Wednesday, December 26, 2018
LIT, Room 310
A.V. Nechaevskiy

Methods and simulation tools for distributed storage and processing systems of big experimental data for high-energy physics problems

(Materials of a PhD thesis)
Modern scientific experiments in the field of high energy physics require the development of systems for storing and processing big data streams that are based on grid and cloud technologies. Simulation tools are used to increase the efficiency of development and optimization of such systems. To increase the accuracy of simulation results it is proposed to use statistical work quality indicators of some real system. For these purposes the simulation program are combined with a real monitoring system of the grid-cloud service through a special database. Such an approach allows to use monitoring data for dynamic correction of model parameters. The software package SyMSim, developed in LIT on the basis of this approach for simulating data processing and storage systems, allows to carry out preliminary research of computing variants taking into account the load of resources and choosing the most efficient solution. After approving the proposed approach on the data obtained at the Tier1 CMS center in LIT, the advantages of this approach are shown on the example of BM@N infrastructure simulation.