Data management oriented application

The candidate, in a Digital Data Library for the CRISISlab project, will have to setup an environment for the management of big data sets, mainly from the economic-financial sector, define ETL procedures in order to feed sql and no-sql databases. In particular the candidate will have to setup an ETL test environment based on virtual machines for a distributed filesystem and database, to write ETL procedures as a template for importing the original data sets into the final databases, to configure computation nodes inside a multicore environment, both with standard and GPU cores, and the associated installation of scientific libraries. The test environment chosen and documented by the candidate will be based either on standard hypervisor virtualization technique (VMvware, Citrix) or on container-based virtualization (OpenVZ, ZEROvm).Moreover a study of the most suitable libraries for the massive parallelization of algebraic code on GPUs. At the end a documentation of all the configuration and installation passages (packages, configuration files and drivers) for all the installed software is requested.

For the scientific applications programming skills, good mathematical skills will be required and candidates' knowledge will be evaluated specifically regarding:

• Programming Languages (C/C++, Python, R);
• Mathematical programming languages (Matlab, Octave, Scilab, R);
• Linux compilation techniques
• Networking and demons configuration;
• Virtual machines setting and administration.

1 month