Chief Of The Year: Amazon CTO Werner Vogels

Amazon's external-facing CTO is helping to devise a cloud computing architecture with customer requirements built-in.

By John Foley, InformationWeek
Dec. 20, 2008
URL: http://www.informationweek.com/story/showArticle.jhtml?articleID=212501217

Werner Vogels, CTO, Amazon -- InformationWeek's Chief of the Year for 2008 -- Photo by Brian Smale
Werner Vogels
CTO, Amazon

InformationWeek's
Chief of the Year
for 2008

Photo by Brian Smale
From an eighth-floor conference room at Amazon.com's headquarters on Beacon Hill, with an expansive view of downtown Seattle to the north, Werner Vogels contemplates how large companies might use cloud computing. He points to system automation and "autoscaling" beyond what IT departments have implemented internally. He envisions "partner clouds" of shared IT resources, data, and applications. He even suggests FedEx could one day run its vital package-tracking application in the cloud.

"For most of these things, it's really Day 1," says Amazon's bearish CTO--he's 6 feet, 5 inches tall--dressed, as usual, in gray and black.

If we're on the cusp of the computer industry's next major architecture, the one beyond client-server, Vogels has played a key role in getting us to this point. A former researcher in Cornell University's computer science department, where he specialized in large-scale distributed systems, Vogels joined Amazon in 2004 to help the e-retailer design and scale its IT infrastructure to handle workloads many times Amazon's own. "Amazon was reaching a point where they needed to rethink how they were building their software--what scale really meant, what reliability really meant," he says.

InformationWeek Reports

The overhaul was required to support the emerging Amazon Web Services business, which over the past four years has come to represent the state of the art in on-demand, pay-as-you-go computing. With a user ID and a credit card, application developers and others can provision servers, storage, and other IT infrastructure with unprecedented ease.

Vogels' name and face are often associated with Amazon's cloud, but AWS isn't a one-man show. Senior VP Andy Jassy conceived the business model five years ago and has had his hand at the wheel ever since. VP Charlie Bell is the lead technical manager of AWS. VP Adam Selipsky is the liaison to the 440,000 developers who have signed up so far. Vogels, Bell, and Selipsky report to Jassy, and Amazon balked at our suggestion that one of them could be singled out. But we did it anyway, selecting Vogels as InformationWeek's Chief of the Year, our highest editorial honor. Here's why.

Amazon's 50-year-old CTO has emerged as the right person at the right time and place to guide cloud computing--until now, an emerging technology for early adopters--into the mainstream. He not only understands how to architect a global computing cloud consisting of tens of thousands of servers, but also how to engage CTOs, CIOs, and other professionals at customer companies in a discussion of how that architecture could potentially change the way they approach IT.

Amazon aspires to be "the earth's most customer-centric company." Vogels' job description as "an external-facing technologist" isn't just consistent with that mission statement, it's cutting edge. Too many CIOs, CTOs, and IT organizations as a whole remain internally focused. They get treated like a cost center because they are a cost center--they don't venture outside their organizations. Vogels, in contrast, is constantly on the road talking with customers about what Amazon can do to address their computing needs in its data centers, then reporting back to the rest of the AWS team with ideas on how to make that happen. If you think Vogels' situation is different because he works for a "vendor," think again. CIOs and CTOs in a variety of industries must get better at articulating the business value of their technology to customers. Vogels even talks about Amazon building a "customer-oriented architecture."

DIG DEEPER
EXEC INSIGHTS
For more on the strategies of leading IT organizations and the execs calling the shots, download our annual InformationWeek 500 report, free for a limited time
>> See all our Reports <<
"Everybody has a lot of questions," says Thorsten von Eicken, CTO of RightScale, a startup with a management platform for AWS and other cloud services, and a former colleague of Vogels at Cornell. Vogels, he says, can both "explain the vision" of cloud computing and dive into the technical minutia.

