Task shape classification and workload characterization of google cluster trace

Abstract

Understanding workload characteristics is crucial for optimizing and improving the performance of large scale data produced by different industries. In this paper, we analyse a large scale production workload trace (version 2) [1] which is recently made publicly available by Google. We discuss statistical summary of the data. Further we perform k-means clustering to identify common groups of job. Cluster analysis provides insight into the data by dividing the objects into groups (clusters) of objects, such that objects in a cluster are more similar to each other than to the objects in other clusters. This work presents a simple technique for constructing workload characteristics and also provides production insights into understanding workload performance in cluster machine.