Abstract
To cope with real-time data analysis as the amount of data being exchanged over the network increases, an idea is to re-design algorithms originally implemented on the monitoring probe to work in a distributed manner over a stream-processing platform. In this paper we show preliminary performance analysis of a Twitter trending algorithm when running over BlockMon, an open-source monitoring platform which we extended to run distributed data-analytics algorithms: we show that it performs up to 23.5x and 34.2x faster on BlockMon than on Storm and Apache S4 respectively, two emerging stream-processing platforms. © 2013 Springer-Verlag Berlin Heidelberg.
Cite
CITATION STYLE
Simoncelli, D., Dusi, M., Gringoli, F., & Niccolini, S. (2013). Scaling out the performance of service monitoring applications with BlockMon. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7799 LNCS, pp. 253–255). Springer Verlag. https://doi.org/10.1007/978-3-642-36516-4_26
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