If all goes as planned, Amazon's cloud will serve as an extension of corporate data centers for new applications and overflow capacity, so-called cloud bursting. Over time, Amazon will then take on more and more of the IT workload from businesses that see value in the model. Customer-centric? What Amazon's doing goes beyond that. Amazon's cloud becomes their cloud; its CTO, their CTO.
EVOLUTION OF AWS
Standing in the lobby of Amazon's headquarters, a 1930s Art Deco building that the $14.8 billion-a-year company has outgrown, I see Amazon's customer focus in action. The receptionist, a burly guy with tattooed arms, is on the phone for five minutes, patiently helping an author who called fretting over inaccuracies in the description of his book on Amazon's site. Problem solved.

My first meeting was with Jassy, a soft-spoken New York transplant and sports nut with seven TVs in his basement, who served as technical assistant to CEO Jeff Bezos in 2002 and 2003 before launching Amazon's Web services business. As Jassy explains it, the seed for AWS was planted in 1999 when the company decided to "decouple" system components to support its strategy of letting other e-commerce companies integrate the features of Amazon with their own sites. "We found religion around SOA, and it changed our thinking about systems and components," he says.

As the number of third-party Web sites tapping into Amazon's functionality to sell products approached a million, the APIs through which it exposed its underlying technology as services grew ever more important. In 2003, Bezos asked Jassy to develop a business plan. The next year, Amazon introduced its Simple Queue Service, a hosted queue for buffering messages between distributed applications, into beta testing.

It was the first of many Amazon infrastructure components to be offered as services. In 2006, Amazon introduced Simple Storage Service (S3) and Elastic Compute Cloud (EC2), foundational pieces for AWS users. Over the past 12 months, it has stepped up the rollout, letting customers launch Red Hat Linux, OpenSolaris, and Windows Server operating systems and MySQL, Oracle 11g, and SQL Server databases as virtual "machine images" in its cloud. Other recently introduced capabilities include persistent storage, remappable IP addresses, user-selected geographic zones for EC2 instances, fulfillment services, development tools for integrating with Facebook and Salesforce.com, and a content delivery network called CloudFront (see timeline, "Amazon Web Services", below).

Jassy, like everyone else at Amazon, is evasive on certain questions about AWS. He won't say how many data centers the company operates, where they're located, how much revenue AWS pulls in, or how fast it's growing. ("They're extremely secretive--paranoid," says one business partner.) But Jassy will say this much: "Much to our surprise, enterprise adoption is happening much faster than any of us anticipated."

Amazon offers a few metrics as proof points. There are 440,000 registered AWS developers, 29 billion objects stored in S3, and, earlier this year, the amount of network bandwidth consumed by AWS surpassed that required for Amazon's retail site.

In terms of revenue, Amazon lumps AWS sales into its nonretail "other" category, along with its Enterprise Solutions Web hosting business (customers include Target and Marks & Spencer) and co-branded credit cards. In the first nine months of 2008, revenue from "other" operations totaled $367 million. That's only 3% of Amazon's $12.5 billion in total sales over that period, but here's the cause for optimism: While Amazon's overall business grew 36% through the first nine months of 2008, revenue from those other operations climbed 45%. It's reasonable to surmise that AWS is fueling that growth.

Jassy describes cloud computing as the undifferentiated heavy lifting of IT infrastructure--or the "muck"--and says businesses stand to benefit by off-loading that task of keeping it all running. He points to the 70-30 trap (sometimes characterized as 80-20), where 70% of IT budgets are spent on maintenance and only 30% on innovation. "One of the mantras of the business is to flip that on its head," he says.

Amazon Web Services
2004 Simple Queue Service, a hosted queue for storing messages is introduced
2005 Mechanical Turk, an online marketplace for on-demand workers, debuts
2006 Simple Storage Service (S3) provides on-demand access to Amazon's data store; Elastic Compute Cloud (EC2) goes into testing;
2007 DevPay, SimpleDB, and Flexible Payments services go into beta testing; Red Hat Enterprise Linux becomes available on EC2
2008 Elastic Block Store, persistent data storage for EC2, and CloudFront CDN introduced

Elastic IP Addresses and Availability Zones give users control over IP addresses and geographic deployment of EC2 instances, respectively

AWS Premium Support launched

Microsoft, Oracle, and Sun server software become on-demand options on EC2
CHIEF CLOUD OFFICER
Vogels' assignment is to serve, in effect, as a chief cloud officer for the computer industry. A high-mileage schedule has him on the road much of the time. Earlier this month, he flew to San Francisco to attend a CTO roundtable on cloud computing hosted by the Association of Computing Machinery (ACM), and, from there, to London, Amsterdam, and Paris, where he spoke at LeWeb, an industry shindig sponsored by Google, Microsoft, and a dozen other tech companies.

