IMPROVED REAL-TIME DATA ELASTICITY ON STREAM CLOUD

Most of the applications in cloud domains such as online data processing, Froud detection, large scale sensor network etc. where large amount of data should processed in real time. Earlier, for data stream processing, the centralized system environment was using with store and then process paradigms. After that some advancement has been introduced with distributed environment for data stream processing. Data Stream processing using novel computing paradigm which take query as input and splits that query into multiple sub queries and process the data on multiple sub clusters in such a way that reduces the distribution overheads. This kind of application generates very high input data which needs to process with the available clusters So High availability and elasticity are two key characteristics on the cloud computing services. High availability ensures that the cloud applications are sensible to failure. Elasticity is a key feature of cloud computing where availability of resources are related with the runtime demand. So in this paper we present a comprehensive framework for obtaining elasticity and scheduling technique for highly availability.