Digital twins for solving management and development tasks of distributed data acquisition, storage and processing centers

Seminars

Laboratory of Information Technologies

Date and Time: Thursday, 28 March 2024, at 11:00 AM

Venue: Conference Hall, Laboratory of Information Technologies, online on Webinar


Seminar topic: “Digital twins for solving management and development tasks of distributed data acquisition, storage and processing centers”

Speaker: Daria Priakhina

Abstract:

A new method for creating and using digital twins (DTs) of distributed data acquisition, storage and processing centers (DDCs) is developed in the PhD thesis. The main goal of creating a DT is to improve DDC technical characteristics and to make decisions on choosing equipment configurations within the task of scaling and developing centers, managing resources and processes occurring in the DDC. The proposed method differs from existing ones in the ability to simulate such processes as data storage and processing, taking into account the characteristics of data flows and jobs, the probabilities of failures and changes in the equipment performance and other processes occurring in the simulated system. Algorithms for the implementation of the method and a data model are presented.

The results of the experimental operation of the software are demonstrated in the creation of the DT of the computing infrastructure of the BM@N Experiment and the computing system of the online data filter of the SPD Experiment of the NICA Complex. Recommendations on the results of the DT functioning are taken into account in the design and development of computing infrastructures for large-scale experiments in the field of high-energy physics. The results of the software application prove the efficiency and high quality of the models and algorithms developed in the thesis.

(based on the PhD thesis)