On Dec. 10 at LeWeb, Vogels announced the expansion of Amazon's servers-by-the-hour EC2 service to Europe. A regional cloud minimizes latency and improves redundancy and, equally important, lets companies comply with European regulations requiring that customer data be stored locally. "This addresses the requests of many of our European customers and from companies that want to run instances closer to European customers," Vogels wrote in his All Things Distributed blog.

Vogels started his blog in 2001 while at Cornell, and he continues to post regularly, offering a tech guru's point of view on AWS services and features and an occasional deep dive into the guts of the Amazon infrastructure, such as his treatise on internally developed technology for high-availability storage. The technology, called Dynamo, integrates a bunch of arcane distributed system techniques such as DHTs, consistent hashing, versioning, vector clocks, quorum, and anti-entropy-based recovery. "As far as I know, Dynamo is the first production system to use the synthesis of all of these techniques, and there are quite a few lessons learned from doing so," Vogels wrote. Dynamo isn't exposed as a Web service, though it underlies S3, he added.

Vogels, a native of the Netherlands, developed his technical chops at Vrije University in Amsterdam, where he earned a Ph.D. in computer science. At Cornell, his research focused on large-scale distributed systems, including networking protocols, middleware, and cluster management. "In my eyes, Amazon is probably the world's largest distributed system," he says.

Vogels' academic and research background gives him added authority when talking up cloud computing. Oh, sure, he wears a marketing hat when on stage at industry events, but underneath that hat is the mind of a technologist who hears "cloud" and thinks of loosely coupled software services on a global distributed network operating with near-perfect availability.

Vogels Vitae
Joined Amazon as director of systems research in 2004; promoted to CTO and VP the next year

Computer science researcher at Cornell University from 1994 to 2004, specializing in large-scale distributed systems

Research engineer with INESC, a computer science research institute, from 1991 to 1994

Doctorate in computer science from Vrije University in Amsterdam

Writes All Things Distributed blog and uses Facebook, FriendFeed, LinkedIn, and Twitter
INTERNAL INNOVATION
Within Amazon, a dedicated development team works on each of the major Web services--EC2, S3, SQS, SimpleDB, CloudFront. It's an organizational structure intended to give those teams maximum autonomy. "We believe the groups should own their customers down to the metal," says Bell.

A 10-year company veteran, Bell knows Amazon's IT infrastructure as well as anyone. I first interviewed him eight years ago when Amazon was on a shopping spree for commercial software, licensing products from Excelon, Oracle, Manugistics, SAS Institute, and others. At the time, Bell made the point that Amazon's differentiating features--its personalization and customer-contact capabilities, for example--were internally developed. "You could characterize us almost as a software company," he said.

That's even truer today than it was then. As the demands on Amazon's IT infrastructure have increased through the success of its retail, Web hosting, and AWS businesses, Amazon is building more software and buying less. "There's hardly any third-party software left," says Vogels. "If I could buy things, I would. But vendors haven't reached a point where they can deliver software that can reliably work at Amazon scale."

(Also in that Nov. 6, 2000, InformationWeek story, titled "Amazon's IT Agenda," was a comment by Amazon chief programmer Jeremiah Wilton that foreshadowed the hiring of Vogels a few years later. "Our greatest challenge, from my point of view, has been scaling our systems," Wilton said.)

Bell is responsible for making sure that Amazon's Web services work together in a coordinated fashion, as well as for the business side of AWS. The general managers of each Web service--including Peter De Santis, general manager of EC2, and Alyssa Henry, general manager of S3--report to Bell. Their teams are spinning out new services and features every week or two, and increasingly those capabilities--user-designated "availability zones," block storage, EC2 in Europe, the CloudFront content delivery network--are geared for large companies.

