Research On Job Scheduling Algorithm For Forest Resource Information In Cloud Computing

As time goes on the data scale of forest resource is getting larger, and the type is more various and complex. As a result, it is harder to detect and process, in addition with user’s different demand for data processing, the traditional and single calculation mode has been unable to adapt to the vast amounts of data manipulation, so the cloud computing based on large-scale computer cluster has become the main way to enhance the performance of data processing in the future. Thus it is of great significance to research on key technology of cloud computing.Due to its reliable and efficient parallel processing ability, Hadoop has become the mainstream open source clouds computing platform, which has implemented the distributed file system HDFS and the parallel data computing model MapReduce successfully. MapReduce is the core technology of Cloud Computing; it is not only a distributed programming model, but also an excellent job scheduling model. Job scheduling technology has become one of the most popular topics which directly relates to system resources usage and overall performance of the cloud computing platform.This paper analyzes the implementation process of the MapReduce programming model and job scheduling principle, which focuses on the algorithm ideas and concrete realization of the existing job scheduling algorithm, each has its advantages and disadvantages. To solve the problem that the existing algorithms rarely consider the double constraints about time limit and highest budget, the paper proposes an improved cloud computing schedule algorithm DBS (Deadline and Budget Scheduler) face to the dynamic cloud computing environments. The new algorithm can calculate job weights use the time limit and budget, allocation the minimum resource slots, and maximize throughput of the cluster in the case to meet the needs of users. It could control the allocation of cluster capacity by adjusting the job weight and the minimum slot number dynamically.To verify the validity of the algorithm, this paper implements DBS algorithm in Hadoop platform compares its performance with FIFO algorithm. The result shows that DBS can reduce the job response time, maximize the number of jobs running in the cloud cluster on the basis of satisfying the user’s time limit and budget, which achieve the fairness of user’s demand.