Learn More
• A decoupled MapReduce computing-storage system for cloud computing is proposed. • Data perception mechanism after decoupling makes tasks read the closest data. • We design a VM placement strategy to exploit the data locality of tasks. • A load-aware data placement strategy is complementary to the VM placement. a b s t r a c t MapReduce as a service enjoys(More)
In recent years, there are various sources that generates data in petabyte and terabyte, known as big data and its generated by human, machine, sensor etc. So the solution leds with big data is Apache Hadoop has attracted strong attention because of its applicability on processing for the large data sets. This paper present the review about the big data and(More)
  • 1