From the last few decades workflow in cloud computing becoming powerful technology for complex computational tasks in the area of a distributed computing environment and cloud computing. Important requirements for Workflow in cloud computing are the allocation of suitable resources for complex computational tasks execution by considering different parameters such as reliability, efficiency, throughput, time, speed, resource utilization, and cost optimization, etc. When the amount of stored data increased in private and public clouds, data storage cost also increases to overcome with this problem different researchers classify the stored data as active and inactive for cost reduction. For this purpose, they use the hybrid filtering approach algorithm that can track the record of active and inactive clouds in the main memory clouds framework. In many cloud computing environments, data replication strategies are used in clouds to reduce data communication costs and access latency and improve data availability. An algorithm proposed for data replication in the cloud is also defined with different determinant levels through which the cost of workflow decreases significantly and communication cost for workflow application will also be reduced in the paper. As the resources in clouds change dynamically for this adaptive data placement strategy is proposed that consists of two stages which will cluster datasets dynamically each time when new datasets generated and group data sets in a data center that‟s the base is data dependency and resource availability at run time. While the workflow is running in cloud data movement can be effectively reduced with an adaptive data placement strategy using the bond energy algorithm.
Volume 12 | Issue 7
Pages: 815-822
DOI: 10.5373/JARDCS/V12I7/20202066