SOLVING OPTIMALITY AND STORAGE CONSTRAINT PROBLEMS IN DATA STAGING CLOUDS

Abstract

Recently the studies on cloud computing remarkably increased. With data accessibility demands to the cloud computing, data availability improvement seems to be a challenging problem to maintain high-fidelity and time-bounded service expectations in clouds. Increasing in number of cloud data users, the problem of producing the requested data, available to the users becomes an important issue for Cloud service Providers to guarantee high-quality services. A particularly engaging an approach to improve such data availability is to stage the requested data to some other sites and cache the data for a period of time. The staging is a process of storing data at intermediate are between source and destination. Scenario is a mobile user who may want a service to advise probable alternative way to the destination depending on current traffic patterns. This service requires not only building the queried data   available at specific nodes along the current path, but also staging the data on key places along with the paths based on navigation history.  This project is an implementation of simulations of data staging problem by leveraging the genetic algorithms techniques to optimally migrate, duplicate, and cache the common data items in clouds with or without some practical resource constraints in an efficient way while reducing the monetary cost for sending and caching the data items. Monetary cost is primary interest as on one hand, it is a very useful concept to reflect the individuality of various network features such as network bit rate, link delay, and storage usage, and the provision of the resources in cloud systems are usually based on payas-you-go fashion, and thus effective use of the platforms within budget constraint is always the user’s concern.