Select your localized edition:

Close ×

More Ways to Connect

Discover one of our 28 local entrepreneurial communities »

Be the first to know as we launch in new countries and markets around the globe.

Interested in bringing MIT Technology Review to your local market?

MIT Technology ReviewMIT Technology Review - logo

 

Unsupported browser: Your browser does not meet modern web standards. See how it scores »

A key problem for Amazon’s Elastic Compute Cloud service, critics say, is that it relies on technical wizardry that’s difficult to pull off reliably. It uses a technology called Elastic Block Store to record and retrieve data that’s being actively used by applications running on the service. It’s easy for cloud systems to store data reliably if there’s plenty of time to distribute the data across all of the machines that make up the cloud. In that case, data might not be entirely consistent at all times, but it will get that way eventually. But Elastic Block Store is built to have it both ways—to be reliable, but also to record changes throughout the system very quickly.

Steve Tuck, general manager of Joyent Cloud, says that such a balancing act is too precarious to be reliable. “You are setting yourself up for this type of [outage],” he says.

Joyent’s cloud avoids taking data that’s being actively used and distributing it over the network. Instead, it distributes that data over a number of computers locally to keep it safe and accessible, and stores less frequently used resources more distantly in the cloud.

Nasuni, a startup based in Natick, Massachusetts, also boasts a different architecture that, it says, can make cloud computing services more reliable. Nasuni believes companies will be more comfortable trusting the cloud as part of a reliable system that includes local storage. According to CEO Andres Rodriguez, a system such as Elastic Block Store can be “a ticking time bomb” within the Elastic Compute Cloud. He says that because the system needs to be responsive, it can’t safely distribute changes with enough time to guarantee that it won’t lose data.

Nasuni offers a solution that keeps active data local, and it’s careful about what it moves to the cloud. It might move backup information to the cloud, for example, but store frequently used documents locally.

Rodriguez argues that making these architectural changes won’t diminish the power of the cloud. “What almost no one understands is the capability of that massive data structure in the cloud,” he says. “It’s a commonly shared, unlimited, immutable, self-protected data structure.”

Companies such as Nasuni and Joyent hope that businesses will be more comfortable using the cloud if their active data is kept close, with cloud storage reserved for data that doesn’t need to be as consistent and immediately accessible. “We’re trying to take everything that’s revolutionary about the cloud and keep it under covers,” says Rodriguez. 

1 comment. Share your thoughts »

Tagged: Computing, cloud computing, Amazon, cloud

Reprints and Permissions | Send feedback to the editor

From the Archives

Close

Introducing MIT Technology Review Insider.

Already a Magazine subscriber?

You're automatically an Insider. It's easy to activate or upgrade your account.

Activate Your Account

Become an Insider

It's the new way to subscribe. Get even more of the tech news, research, and discoveries you crave.

Sign Up

Learn More

Find out why MIT Technology Review Insider is for you and explore your options.

Show Me