DYNAMIC SCHEDULING ALGORITHM FOR REDUCING START TIME IN HADOOP

Map Reduce is a model associated with a programming and implementation method and is used for formulating on large datasets. The main challenge is scaling of start blocks and present implementations might end in a block of scale back tasks. In this work, In this work, a new start up model is proposed using temporal constraints and hence, the map task gives a massive output then the performance of Map Reduce reduces drastically. Through this analysis the map reduce planning mechanism is modified to reduce the waste resources in the system slot. This tends to an end within the scale back tasks waiting around the proposed model scale back the planning policy for reducing the waiting of scales back tasks and begin times within the Hadoop platform. It also decides the beginning time and purpose of every scale back task dynamically based on the context of each job, together with the task completion time and therefore the size of map as output. Thereafter, scale back completion time and system average latent period job completion time have been estimated. The experimental results illustrate that the scale back completion time has been decreased sharply due to the rise of the temporal rules and map reduce techniques.