Abstract
Object storage systems are a flexible class of storage where data are architectured and stored in the form of units called objects. Objects may store structured, semi-structured or unstructured data, so it is well suited for big data in cloud environments. It was introduced particularly for academic purposes at Carnegie Mellon University and the University of California at Berkeley. Till then, large-scale file systems were practiced for their search engine. The earliest instance of object storage used in an organization was EMC’s Centera platform (2001). This article plays an important role in the industry and research in the field of understanding the available storage systems for supporting big data features, whereas our analysis of the object-based storage system is based on the classification of the storage systems and the comparative study of these systems’ architecture. It helps to distinguish the basic characteristics of object-based storage systems with respect to the implementation of these storage systems. Furthermore, it discusses the future perspective, research challenges and limitations.
