Thesis Open Access

PIRE ExoGENI – ENVRI: Preparation for Big Data Science

Konstantaras, Stavros; Grafis, Ioannis

Thesis supervisor(s)

Oprescu, Ana; Zhao, Zhiming

Big Data is a new field in both scientific research and IT industry focusing on collections of data sets which are so huge and complex that create numerous difficulties not only in processing them but also in transferring and storing them. The Big Data science tries to overcome problems or optimize performancebased on the “5V” concept: Volume, Variety, Velocity, Variability and Value. A Big Data infrastructure integrates advanced IT technologies such as Cloud computing, databases, network and HPC, providing scientists with all the required functionality for performing high level research activities. The EU project of ENVRI is an example of developing Big Data infrastructure for environmental scientists with a special focus on issues like architecture, metadata frameworks, data discovery etc. 

In Big Data infrastructures like ENVRI, aggregating huge amount of data from different sources, and transferring them between distribution locations are important processes in the many experiments [5]. Efficient data transfer is thus a key service required in the big data infrastructure. 

At the same time, Software Defined Networking (SDN) is a new promising approach of networking. SDN decouples the control interface from network devices and allows high level applications to manipulate network behavior. However, most of the existing high level data transfer protocols treat network as a black box, and do not include the control for network level functionality. 

There is a scientific gap between Big Data science and Software Defined Networking and -until now- there is no work done combining these two technologies. This gap leads our research on this project. 

Files (1.2 MB)
Name Size
2014.masterthesis-1.pdf
md5:a2adffe83f24bae4e25cd5f5933e86e1
1.2 MB Download
74
30
views
downloads
All versions This version
Views 7474
Downloads 3030
Data volume 37.4 MB37.4 MB
Unique views 7373
Unique downloads 2929

Share

Cite as