Open Data Center Alliance tackles big data analysis
By James Niccolai
The Open Data Center Alliance, a customer group that shares tips about cloud deployments and tries to nudge vendors into supplying the products they want, has added big data to the list of IT topics it covers.
The alliance was set up in 2010 to provide a collective voice for enterprise customers, who use their buying power to try to influence the data center products vendors sell, with an emphasis on interoperability and open standards.
The alliance defines technology requirements and usage models for various aspects of cloud computing, such as identity management and data security. Its members agree to follow those usage models, giving vendors a greater incentive to provide the products and capabilities they ask for.
The ODCA has more than 300 members, including big firms like BMW, JP Morgan Chase, Lockheed Martin and Capgemini. It’s holding its annual Forecast event in San Francisco this week, where it announced the completion of three new usage models, for software defined networking, scale-out storage, and information as a service, or basically big data analysis.
Like the other usage models, they lay out the group’s main technology requirements for each area. The usage models also include wording that customers can use when they submit RFPs (requests for proposals) to vendors, and the alliance has built an online proposal engine assistant tool (PEAT) to help prepare RFPs.
“We need standards-based cloud solutions that are designed from the ground up to meet the requirements of enterprise IT organizations, and this is where the ODCA comes into play,” said Mario Mueller, vice president of IT infrastructure at BMW and chair of the Open Data Center Alliance.
Many of the cloud products offered by vendors today are incomplete, he said in an interview, and one of the alliance’s goals is to educate vendors about the features and capabilities their customers want. “If enough of us come together, they will listen,” he said.
The cloud offerings are lacking in areas like security, compliance, and how they work with existing legacy equipment, he said. And they’re often not as interoperable as vendors claim.
One of the benefits for alliance members is that they share with each other results from proof of concept tests, which help guide purchasing decisions. There’s a trade-off with sharing information that might be useful to a competitor, but it’s worthwhile to learn from experiences other members have to offer, Mueller said.
And ultimately, he said, alliance members still compete on the speed with which they can implement their cloud services, even if they share basic architectural tips.
The other advantage of sharing test results is that the alliance can take them back to vendors and call them out on any false claims they make about capabilities or interoperability. “It’s not that the vendor is just telling us stories about what they have to offer,” he said.
Still, Mueller acknowledged it’s been a challenge trying to influence how vendors build their products. That’s partly because the members who employ its usage models don’t always tell their vendor they’re an ODCA member, Mueller said, so vendors don’t know how influential the alliance is becoming.
Over time, and as more members join the group, Mueller said he’s “100 percent confident” that vendors will act on their requests. “It’s what customers want, and the vendors that succeed in the long term will be the ones that listen to their customers,” he said.
Vendors can also join the alliance, which gives them benefits like being able to see usage models as they’re developed. SAP, Infosys and UOL, a large ISP in Brazil, announced Monday that they have become members. They join Microsoft, Dell, Hewlett-Packard and dozens of others, though some big names, like IBM and Oracle, are not yet on board.
The alliance also updated several of its existing usage models Monday, including those around data security, long-distance migration, and reporting and reducing carbon footprint. The Forecast conference continues Tuesday.