IN THE CLOUDS

Cloud computing will handle more of the heavy computational lifting called for by the scientific and engineering sectors, say researchers at Argonne National Laboratory in Illinois.

Definitions vary, but cloud computing is essentially a form of distributed computing that offers the capability to tap into a vast network of computing resources through the Internet, said Kate Keahey. She heads the cloud-computing infrastructure called Nimbus, which recently went online at Argonne.

When computer users are said to be operating "in the cloud" they're actually running simulations and other computer work on customcrafted virtual computers that exist on distributed computing platforms, Keahey said.

Researchers who sought to analyze traffic patterns on the nation's highways, for example, could upload their data in the cloud and then have multiple computers crunch the data. The computers would then present the results back to the researchers as if a single machine had completed the work, Keahey said.

Though cloud computing evolved from grid computing, it does differ from that computational method, she added. With grid computing, researchers submit the work they want computed to a batch scheduler, which puts the job in a queue for a specific set of computing resources - for ple a supercomputer - to work on.

"This means you have no control over when your job might execute," Keahey said. "You're pretty much at the mercy of how that particular grid asset is set up. If its configuration doesn't quite match the complexities of your job, fixing the problem may get very complicated."

Many cloud- computing platforms tell users ahead of time how much computing capacity is available from the cloud so the work can be done faster, she said. Users can also configure a virtual machine within the cloud to meet the particulars of the jobs they're trying to accomplish, Keahey said.

After users have configured the type of virtual machine they need for their work they can go to different cloud computing providers and recreate the system they need to get their jobs done, she added.

Keahey and her team developed the Nimbus open-source cloud-computing infrastructure to allow scientists working data-intensive research projects to use such virtual machines with a cloud provider.

YOUR COMPUTING INFO HUB

MEMAGAZINE.ORG