Using Object-Oriented Big Data Analytics to Reveal Server Performance Dead Zone

So far, big data analytics have proved worth by reaping fruitful achievements in business intelligence, health care and so on, which aims to reveal efficiently hidden and unique information from pre-existing large datasets. Therefore, big data analytics are sprouting in almost all areas, expecting to find new values from musty archives or continuous wave of newly generated data. This paper introduces how we utilize big data technologies to establish an object-oriented analytic architecture for IT operations, which evolves from traditional and coarse statistics into fine-grained and in-depth analysis. Moreover, this paper demonstrates applying the architecture to peek inside practical problem of the server performance dead zone. The peering process consists of applying various analysis models iteratively on large set of server logs. Our work can be considered as an attempt to exploit how object-oriented big data analytic benefits IT system operations and optimization.