Front Inner Page - Volume 1 No.3 December 2014

  • » Back to Index

  • Title

    :

    A Novel Optimization Framework for Dynamic Map Reduce in Distributed Networks

    Authors

    :

    V.Lijina1, M.Jebeen2

    Keywords

    :

    -

    Issue Date

    :

    December – 2014

    Abstract

    :

    Dynamic Map Reduce is a current computing standard for data processing that overcomes the problem of un-improved resource allocation in pseudo distributed environment. Map Task or Reduce Task uses detection Algorithms to identify the resources which is not utilized and uses that particular resource in efficiency by shifting resources with different clusters. We have three major steps to implement Dynamic Map Reduce. Self-motive Slot Allocation it is used to overcome the slot allocation restraint using Decision Tree Approach. Execution Performance balancing it is to steadiness the performance in jobs using optimized Load Balancing Mechanism. Self-Adaptive Map Reduce with decision tree for pre scheduling it is to advance the data to load in idle slot. Finally it improves the performance of Map Reduce workloads. To efficiently manage the limited physical resources could improve the usage of both CPU and disk I/O resources under heterogeneous workloads. I proposed three provisioning polices for dynamic map reduce also prove my system through experiments on a real multi cluster.

    Page(s)

    :

    1-5

    ISSN

    :

    2347- 4734

    Source

    :

    Vol. 1, No.3, December 2014

    Download

    :


  • » Back index