That pace will continue into 2009, with EC2 monitoring, a Web-based management console, load balancing, and autoscaling due soon. Amazon's ability to propagate Web services goes back to architectural decisions made before the term cloud computing was first uttered. Bell puts it this way: "If you're going to build a service in the company, you're going to build it as a Web service, and you're going to build it to be exposed."

In its boilerplate, Amazon says 440,000 developers have signed up for AWS. In fact, they're not all developers; they range from entrepreneurs to enterprise IT pros and everyone in between. Selipsky's team supports that burgeoning population and plays to its diversity. "We want to support any programming language that our customers want--Java, C#, Python, Ruby, PHP," he says.

Charlie Bell

'Werner provides a great translation of what we're doing to the outside world. He's a great communicator,' says Amazon veteran Bell, who oversees Amazon Web Services development.
AWS is a popular platform among startups, Web companies, and software-as-a-service companies. Increasingly, Amazon's customers are household names: Nasdaq, The New York Times, Philips, SanDisk. Eli Lilly is using EC2 to deploy SQL Server/Windows Server instances as needed for research data. The Indianapolis Motor Speedway uses AWS for Web site mirroring, video streaming, and digital image archiving. Selipsky says there are a "slew" of other enterprise customers, including a hedge fund company and a mutual fund company, that haven't been disclosed.

As the early examples demonstrate, enterprise adoption tends to be limited to specific applications; CIOs don't want to commit too much, too soon. Vogels says we've only scratched the surface. When I ask whether FedEx, one of the largest and most sophisticated IT users, could seriously consider AWS as an alternative to running mission-critical apps in its own data centers, he doesn't hesitate. "I would love them to do that," he says. "I think that would be a great benefit to them."

The world of research, Vogels' old stomping ground, is another target market. This month, Amazon introduced AWS Public Data Sets, a no-cost repository for government, research, and other public data. And Amazon and Harvard Medical School co-sponsored a daylong forum in Boston where doctors, biomedical researchers, and other experts discussed how to apply cloud computing. Harvard's Laboratory for Personalized Medicine already uses Oracle on EC2 for genetic testing models and simulations.
SPREADING THE WORD
After spending half a day at Amazon's headquarters, I still had questions for Vogels, but, traveling somewhere in Europe, he had gone silent. Nevertheless, it's easy enough to follow Vogels. He's all over the Web--blogging, Twittering, and talking up cloud computing in interviews and on video.

With his Dutch accent, Vogels comes across as a bit of a Renaissance man--a music buff, photographer, and patron of the arts who speaks four languages and rides a motorcycle. His Twitter posts alternate between references to server optimization and network latency and an appreciation for "marvelous absurdistic physical theater."

I click on a video replay of Kara Swisher's on-stage chat with Vogels at LeWeb. Amazon's CTO tells the audience that the conversation with customers has shifted from cloud security to application deployment. Customers want to know how to move existing applications into the cloud, how much engineering work is involved, and how to exploit the functionality that Amazon makes available, he says.

Vogels, surprisingly, tells Swisher that security isn't the big issue it once was with customers, but he admits that AWS--which has experienced a number of hours-long service outages this year--has work to do in the area of reliability. "There's no excuse for any downtime or failure," he says. "One hundred percent availability is the only goal that you can have."

It's a reminder that, even though Amazon is four years into cloud computing, it's only recently begun addressing the stringent requirements of enterprise customers, and that work is unfinished. Amazon's flagship EC2 service, in beta testing for two years, became generally available just two months ago. Its SimpleDB relational database service is in beta testing. EC2 monitoring, management, load balancing, and autoscaling are still on the road map.

As my deadline approaches, Vogels resurfaces. I ask him about lingering concerns over data governance, security, and reliability in the cloud. He responds via e-mail that Amazon works with customers to address these concerns, but that "no customer has exactly the same needs as another."

In other words, with Bell driving AWS development and Selipsky managing front-line relationships, Vogels will need to keep racking up the frequent-flier miles. As Amazon reaches out to customers, his ability to articulate the benefits of cloud computing will have a lot do with whether they return the embrace. Years of engineering and cloud development are behind Amazon; the bigger job lies ahead.