Coarse-grained workload categorization in virtual environments using the Dempster-Shafer fusion

Abstract

Given a number of known reference workloads, and an unknown workload, this paper deals with the problem of finding the reference workload which is most similar to the unknown one. T he depicted scenario turns to be useful in a plethora of modern information system applications. We\ud name this problem as coarse-grained workload classification, because, instead of characterizing the unknown workload in terms of finer behaviors, such as CPU, memory, disk or network intensive patterns, we classify the whole unknown workload as one of the (possible) reference workloads. Reference workloads represent a category of workloads that are relevant in a given applicative environment. In particular, we focus our attention\ud on the classification problem described above in the special case represented by virtualized environments. Today, Virtual Machines (VMs) have become very popular because they offer important advantages to modern computing environments such as cloud computing or server farms. In virtualization frameworks, workload classification is very useful for accounting, security reasons or user profiling. Hence, our research makes more sense in such environments, and it turns to be very useful in a special context like cloud computing, which is emerging at now. In this respect, our approach consists in running several machine-learning-based classifiers of different workload models, and then deriving the best classifier produced by the Dempster-Shafer fusion, in order to magnify the accuracy of the final classification. Experimental assessment and analysis c1ealry confirm the benefits deriving from our classification framework

Machine Learning Empowered Intelligent Data Center Networking: A Survey

“…FTR, Unspecified Adopted dynamical flow learning (DTL) algorithm, weighted optimization based on Gaussian distribution and smooth mechanism based on difference estimation. et al[134] FTR, KNN, etc. ML classifiers based on different workload models proet al[135] FTR, SVM etc A method for classifying and characterizing data center workloads based on resource usage was proposed Google Cluster Trace (GCT)[136] and Bit Brains Trace (BBT)[137] et al[140] DL, DNN A low overhead, adaptive traffic classifier was proposed Historical flow data from a data center The unsupervised hierarchical clustering algorithm was used to classify the unexecuted jobs, and Davies-Bouldin and other indicators were used to evaluate the clustering quality ClusterData2011_2 [103], which is a data set released by Google Data Center Workload MSE, etc.…”