Arik Hesseldahl

Recent Posts by Arik Hesseldahl

Seven Questions for Rackspace CTO John Engates

In a completely rational world, Rackspace would no longer be an independent company. The Web host that has in recent years veered into the cloud services business has held steady to its independent streak as several of its rivals have been rolled up by larger companies. In fact, assumptions that Rackspace would be next have spurred much of the growth in the company’s stock price, which has in recent weeks been trading at historic highs, and the shares are up more than 35 percent just since January.

No, it’s not an entirely rational company, but rather a “fanatical” one. That’s the word the company uses in its marketing messaging to describe its customer service, but it wouldn’t be inaccurate to use the same word to describe the dedication of CEO Lanham Napier and his team to remain independent and take full advantage of the cloud computing opportunity that lies ahead. From that you could conclude that they see a big opportunity and you’d be right. Rackspace executives are now accustomed to questions from persistent reporters like myself pressing them on the subject of being acquired. The answer, now as before, is still no.

In fact, Rackspace sees the cloud opportunity as being so big that it has made some acquisitions of its own to exploit it. In February, it grabbed Anso Labs in a bit of an “acqhire” deal that brought the team that built NASA’s cloud computing system to Rackspace. That move was part of a wider strategy to promote Openstack, an open source cloud computing platform that Rackspace has spearheaded with partners including Dell, Cisco Systems Citrix, and with a little help from — get this — Microsoft.

These were among the many things I talked over with Rackspace CTO John Engates when he stopped by my office in New York recently.

AllThingsD: So for those who haven’t done their homework, what is Openstack and what is Rackspace’s place in it?

Engates: When Rackspace got into the cloud business, we had to build everything ourselves. The hardware was there but to make a cloud you have to have software. And that software didn’t exist as a commercial product. We went to work building our own software and acquired some companies to help us build it. We looked long and hard at what we could use to build an enterprise scale cloud, and at that point it was all lab scale, you could build a small lab-sized project but you couldn’t run a business on it let alone offer as a service. Amazon had already built their service, and we knew that Google was already a cloud and that Microsoft was working on the Azure platform. We just had to build something ourselves because it wasn’t going to be handed to us on a platter, so since 2007 we’ve been building our own software and in 2010, we asked ourselves whether we wanted to continue building our own software long term or if we just wanted to make sure we had access to the caliber that everyone else was going to have.

So the strength in the cloud is really about software?

I’ll go ahead and say it: There’s nothing really unique about one vendor’s servers or storage networks today that one vendor has a real advantage over the other when it comes to the cloud. But the software that makes the cloud the cloud is the real difference. That’s why Dell and Citrix and Cisco have all joined on to the OpenStack project, even Microsoft is contributing code. I know that sounds weird, but the reason that Microsoft is interested is because Openstack is hypervisor-agnostic. You can run Xen or KVM or Microsoft’s Hyper-V or even VMWare.

Is Openstack really the basis of your cloud offering or will you support other cloud platforms?

We will. It is today a big part of our cloud strategy. Our Cloud Files project sits 100 percent on Openstack. They’re one and the same. Our cloud compute product is being migrated to Openstack by the end of the year. But we also have a large VMWare-based business. Why not? We’re sort of technology agnostic. We’re not tied to any one vendor.

Historically, I think of Rackspace as a Web hosting company, one that was always outside of my price range as an individual. The cloud is supposed to be flexible enough that even the smallest operations can use it. Do you have that flexibility now?

We’re a host and in that business, but we don’t play in that lower tier of the market where the Mom and Pop-style operations are, but for big operations. We start at the small business and go up. We’ve always offered a 100 percent dedicated server, and you can’t offer hosting like that for $20 a month. But with the cloud, because it’s inherently multi-tenant, we can start to reach some of those smaller customers. Our smallest Linux compute instance starts at about 1.5 cents an hour, and if you do the math it works out to about $10 a month, so we can reach the really small customers now.

So tell me about the overall state of your business. Openstack is obviously a big part of your strategy. But what are you hearing from customers, and what do they need right now that they’re not getting?

I think our customers are not unlike most other cloud or hosting customers. They want things to go faster and they want it cheaper and better. A lot of things are obviously moving in that direction and the cloud helps get them there. They look at it as a way to automate, to get things online. Some people worry about the shared nature of the cloud because they worry about security. They want some better sense of security, and one of the things we’ve done is leveraged our hosting business as a way to bring together a hybrid offering. An example I like to use is Under Armour, the sports clothing company. They use some of our cloud resources and dedicated services in concert with each other. They do all their e-commerce on the dedicated site so they can meet all their security requirements but also use the cloud to be much more responsive on their site when they need to be.

Does that combination help companies get over their concerns about security? There is still an ingrained resistance to putting critical files off premises.

It helps a lot, because it’s something that they can wrap their head around. When they know they have a dedicated firewall and server. They get that. When it goes to the cloud, it’s a little bit more of a stretch because its most abstract. There are always concerns about what else might be going on on that server next door. You hear the phrases “noisy neighbors and nosey neighbors” a lot these days in relation to the cloud. You don’t want either one of them. We’re getting better at isolating customers from those, but still there are some things that belong in a dedicated environment.

So your business has been growing like crazy, and many of your competitors have been acquired by telecom players. Terremark became part of Verizon, then Saavis was acquired by Centurylink. The conventional wisdom has long been that Rackspace was going to be the next one taken out, and yet here you are still independent. What gives?

Lanham Napier, our CEO, has said many times that Rackspace doesn’t want to be acquired because we think the opportunity is huge. The opportunity is, and that’s why some of these telecom players are grasping at other companies that they think will help them get there. We’re a public company and in theory someone could certainly show up with enough cash, but sometimes the people who show up with that kind of cash do so because you’ve sent a signal that you want to be acquired, but Lanham I think has done a pretty good job of saying that we don’t want to be acquired. I think it’s because he thinks we have something special.

And what’s that?

We like to talk about fanatical customer support, and that’s more than a marketing pitch. With us, its really the real deal. I like to compare us to Zappos. They were customers, and they had sort of a funky, weird corporate culture but in a good way, and I think we’re a little like that. We’re all about having a great customer service outcome, and that’s hard to do at a telco. They’re the phone company. They’re not exactly known for great customer service outcomes.


Latest Video

View all videos »

Search »

Just as the atom bomb was the weapon that was supposed to render war obsolete, the Internet seems like capitalism’s ultimate feat of self-destructive genius, an economic doomsday device rendering it impossible for anyone to ever make a profit off anything again. It’s especially hopeless for those whose work is easily digitized and accessed free of charge.

— Author Tim Kreider on not getting paid for one’s work