An Analysis of Resource Utilization in Cloud Computing using Alibaba Cluster Trace

Abstract

—Cloud computing usage is rising day by day due to their ‘On Request’ service. Increasing the usage of cloud
computing, there is a need for better utilization and effective resource allocation. For better utilization, one needs to
apply new methods, real environment or setup, or there is a need for real-time cloud datasets. Alibaba group released
the dataset ‘Alibaba cluster trace’ in 2017, which is used in this paper. Alibaba cluster trace dataset consists of
machines, online services, and offline services. Dataset has 11101 Container events(online services) and 12935 offline
batch jobs located in 1300 machines over 12 hours. All services are completed in 5 minutes interval of 12 hours. The
paper ‘Imbalance in the cloud: An analysis of alibaba cluster trace’ describes the study of dataset in depth. Also,
reviewed other various research papers based on alibaba cluster trace dataset which is described in this paper and also
surveyed the various methods and algorithms. Assignment of the algorithms is done in this paper for better utilization on
‘alibaba cluster trace’ dataset.