are our adopters — and contributors,” was originally published at InfoWorld.com. Get the first word on what the important tech news really means with the InfoWorld Tech Watch blog. For the latest developments in business technology news, follow InfoWorld.com on Twitter.

Paul Krill

Paul Krill is editor at large at InfoWorld. Paul has been covering computer technology as a news and feature reporter for more than 35 years, including 30 years at InfoWorld. He has specialized in coverage of software development tools and technologies since the 1990s, and he continues to lead InfoWorld’s news coverage of software development platforms including Java and .NET and programming languages including JavaScript, TypeScript, PHP, Python, Ruby, Rust, and Go. Long trusted as a reporter who prioritizes accuracy, integrity, and the best interests of readers, Paul is sought out by technology companies and industry organizations who want to reach InfoWorld’s audience of software developers and other information technology professionals. Paul has won a “Best Technology News Coverage” award from IDG.

More from this author