Method for determining whether to dynamically replicate data

作者: David Vengerov , George Porter

DOI:

关键词: Real-time computingData segmentComputer clusterWorkloadReplicateNode (networking)Replication (computing)Computer science

摘要: The disclosed embodiments provide a system that determines whether to dynamically replicate data segments on node in computing cluster stores collection of segments. During operation, the identifies segment from is predicted be frequently accessed by future tasks executing cluster. then slowdown would result for current workload if were replicated node. also benefit associated with replicating If less than benefit, replication replicates

参考文章(7)
Berthold Reinwald, Mir Hamid Pirahesh, Saileshwar Krishnamurthy, Mehmet Altinel, Chandrasekaran Mohan, Christof Bomhoevd, System and method for adaptive database caching ,(2008)
Bruce D. Holenstein, Gary E. Strickler, Paul J. Holenstein, John R. Hoffmann, Mixed mode synchronous and asynchronous replication system ,(2009)
Raghu Krishnamurthy, Deepak Tawri, Abhay Kumar Singh, Gopal Sharma, Techniques for non-disruptive transitioning of CDP/R services ,(2007)
Apratim Purakayastha, Norman Howard Cohen, Stefan Hepper, Thomas Schaeck, Veronique Perret, System and method for using portals by mobile devices in a disconnected mode ,(2003)