Since 1986 - Covering the Fastest Computers in the World and the People Who Run Them

Language Flags
July 12, 2010

New Research Helps Spot Cloud Problems Before They Arise

Nicole Hemsoth

Bottlenecks are one of the biggest banes to cloud computing, especially for users who require solid performance and predicatability. To get to the heart of the problem, researchers from North Carolina State University have devised a solution that will allow infrastructure providers to to spot and predict these problems and address them before they have an impact on performance. This is, of course, good news for end users but can also save infrastructure providers a great deal of money since when these issues occur they are often in violation of ther own SLAs–a costly mistake. This research allows for prediction of abnormalities via the crafting of models that look at system activity in a host of different contexts, thus making a profile of normal system behavior so that when something is out of synch, it is possible to see the exact root of the problem and allowing for time to correct the issue. This is what one of the researchers called a “context aware” prediction approach that leads to “an 80 percent lower rate of false alarms.”

Full story at R&D Magazine