The key objective of this work is to
investigate the phenomenon of emerging in-memory analytics applications and identify
current big data application that follows in-memory architecture. In addition,
this chapter also briefly introduce some of existing problems and challenges
within traditional approach and how in-memory data processing overcome these bottlenecks
were discussed. The following sections, 1-4
dive into deeper details of the in-memory analytics and its driving forces.
Section 5 discusses NVRAM impact on the column data format, column store, and
persistency layer. Section xx Organizes existing approaches that improve the
performance of query processing in a taxonomy of categories related to the main
problem they solve and classifies existing work according to optimization goal.
Finally, Sect. 5 identifies opportunities for future work in the field.