Virtualized Big Data Benchmarks

Synonyms

Cloud big data benchmarks; Virtualized hadoop benchmarks

Definitions

Virtualized big data benchmarks measure the performance of big data processing systems, such as Hadoop, Spark, and Hive, on virtual infrastructures (Ivanov et al. (2014)) (The term virtual infrastructure could either mean on-premise or cloud infrastructure. While virtualization is not strictly necessary to support cloud computing, it is typically a foundational element of all major cloud computing services.). They are important for making quantitative and qualitative comparisons of different systems.

Historical Background

Big data applications are resource intensive and, as such, have historically been deployed exclusively on dedicated physical hardware clusters, i.e., bare-metal systems. As big data processing moved out of the specialized domain of Web 2.0 companies into the mainstream of business-critical enterprise applications, enterprises have been looking to take advantage of the numerous benefits of...