Heuristic data placement and replication for scientific workflow in cloud computing
Vishali, Parminder Singh, Avinash Kaur, Manpreet Singh · 2017
Cloud computing has turned up as an emerging platform for individual or personal computing. In order to improve overall performance of cloud data placement is an important task. Data placement is a prime issue which aims at minimizing the cost of inter node transfers of data in the cloud, the performance of the entire cloud system get improved by eradicating this issue. Many authors have proposed different techniques for optimizing the data placement strategy in scientific workflow. The strategy used for one application may not used for another application. The appropriate data placement strategy reduces the scheduling overhead, the cost of data processing, availability of data becomes high, bandwidth will be less consumed, scalability will be improved and fault tolerance will be increased. It is impossible to satisfy all the conditions to place the datasets at appropriate position where all tasks can access data with the minimum data transfer cost and fulfillment of SLA.